Formation of the sameness-difference concept by Japanese monkeys from a small number of color stimuli.

Fujita (1983) taught monkeys to tell same from different using only a few colored cards.

The monkeys played a matching-to-sample game. A center card appeared. They had to pick the card that looked the same from two side cards.

The twist: reinforcement came on a variable-interval schedule inside each trial. The longer the monkey waited, the more likely a peck would pay off.

The monkeys learned the sameness-difference concept faster and transferred it to new colors better than monkeys trained with the usual fixed-peck contingencies.

In plain words, the timing of rewards, not just getting them, sharpened the concept.

Nakagawa (2005) later showed rats also build untrained relations after many-to-one matching, proving the process crosses species.

Brown et al. (1994) pushed the idea further in pigeons, finding full transitivity—another sign that equivalence is a basic behavioral process, not just a human trick.

Paul (1983) ran a similar pigeon study the same year, but swapped in ratio requirements instead of VI schedules. Both papers show that the contingency itself can act like a cue, yet K’s VI boost is unique for concept speed.

If you run matching-to-sample programs with learners who stall, try slipping in brief VI schedules instead of immediate reinforcement. Start with short, unpredictable intervals—two to five seconds—then stretch them as concepts firm up. The monkey data say this small shift can cut training time and improve transfer to new stimuli, saving you sessions and keeping motivation high.