Date of Award


Document Type


Degree Name

Doctor of Philosophy (PhD)




Multiple unit recording is a relatively new electrophysiological recording technique which records spike potentials from many neurons. In the Halas laboratory and others, systematic changes have been observed during classical conditioning but none under other conditioning procedures . Recently, some have interpreted changes in evoked potentials and multiple unit potentials as due to emotional arousal (fear) and thus related more to a conditioned emotional response (CER) than to conditioning per se. If a CER occurs, and not a learning modification, then during classical aversive conditioning a high sustained neuronal response ought to occur. This would reflect the high degree of emotionality displayed behaviorally by the animal throughout classical conditioning. On the other hand, during instrumental avoidance conditioning a decreasing level of neuronal response ought to occur which would reflect the reduced emotionality displayed during this kind of conditioning procedure.

Thirteen permanent macroelectrodes were implanted in each of seven cats. A total of twenty sites were implanted in the classical auditory pathways, the reticular system (myelencephalon, mesencephalon, and dien- cephlon), the limbic system, and the cortex. Standard electrophysiological techniques were used- to display the multiple unit activity on an oscilloscope. The tracings were photographed and then were judged as to the level of their neuronal activity. A three day running average was computed which served as the neuronal dependent variable. Counterbalanced classical aversive, instrumental avoidance, and discrimination conditioning procedures were applied. A 1500 hz tone served as the conditioned stimulus and a mild shock to the right hindpaw served as the unconditioned stimulus.

For all sites taken together a significant linear negative neuronal trend emerged for classical aversive conditioning, and a significant linear positive neuronal trend emerged for instrumental avoidance conditioning. Moreover, the neuronal responses for classical extinction decreased further, and the neuronal responses for instrumental extinction increased significantly. These results do not bear out the prediction stated earlier. The emotional arousal hypothesis is untenable since one would expect the opposite kinds of trends. Inflections in the neuronal trends in the auditory sites during the two types of conditioning suggested that a neural associative process may have occurred in these sites. The neuronal activity of the reticular structures tended to follow the increasing neuronal trends found in the auditory centers during instrumental conditioning but not during classical conditioning. Some high positive neuronal-behavioral correlations (as high as .96) occurred during instrumental avoidance conditioning and some high negative correlations (as high as -.88) occurred during classical conditioning. Due to equipment problems little invariance across cats occurred for discrimination learning; and thus, the data was not presented.

As the conditioned stimulus and the unconditioned stimulus for the two types of training procedures were kept as congruent as possible, the pervasive, remarkably different neuronal patterns which resulted imply that classical aversive and instrumental avoidance conditioning share very little in common in the CNS. Apparently, two quite different reactions occurred in rather widespread areas of the brain. Clearly, a CER cannot account for the trends in the data found in this study. Nor is the explanation for these contrasting results, particularly during extinction after classical and instrumental conditioning, readily available through any other theory of learning known to the experimenter. The inflection points in the neuronal trends in the subcortical auditory sites appeared to occur first in the lowest center (cochlear nucleus) and latest in the highest center (medial geniculate) for both types of conditioning (one exception occurs in the inferior colliculus during classical conditioning). These data suggest that the subcortical auditory nuclei function as a unit during learning. There is already considerable data showing that these same nuclei function as a unit dur- in the transmission of sensory information which therefore indicate that these nuclei are capable of mediating several quite different functions .