Abstract
Twenty to seventy per cent of the reinforcements scheduled for pigeons' fixed‐ratio 80 performances were replaced by a 4‐sec timeout. Pauses after reinforced ratios were unchanged at 80% reinforcement, but were lengthened at lower reinforcement percentages. Pauses after nonreinforced ratios were shorter than post‐reinforcement pauses. When 50% of the reinforcements arranged by a variable‐interval 60‐sec schedule were replaced by a 4‐sec timeout, pauses after reinforcement omission increased. Both frustrative nonreward and reinforcement aftereffects notions can explain the fixed‐ratio results; neither easily explains the variable‐interval data.