Memory for Asymmetric Events

The study of memory for event duration in pigeons has, for the most part, centered on an unusual behavioral phenomenon and an interesting theoretical explanation for it. A typical experiment on memory for event duration involves trials initiated by a sample stimulus that is either short or long in duration (e.g., 2 vs. 10 sec of houselight) and that is later followed by the simultaneous presentation of two choice stimuli (e.g., red and green). A response to one stimulus (e.g., green) is reinforced on trials initiated by the longer sample stimulus, and a response to the other (red) is reinforced on trials initiated by the shorter sample stimulus. The unusual finding is that the forgetting functions following the two samples are reliably asymmetric, in that pigeons tend to choose the alternative associated with the short sample following a lengthy retention interval, regardless of which sample was actually presented. As a result, performance following the long sample decreases rapidly to well below 50% correct as the retention interval increases, whereas performance following the short sample remains accurate at well above 50% correct (Fetterman, 1995; Spetch, 1987; Spetch & Wilkie, 1982, 1983). This increasing bias to choose the alternative associated with the short sample as the retention interval increases is called the choose-short effect.