Taste and odor in conditioned flavor preference learning

Abstract

Almond and peppermint extracts were combined with salt and citric acid as cues in conditioned flavor preference conditioning. In Experiment 1, extracts overshadowed tastes, although tastes and extracts conditioned equally well when presented in isolation. In Experiments 2 and 3, tastes and extracts were conditioned in isolation prior to conditioning of a taste/extract compound. The conditioning history of the tastes and extracts did not affect the overshadowing of taste by extract. The results of Experiment 4 showed that rats could learn to discriminate between a taste and extract presented in isolation vs. the taste/extract compound. Thus, extracts do not interfere with sensing the tastes. We suggest that a taste/extract compound produces a configural stimulus that is more characteristic of the extract than the taste.

Cue interactions in flavor preference learning: A configural analysis.

“…In contrast to the results of the experiments reported by Holder (1991) and Capaldi and Privitera (2008) these experiments show a clear overshadowing effect in flavor preference learning. It was noted in the Introduction that the only previous demonstration of overshadowing of flavor preference learning, reported by Capaldi and Hunter (1994), could have been a consequence of greater consumption during training in their control than their overshadowing group carrying-over to the test. It is difficult to apply the same analysis to the current results for there was no indication during training of a difference between Group Overshadowing and Group Paired-Control in their consumption of the solution that contained either peppermint (Experiment 1A) or lemon (Experiment 1B).…”

“…They further noted that it is overly simplistic to account for their results in terms of differences in the salience of the taste or odor, but that rather taste/odor conditioning results in the formation of a configuration that is particularly similar to the odor element alone. Among other things, this proposal can explain the failure of Capaldi and Hunter (1994) to observe blocking of the odor by the taste because initial training to the taste would not generalize to the taste/odor compound sufficiently to reduce the error term for the compound. It can also explain the failure of the odor to block the taste because the failure to generalize between the odor/taste compound and the taste at test would mean that there was little or no conditioned response to the taste regardless of whether the odor had been paired with the US prior to compound training.…”

“…The idea that generalization between a compound and either of its two elements might differ, despite those elements being equally salient, would require that either one element shared more features with the compound than the other or that the G parameter can vary as a function which element is being compared to the compound. Unfortunately, it is not possible to make a principled statement of whether either of these possibilities would be true in the case of the stimuli used by Capaldi and Hunter (1994).…”

“…Capaldi and Hunter (1994) noted that conditioning of a taste/odor compound resulted in substantial generalization to the odor but not the taste when they were presented alone at test. They further noted that it is overly simplistic to account for their results in terms of differences in the salience of the taste or odor, but that rather taste/odor conditioning results in the formation of a configuration that is particularly similar to the odor element alone.…”

“…Capaldi and Hunter (1994) gave rats in a taste group paired presentations of one taste with sucrose and another taste with saccharin. Rats in an odor/taste group received pairings of one odor/taste compound with sucrose and pairings of another odor/taste compound with saccharin.…”