Fixed-interval behavior maintained by conditioned reinforcement.

The team worked with pigeons in a small chamber. A key lit up for half a second each time a timer ended.

No food appeared during these brief flashes. Food only came after many flashes. The birds still pecked the key in a smooth scallop pattern.

The birds paused right after food, then pecked faster as the next food time neared.

The short light alone kept this pattern going. It acted like a tiny reinforcer even though it was not food.

Gardner et al. (1977) later used the same setup but swapped food for cocaine at the end. The birds still showed the same scallop, proving the brief flash works with many kinds of final pay-offs.

Nevin (1969) mapped the classic two-state pattern: long pause, then sudden jump to fast pecking. Gibbon (1967) shows this pattern can live on conditioned reinforcers alone.

Downing et al. (1976) mixed food and flashes in random order. They found random mixing made the flash act even more like food. Gibbon (1967) is the base case—flash with no food—that their sequencing builds on.

You can keep a learner on track with tiny, cheap cues. A thumbs-up, a beep, or a brief visual can bridge long gaps before the real prize. Use brief praise or tokens during DTT or while waiting for a break. The learner stays in rhythm without constant edible feed.