Do contingency‐conflicting elements drop out of equivalence classes? Retesting Sidman's (2000) theory

Silguero and team asked a simple question. Do conflicting reinforcers get kicked out of an equivalence class?

They taught adults to match symbols in the usual way. Then they added a twist. Some symbols now earned points while others lost points.

The key test came next. They checked if the money-losing symbols still belonged to the class. They tried training the money part first, then the matching part. They also tried the reverse order.

The bad-news symbols stayed in the class every time. Training order made no difference.

Even when people learned "this picture costs me money" first, they still treated it as equal to the other pictures later. The equivalence class held together despite the conflict.

Allen et al. (2001) saw something different. In their study, some people's classes fell apart after conflicting training. The difference? D's team used much harder conflicting tasks. This suggests classes only break under heavy stress.

Davison et al. (2002) showed you can teach people to switch classes on and off using context cues. But their method needed tons of practice. Silguero's simpler approach shows the default is stability, not flexibility.

Mason et al. (2025) found fragile transfer in rats. The animal data lines up with the human finding - once classes form, they resist change.

Your client's stimulus classes won't fix themselves when problems appear. If a safety sign gets linked to scary stimuli, that link stays put. You'll need active intervention like context training or re-teaching. Don't assume conflicting experiences will naturally boot the problem stimulus from the class.