Delayed discrimination and delayed matching in pigeons.

Smith (1967) tested pigeons on two tasks: delayed discrimination and delayed matching. The birds had to remember a color or line tilt for a few seconds. The delay between the sample and the choice grew longer across trials.

The goal was to see which task broke down first as the wait got longer.

Accuracy dropped for both tasks as the delay increased. Delayed matching fell apart faster. Birds still picked the right key in simple discrimination even after a long wait, but they forgot the matching sample sooner.

Gentry et al. (1980) built on this by letting pigeons choose between two future reinforcers. They showed that absolute delay, not just the gap, steers choice. The 1967 paper first mapped the decay curve; the 1980 paper used it to predict preference.

Weil (1984) moved the same delay idea into everyday responding. Pigeons pecked less when every reinforcer was late, even if food stayed frequent. The 1967 finding—delays hurt performance—now explained why free-operant response rates sank.

Horner-Johnson et al. (2002) later copied the decay shape with humans and money. Hyperbolic discounting that started in the pigeon lab fit college students picking $20 next week versus $40 next month. Animal work from 1967 became the blueprint for human delay research.

When you add a wait between instruction and consequence, expect faster forgetting in conditional-discrimination programs such as listener responding or match-to-sample. Start with short delays and thin them slowly; the 1967 curve tells us matching tasks will need more review than simple discrimination. If you see sudden drops in accuracy, check the delay first—it may be the lone variable erasing the skill.