Development and function of consequence classes in operant behavior.

Embregts (2000) wrote a theory paper. He asked: what if we stop looking at single reinforcers or punishers?

Instead, we group consequences that work the same way into a "consequence class." The paper shows how this new unit can tidy up old data and guide future tests.

The author found that many different consequences shrink or grow behavior in the same direction.

When consequences share a function, they can swap places and still steer choice. Grouping them into classes makes patterns easier to see and teach.

Flanagan et al. (1958) first said "consequences control stuttering." Embregts (2000) keeps the idea but bundles those consequences into classes.

Deluty (1976) and DARDANO et al. (1964) showed punishment on one choice pushes responding to the other. Their data are early bricks in the consequence-class wall.

Bland et al. (2018) proved an S- can punish. That single case becomes cleaner to explain when you call the S- part of a "punisher class" instead of a one-off event.

Fontes et al. (2025) found relative punishment, not absolute rate, drives choice. This supports J’s claim: what matters is class membership, not the specific punisher.

You can stop memorizing endless reinforcer or punisher lists. Ask: what class does this consequence belong to? If two rewards both increase requesting, they are in the same class—use either one when the first is unavailable. If a red card and a brief break both cut disruptive behavior, treat them as interchangeable class members and rotate to reduce satiation. Thinking in classes saves planning time and gives you flexible, ethical options during sessions.