Acquisition of a spatially defined operant with delayed reinforcement.

Researchers put rats in a box with one lever. Each lever press earned food, but the food arrived 30 seconds later. A tone could sound right after the press. The team asked: Will the rats learn the lever press at all with this long wait?

They also tested if the immediate tone helps the rats connect press to payoff. Some rats got the tone every press. Others got only the delayed food. The scientists counted how fast each group learned to press.

Every rat learned to press, even without the tone. The 30-second delay did not block new learning. It just took more sessions than typical immediate-reinforcement studies.

The group with the immediate tone picked up the skill faster. The tone acted like a tiny bridge across the delay. Still, the rats without any tone reached the same final performance.

Davis et al. (1994) repeated the idea with Siamese fighting fish. The fish swam through a ring for delayed visual reinforcement. Both studies show the same core point: new responses can lock in even when payoff is late.

LeSage et al. (1996) added d-amphetamine to the same rat setup. The drug did not stop learning under 8-s delays unless the dose was so high it disrupted all movement. Together the papers stretch the finding across species and drug states.

Byrne et al. (2019) moved past acquisition. They showed that once the response exists, its duration also comes under delayed-reinforcement control. The 1993 study opened the door; later work walked through it.

If you run early intervention, remember that a reinforcer can still work even when it cannot follow the target within seconds. A brief praise or token delivered 20–30 s later can shape a brand-new skill, especially if you give an immediate bridge like a click or “nice.” The rat data say the bridge helps, but the absence of it does not doom the program. Keep teaching; the learner can still connect the dots.