FLAVOR-FLAVOR AND COLOR-FLAVOR CONDITIONING IN HUMANS

Pavlovian flavor-flavor associative learning has been suggested as one of the determinants of humans' food likes and dislikes (e.g., Rozin & Zellner, 1985). However, few experimental studies exist which directly examined this proposal. In the present study, the hypothesis was tested that the contingent presentation of a neutral flavor (CS) with an already (dis)liked flavor (US) should result in an increase (or decrease) in liking for the originally neutral flavor. Second, the question was addressed whether or not a perceptual characteristic of food other than its flavor (c.q. its color) may also acquire evaluative value when paired with an appetitive or an aversive flavor US. Finally, the relationship was studied between evaluative conditioning and subjects' explicit knowledge of the stimulus relationships presented during the conditioning phase. In a differential conditioning paradigm, subjects drank 12 CS+ US compound fluids and 12 CS- fluids. Sugar was used as a positive US, and Tween20 as a negative US, while either the flavor or the color of the drink was scheduled as the CS. As indexed by postconditioning visual analog scale ratings, a strong differential evaluative conditioning effect was observed in the flavor-flavor, negative condition. The evidence for positive flavor-flavor conditioning was weak at best, and no evidence was obtained for color-flavor conditioning. On the other hand, none of the subjects in flavor-flavor groups evidenced any explicit knowledge of the CS-US relation, whereas about half of the color-flavor subjects did so. These results are related to the theoretical issues of positive/negative asymmetry, selective associations, and implicit/explicit knowledge dissociations. © 1990.