Reversing the course of forgetting

Abstract

Forgetting functions were generated for pigeons in a delayed matching-to-sample task, in which accuracy decreased with increasing retention-interval duration. In baseline training with dark retention intervals, accuracy was high overall. Illumination of the experimental chamber by a houselight during the retention interval impaired performance accuracy by increasing the rate of forgetting. In novel conditions, the houselight was lit at the beginning of a retention interval and then turned off partway through the retention interval. Accuracy was low at the beginning of the retention interval and then increased later in the interval. Thus the course of forgetting was reversed. Such a dissociation of forgetting from the passage of time is consistent with an interference account in which attention or stimulus control switches between the remembering task and extraneous events.

Keywords: delayed matching to sample; forgetting functions; pigeons; remembering; retroactive interference; short-term memory.

Fig. 1

Mean proportion correct as a function of retention interval for conditions with a dark retention interval (filled circles), the houselight (HL) lit throughout the retention interval (filled triangles), and the houselight lit for the first 1.5 s or 3 s of the retention interval (unfilled circles and triangles, respectively) in Experiment 1 (left panel, n = 5) and Experiment 2 (right panel, n = 4). Error bars are standard errors of the mean.

Fig. 2

Mean proportion correct as a function of retention interval for conditions with a dark retention interval (filled circles), the houselight lit throughout the retention interval (filled triangles), and the houselight lit for the first 1.5 s (left panels, unfilled circles) or the first 3 s (right panels, unfilled triangles) of the retention interval, or from 1.5 s into the retention interval (center panels, unfilled squares), for individual pigeons in Experiment 1. Mean proportions correct were based on eight sessions for each condition and were averaged over replications of the baseline condition with dark retention intervals. Note. The same functions for the dark retention interval and with the houselight lit throughout the retention interval are plotted in all three sets of panels.

Fig. 3

Measures of discriminability, Log d, as a function of retention-interval duration for individual pigeons in Experiment 1, for the condition with the houselight lit for the first 1.5 s of the retention interval (unfilled circles), and with exponential functions in the square root of time fitted to the data for conditions in which the retention interval was dark (upper, filled circles) or in which the houselight was lit throughout the retention interval (lower function, filled triangles).

Fig. 4

Proportion correct as a function of retention interval for conditions with a dark retention interval (filled circles), and with houselight lit for the first 3 s of the retention interval (unfilled triangles), or throughout the retention interval (filled triangles), for individual pigeons in Experiment 2.