The unit of selection: what do reinforcers reinforce?

Shearn et al. (1997) wrote a theory paper. They asked: what exactly gets picked when reinforcement happens?

The authors said reinforcers do more than make a response stronger. They change the cues that come before the response, even private thoughts or feelings.

No new data were collected; the piece walks readers through logic and earlier findings.

The paper claims the real unit of selection is the whole antecedent package, not the single response.

That package can include hidden events like self-talk or bodily sensations. If those inner cues shift, the response changes even though the consequence stays the same.

Jimenez-Gomez et al. (2025) later tested kids and adults in a lab choice game. People picked the option that signaled better future pay, even when past pay was equal. Their data line up with the 1997 idea: reinforcers work as signals, not just strength builders.

Peterson et al. (2016) tried the same logic with children with autism. Letting a child pick the reinforcer before the task raised work levels for half the group. This extends the 1997 paper from theory to everyday therapy.

O'Hora et al. (2014) separated two paths: rule cues versus rule consequences. Both shaped new instruction-following, showing antecedent control is real and measurable. Their single-case design gives clinicians a ready model.

If reinforcers mainly change antecedents, you should scan for the cues you might be strengthening. Check for private events the learner feels right before the target behavior. Try offering choice or other clear signals early, not after the response. This small shift can boost the power of your reinforcement plan without adding extra tokens or candy.