ABSTRACT
In the study of discrimination learning, there is a long and rich tradition of inquiry concerning the extent to which behavior is controlled by the absolute or the relative characteristics of stimuli. Historically, this controversy focused on transposition experiments. The controversy over absolute and relational factors in discrimination learning continues, but the focus of the controversy has shifted. The extent to which nonhuman animals possess and use abstract relational concepts is one of the primary foci of the controversy over absolute and relational control. A closer examination of the manner in which durations were determined and of the resulting distribution of duration pairs shows why some predictability by absolute duration remained. The explicit arrangement of a number of duration pairs for which absolute duration of one of the pair members is uninformative about duration ratio is responsible for the lower maximum accuracy achievable by completely nonrelational means.
