Nervous Systems 843 



cific sequence. After repetition of Si~^Ri a second stimulus, S-, may now 

 elicit its own response, R2, although before the repetition of Si, stimulus So 

 failed to evoke any overt response. A weak tactile stimulation to a rat, for 

 example, mav elicit no motor response unless it follows strong visual stimu- 

 lation. The stimulation by light lowers the threshold for tactile stimulation, 

 even though the responses are different. A response which has been extin- 

 guished by repetition can be "disinhibited " or brought back by generalized 

 sensory stimulation. One sense modality directly modifies the response to 

 another. 



3. In pseudoconditioning the stimulus S2 comes to elicit the same response, 

 Ri, as does Si. In some instances the stimuli are not presented together. In 

 other instances they are presented in combination. The conditioned response 

 (RC) is very similar in latency and form to the unconditioned response 

 (RUnC); persistence is limited. Examples are conditioning of the knee jerk 

 or of the pupillary reflex. Trial or error learning by lower animals and much 

 "instinctive" behavior appear to depend on pseudoconditioning mechanisms; 

 this type of modification of behavior is dehnitely limited by existing neural 

 connections. 



4. True conditioning depends on temporal contiguity of the uncondi- 

 tioned stimulus UnCS, and the conditioned stimulus CS. The RCS differs 

 somewhat from RUnCS; it has a longer latency, and the new response may 

 persist for days unless it is extinguished. The RCS may be anticipatory in 

 character. Conditioned salivary reflexes of the classic Pavlov type and the 

 neural modifiability for maze learning and problem-box solving (instrument- 

 al conditioned reflexes) represent true conditioning 



5. The most complex behavior changes require motivation separated in 

 time and space from the CS. Such behavior is "insightful," and the animal 

 responds to a new situation, such as a multiple choice problem, in the light 

 of memory of complex past experience. Complex learning may involve sym- 

 bolic processes, as in reasoning. 



The preceding classihcation is arbitrary and has no experimental phy- 

 siological basis, but it may provide a background for considering different 

 kinds of animal "learning." Three theories of conditioning have been pro- 

 posed: (1) New synaptic connections mav be established during condition- 

 ing; the time required for conditioning is so brief that morphological change 

 seems unlikely. (2) Persistent synaptic facilitation may occur; electrical 

 studies have not given evidence for long-lasting changes in neurone excit- 

 ability. (3) Conditioning may start conduction in reverberating chains and 

 thereafter impulses pass continually around the chain, giving off motor dis- 

 charges as called for; opposed to this is the fact that memory continues after 

 anesthesia. None of these hypotheses is satisfactory for long-term learning, 

 and a new approach is needed if the physiology of conditioning is to be un 

 derstood. Examination of the kinds of conditioning which can occur in dif- 

 ferent animals and in various nerve centers may aid in understanding the 

 mechanisms of persistent responses. 



Neural Basis for Conditioned Responses in Animals. In Protozoa altera- 

 tion in sign and rate of response to repeated stimulation has been reported; an 

 amoeba avoids a bright light by decreasing pseudopodial contacts,-^^ and a 

 Paramecium turns around in a capillary tube with fewer partial turns after 



