Reversal of baseline relations and stimulus equivalence: I. Adults.

The team asked: what happens to equivalence classes when we flip the original training rules? They worked with neurotypical adults in a lab. First they taught A-B and B-C match-ups until classes formed. Then they reversed the A-B baseline so A1 now pointed to B2. Finally they probed symmetry and transitivity to see what survived.

Symmetry probes tracked the new baseline: adults quickly matched B1 to A2. Transitivity and full-equivalence probes mostly stayed with the old classes: participants still picked C1 after A1 even though A1-B1 was no longer reinforced. The result shows a split: some relations bend with new contingencies while others hold firm.

Mace et al. (1990) saw the same split five years earlier, so the adult pattern is stable. Lerman et al. (1995) part II tested kids aged 5-7 with the same steps and got the opposite outcome: most children's equivalence classes collapsed after the reversal. The studies seem to clash, but the method is identical—only age differs—so the contradiction points to a real developmental gap, not a flaw.

Arntzen et al. (2018) later showed that adding a brief delay-training warm-up helps adults reorganize classes faster. That finding extends the reversal idea by giving us a tool to speed up the change when we want it.

If you teach conditional discriminations to adults, do not assume later reward changes will wipe out all emergent relations. Probe transitivity after any contingency shift; it may still be intact and could interfere with new goals. With young learners, expect the opposite—monitor carefully because their classes can fall apart. When you need flexible classes in adults, borrow the Arntzen delay warm-up to make reorganization easier.