The social control of generalized imitation.

Steinman (1970) watched kids copy adult actions. Some copied moves earned candy. Some earned nothing.

The team then told kids, "Only copy the moves that win treats." They wanted to see if words alone could stop the useless copies.

Kids kept copying both rewarded and non-rewarded moves until the rule was spoken. Once told, the useless copies stopped almost at once.

The study says children follow social cues, not just candy. Words acted like a switch.

Carr et al. (2003) and Sun et al. (2025) later used the same social-cue power to help non-vocal children speak. They ran quick motor imitation bursts first, then gave a vocal model. Kids began to talk.

Gladstone et al. (1975) flipped the direction: adults copied the kids. Child behavior still rose, showing the cue itself is the reinforcer, not the candy.

Dal Ben et al. (2019) seemed to disagree. Kids kept using a grammar form even when adults praised a different form. The twist: praise is another social cue, but hearing their own voice match the model felt better, so the form stayed. Both studies agree that social feedback, not toys or candy, drives the behavior.

You can stop unwanted imitation with a short rule instead of pulling reinforcement. Before you place a demand, run two or three easy motor imitations the learner always does. Then give the new vocal target. This simple sequence raises compliance and can spark first words in non-vocal clients.