"Facilitating creativity by regulating curiosity": Comment.

373 few published studies of either the positiveincentive value of food or of conditioned taste aversions in anorexics. Vitousek and Gray’s comments about the positive-incentive value of food in anorexics, although intended as an attack on our position, cry out for more research. Anorexics act in some ways as if they are experiencing aversions for food and in others as if food has a very high positiveincentive value—just the sort of ambivalent behavior that one would expect of starving subjects with widespread taste aversions. We did not suggest that taste aversions are the major causal factor in anorexia nervosa but rather that for some anorexics, taste aversions might make it easier for them to refrain from eating in the face of starvation. Vitousek and Gray (2002) responded by pointing out that the participants in the Keys et al. (1950) study managed to voluntarily starve themselves without apparent taste aversions. In contrast to most anorexics, however, the participants in the Keys et al. study were not continually encouraged to eat. In addition, they requested that they not be asked to perform functions that involved the handling of food “because the temptations under these circumstances were too great” (Keys et al., 1950, p. 832). Unfortunately, Vitousek and Gray (2002) did not comment on our conclusion that the consumption of meals by anorexics may be contraindicated. The disruptive postingestion effects of meals in starving subjects are well documented. Our suggestion that adverse effects, such as widespread conditioned taste aversions, might be produced in anorexics by meals has immediate implications for the design of treatment programs for anorexia, which often have as their major goal the consumption of meals.