Abstract
We generated analogies based on observed spatiotemporal relations. In Experiment 1, six adults observed arbitrary stimuli in a spatiotemporal sequence (A1 followed by A2, A3, A4, A5 and then A1). Then they were taught to select the first stimulus following the sample in the presence of one contextual cue X (e.g., given A2, select A3) and to select the stimulus immediately preceding that stimulus in the presence of another contextual cue Y (e.g., given A2, select A1). Finally, they received a conditional discrimination (CD) probe with three-stimuli samples (3-SS-CD) in which the sequential relation between the first two stimuli set the occasion for selecting a comparison with the same relation to the third stimulus (A2A3 A1, select A2). Most participants demonstrated emergence. In Experiment 2, the procedure included a second set of B stimuli and a CD probe with A and B stimuli (i.e., A4, A5, B2, as sample and Bs as comparisons). All eight participants demonstrated emergence. Participants also observed new sequences with novel stimuli, without X or Y, and demonstrated emergence of the 3-SS-CD. The results demonstrated a type of analogical responding close to that observed in traditional analogy tasks and found basic learning processes involved in it.