is in behaviors which depend directly upon short-term memory. These may range from phenomena such as the digit span on an intelligence test to the solving of complex problems requiring the individual to keep in mind several aspects of the problem in order to work it out.
Finally, even though the theory does not attempt to describe long-term retention processes, it does tend to ascribe the inadequacy to the process of acquisition and to the immediate stimulus control of behavior. Thus, from the standpoint described here, long-term retention, a process seemingly closely related to the acquisition process, is viewed as "normal" in the retardate, i.e., equivalent to the process in the individual without CNS pathology. This issue, of course, deserves further attention. Perhaps the etiology of the pathology may be of importance. An individual who has had localized brain damage may have difficulty learning; but once the information is stored, presumably in undamaged tissue, we might expect the retention of the information to be adequate. On the other hand the individual with an inadequate CNS owing to genetic factors or endocrine disorder may possess an inadequate storage facility attributable to biochemical or structural characteristics. This amounts to nothing more than speculation. However it does point to meaningful questions for behavioral research involving etiological variables.
Caution should be exercised in attempted application of the findings presented here. Even though they appear germane to such situations as classroom training, the wise practitioner will await further empirical evidence. There is an appreciable gap between theory derived and tested in the laboratory and the application of the finding to the classroom, for example. In the writer's view the gap should be spanned by further research in which the real life situation is simulated more closely. Too often in an area in which the problems seem insurmountable, there is premature application of theory which leads to wasted effort and occasionally to disastrous results.
The research vistas suggested here do not require commitments to the use of st or ni, or to the acceptance of the reverberatory circuit concept. Certainly, clear evidence against the last would not damage the position which is essentially molar. As Sidman has aptly put it, "the importance of data is not affected by the sophistication of the hypothesis that may have generated the experiments" ( 1960, p. 6).
A theory employing two notions, stimulus trace st and integrity of the CNS ni, has been formalized to account for differences between normal and mentally defective humans in behaviors dependent upon short-term memory. st was defined by an environmental stimulus S antecedently, and by a behavioral event B consequently. ni, defined by IQ or other indices of behavioral inadequacy, was introduced as an individual difference construct. The central problem outlined involved the establishment of the relationship between st and ni.
It was hypothesized that the duration is shortened and the intensity of st lessened in the subnormal organism, i.e., the organism with lower ni. A