Because of very real practical constraints, conditions in animal shelters are often reminiscent of those in early primate deprivation studies. Dogs are frequently surrendered to shelters because of behavior problems, and aspects of the shelter environment may induce anomalous behavior, increasing the chances that adopted dogs will be returned to the shelters. Comparative psychologists, psychobiologists, and other behavioral scientists possess the knowledge and techniques to help shelters intervene in this cycle. Experience suggests human interaction and the application of basic conditioning procedures can reduce the impact of the shelter environment, and ease the transition into the adoptive home. A program developed to meet these goals is described. Shelters can provide opportunities both for the training of students in animal-related exercises and for limited applied research. Behavioral scientists stand in a unique position to help transform conditions in animal shelters to the benefit of all involved.Approximately 15 million dogs are either turned out as strays or released to animal-welfare agencies by their owners in the United States each year (Moulton, Wright, & Rindy, 1991). These agencies, facing an endless influx of discarded pets, are able to place only a small percentage of homeless dogs (Moulton et al., 1991). Moreover, during their stay in even a modern, well-run shelter, dogs are subject to a variety of psychological stressors, including novelty, isolation from any former attachment figures, exposure to unpredictable and often intense noise, disruption of familiar routines (such as walks for elimination), and a general loss of control over environmental contingencies. These are precisely the types of events known to activate stress-related physiological systems, particularly the hypothalamic-pituitary-adrenal (HPA) axis, in laboratory ani-

No abstract

The research of this article investigates the possibility that memory for the conditioned response (CR) may be subject to the same sorts of interference that have been found to operate in verbal and in perceptual-motor memory situations. If interfering experiences can indeed contribute to the forgetting of a CR, then the need to postulate a special memory system for this class of behavior would be seriously questioned, and we might move in the direction of a general theory of memory.In the classical conditioning paradigm used, cats that were suspended in a webbed sling received shocks on either forepaw. A shock was the unconditioned stimulus, flexions of one paw or another were the unconditioned responses, the conditioned stimulus could be either a light or a tone, and the CR was a movement of the appropriate paw. By manipulation of the two conditioned stimuli and the two parallel unconditioned responses, we devised and introduced different kinds of potentially interfering experience. Experiment 1, a retroactive design, used the various stimulus and response manipulations of verbal-learning research and determined that the CR did indeed show retention loss when interference was introduced. Experiment 2, a proactive design, measured the effect of prior, random, and uncorrelated presentations of shock stimuli and lights or tones on later conditioning and retention and found that the uncorrelated experience had a profound effect on retention tested 10 weeks after the end of original conditioning, although it had little impact upon acquisition of the CR.The Discussion sections consider theoretical implications of findings as they relate to several topics: (a) the operation of interference paradigms in conditioning that are parallel to those in other types of learning, (b) generality of the concept of spontaneous recovery, (c) transfer effects as functions of experimental manipulations, (d) the implications for training and therapy that lie in our interference data, (e) the possible contrast between the effect of correlated experience on memory and that of a random noncorrelated experience, (f) the influence of a prior random schedule upon subsequent learning and especially upon memory, and (g) the relation of our paradigm to the typical learned helplessness phenomenon.

Three experiments investigating memory for the conditioned response (CR) as a function of proactive inhibition are reported. The proactive operation was the preexposure to quasi-random presentations of the potential CS and UCS. The subjects were cats suspended from a sling with their legs lightly restrained. The possible conditioned stimuli (CS) were light and tone, and the the unconditioned stimuli (UGS) were brief mild shocks to either the right or left paw, which produced a brisk leg jerk.In Experiment 1 all possible combinations of the CS and UCS components of the eventual CR were also present in the preexposure period for one or another group as in the traditional interference paradigms of human paired-associate memory research. The results may be summarized by stating that if any component of the eventual conditioning had occurred in the preexposure period, retention 10 weeks after criterion attainment was drastically reduced. Some inhibition was also found if both components differed as in the A-B, C-D verbal design.Experiment 2 demonstrated that the decline cannot be attributed to a strategy type of interpretation that asserts that when the retention-extinction situation occurs, the cats "backward scan" and judge themselves to be once again in the preexposure period. Performance immediately after reaching the conditioning criterion did not differ between the control group that experienced no preexposure and the experimental group, but it did so after the usual 10-week retention interval.Experiment 3 investigated the role of context in the memory deficits by maintaining the same context in the preexposure, conditioning, and memory test situations or giving the preexposure experience in an environment different from the other two situations. Context change greatly reduced but did not entirely eliminate the proactive inhibition.The following four principles serve to handle the results; of all three experiments: (a) Acquired indifference. An attitude of indifference results from the repeated presentation of stimuli that predict nothing and/or are not predicted, (b) contextual effects. The attitude of indifference will become attached to the environmental context as well as to the specific phasic stimuli, (c) suppression of acquired indifference. Given predictable relationships between stimuli, a new response can be learned, and the attitude of indifference is suppressed, (d) the primacy effect. Other things being equal, initial experiences with stimuli have a priority over subsequent contrary learning. Although, they may be suppressed, they will, across time and in the absence of contrary learning, reassert themselves.It is concluded that the CR is readily forgotten given appropriate interference and does not differ from other kinds of learning in this respect. Furthermore, there is a close parallel between the last two principles above and those stated for human paired-associate learning by Postman, Stark, and Eraser (1967).

To investigate stimulus selection as a function of the time between the onsets of the longer and shorter elements of a compound stimulus (tone and light), the paw flexions and galvanic skin responses of 48 cats to shock were classically conditioned. With modality counterbalanced, each of 3 groups of cats experienced a different CSi-CSz interval: 2,000, 500, and 150 msec. The CSa always preceded shock by 500 msec.; all stimuli terminated together. After 8 wk. of conditioning, cats were tested to either the long or the short stimulus alone. Response strengths of the elements were clearly functions of the CSi-CSa intervals, the shorter stimulus being more effective at the 2,000 and 150 msec, intervals, the longer stimulus at the 500. Redundancy interpretations cannot account for all results.