Open AccessFailure of Retrospective Inference in Rats' Taste Aversion[Note 1. Each of us was financially supported by the JSPS ...]
Author(s) -
Nakajima Sadahiko,
Kawai Nobuyuki
Publication year - 1997
Publication title -
japanese psychological research
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.392
H-Index - 30
eISSN - 1468-5884
pISSN - 0021-5368
DOI - 10.1111/1468-5884.00042
Subject(s) - retrospective cohort study , inference , element (criminal law) , taste aversion , confounding , presentation (obstetrics) , taste , psychology , computer science , medicine , artificial intelligence , surgery , neuroscience , law , political science
Three experiments explored the possibility of retrospective inference in the rat. Experiment 1 revealed that poisoning of an element of a taste compound after single compound poisoning enhanced aversion to the other element, and that presentation of an element without poison after the compound poisoning reduced aversion to the other element. These results were opposite to those predicted by retrospective inference. Experiment 2 eliminated some confounding variables and examined the effect of element poisoning after compound poisoning. The result again was opposite to what the retrospective view predicts. The results, however, accorded with the idea that within‐compound learning was established during compound presentation, and subsequent poisoning or nonpoisoning of one element affected response to the other element via the within‐compound learning. In Experiment 3, the possibility of within‐compound learning was reduced by using sequential presentation of tastes, but there was no indication of retrospective inference even under this condition.