A demonstration of auto-shaping with monkeys.

Researchers paired a light with food for monkeys. The monkeys never had to press a key to get the food. The team just turned the light on, then gave food. They wanted to see if the monkeys would still learn to press the key.

This is called auto-shaping. It was first done with pigeons. The 1968 study asked, 'Does it work in monkeys too?'

All the monkeys started pressing the key. They did it even though pressing was not required. The light-food pairing alone created the new response.

The monkeys also learned to press only when the 'correct' light appeared. They showed both acquisition and simple discrimination.

Mulvaney et al. (1974) ran almost the same study six years later. They got the same key-press acquisition, so the effect is repeatable. But they added a twist: when food no longer followed the key-press, most monkeys stopped pressing. That follow-up shows the first study's acquisition result holds, but maintenance needs more work.

Neuringer (1973) had already shown auto-shaping in pigeons. Hart et al. (1968) simply moved the procedure to monkeys, proving the effect crosses species. Kirby et al. (1981) went further and showed it even works in rats, though the response changed from key-press to lever-nose.

Together these papers build a bridge: pigeons → monkeys → rats. Auto-shaping is not just a bird trick. It is a general process.

If you run stimulus-stimulus pairing with learners, remember this monkey data. You do not need to shape every step. Just pair the target stimulus with a preferred item. The learner may acquire the response without direct prompting. Watch for the first untaught attempts and then reinforce them. This can save you minutes of trial-by-trial shaping each session.