Suppression of random-ratio and acceleration of temporally spaced responding by the same prereward stimulus in monkeys.

Kelly (1973) worked with monkeys who pressed a lever for food. The team added a one-minute tone that always ended with a pellet.

Some monkeys earned food on a random-ratio schedule. Others got food on a slow, fixed-time schedule. The same tone played for both groups.

The tone cut lever presses almost in half for the busy random-ratio monkeys. It sped up the slow fixed-time monkeys instead.

The slower a monkey was before the tone, the bigger the boost it got. One stimulus did opposite things depending on baseline rate.

Hartmann et al. (1979) saw the same double effect in pigeons. Their tone slowed fast key pecks but kept slow ones going, even when pecks cancelled food.

Lancioni et al. (2009) flipped the story. They showed a food-paired light can create brand-new screen pecks in calm monkeys. D’s monkeys already knew the lever, so the tone only tuned speed up or down.

Dardano (1970) used shock instead of food and still got bidirectional control. Short warning flashes cut avoidance hits; long flashes raised them. Same lab, same idea: stimulus timing plus baseline rate sets the direction.

Before you add a signal to any program, clock the client’s current rate. A high-rate stem behavior may drop when you pair a bell with candy, while a low-rate one may surge. Watch both directions and adjust reinforcement density, not just the signal.