Search

Rats tend to prefer flavors previously consumed under low deprivation to flavors previously consumed under high deprivation (Capaldi & Myers, 1982). We attempted to distinguish among possible associative explanations by determining whether this conditioning phenomenon was based upon conditioned preferences, conditioned aversions, or both. We compared preference for flavors presented exclusively under either high or low deprivation with preference for a neutral flavor. In Experiments 1A and 1B the neutral flavor was one that had been randomly paired with both high and low deprivation, whereas in Experiments 2 and 3 the neutral flavors had not been associated with either high or low deprivation. Our results strongly suggest that this conditioning phenomenon is based upon an actual increase in preference for the flavor consumed under low deprivation rather than on any form of aversion conditioning.

In five experiments, rats’ preference for a flavor was greater if the flavor had previously been consumed under low rather than high deprivation. This preference was conditioned in as few as three flavor-deprivation pairings (Experiment 1), and persisted through 28 test days, half under each deprivation level (Experiment 2). Rats never preferred the flavor associated with high deprivation even when calories were increased by giving 40 ml of 8% sucrose or when caloric density was increased to the equivalent of 20% sucrose. The preference for the low-deprivation flavor was greater when saccharin solutions were used rather than sucrose solutions, but the preference did emerge when sucrose solutions were used as testing proceeded and when a lower concentration of sucrose was used. We suggest that these preferences may be a result of flavor-taste associations rather than associations between flavors and postingestive consequences, and that the taste of the solutions under low deprivation is preferred to the taste under high deprivation.

In three experiments, counterconditioning was found to reduce fear less effectively than extinction. In Experiments 1 and 2, the resistance to extinction of avoidance was greater if food was given during extinction of fear to the CS than if no food was given, even when exposure to the CS and numbers of food and no food confinement trials were equated. It is suggested that these results could be attributed to contextual control of fear extinction by the food cue and/or to frustration produced by removing food for the counterconditioning group. Experiment 3 also found counterconditioning to be less effective than extinction and provided evidence that this difference occurs because of contextual control of fear extinction by the food cue. Measuring conditioned suppression of licking, in a test with no food present, less fear was shown if no food had been present during fear extinction, and greater fear was shown if no food had been present during fear conditioning. These results indicate that food is an important part of the context controlling fear and fear extinction. It is suggested that there may be no unique counterconditioning process. Rather, when counterconditioning procedures are employed, rules governing interference paradigms in general may apply. Thus, in a test for fear following counterconditioning, fear will be shown to the extent the test situation is similar to that in which fear conditioning occurred rather than that in which fear reduction occurred.