The formation of linked perceptual classes.

Fields et al. (2002) taught adults to link two kinds of pictures. One group saw fill patterns. The other group saw satellite photos.

They used many examples of each picture pair. This is called multiple-exemplar training. Then they tested if new mixes of pictures would also match.

Nine times out of ten, the adults picked the correct new pairs without more teaching. Seven of those times they got it right on the first test. Two other times they got it right after a short break.

The classes formed just like equivalence classes, even though the pictures looked nothing alike.

Arntzen et al. (2018) later showed that adding a six-second delay and using real photos boosts success. Their adults reached a large share class formation, triple the rate seen with plain shapes alone. This builds on Lanny’s work by showing small tweaks make big gains.

Lantaya et al. (2018) used a simpler go/no-go task and still got equivalence. Together these studies tell us the brain links sights in flexible ways, not just through classic match-to-sample.

Tullis et al. (2021) moved the idea into a classroom. Three autistic children learned categories with instructive feedback and then answered untrained questions. The same emergent learning seen in Lanny’s lab can happen at school.

You can teach broad categories without drilling every mix. Use lots of varied pictures, add a short delay, and let emergence do the rest. This saves hours of trial time and builds flexible, real-world understanding for any learner.