Mnemonics for variability: Remembering food delay

Three experiments with White Carneaux pigeons (Columba livia) investigated memory and decision processes under fixed and variable reinforcement intervals. Response rate was measured during the unreinforced trials in the discrete-trial peak procedure in which reinforced trials were mixed with long unreinforced trials. Two decision models differing in assumptions about memory constraints are reviewed. In the complete-memory model (J. Gibbon, R. M. Church, S. Fairhurst, & A. Kacelnik, 1988), all interreinforcement intervals were remembered, whereas in the minimax model (D. Brunner, A. Kacelnik, & J. Gibbon, 1996), only estimates of the shortest and longest possible reinforcement times were remembered. Both models accommodated some features of response rate as a function of trial time, but only the second was compatible with the observed cessation of responding.