Modification of the frequency of descriptive adjectives in the speech of Head Start children through modeling without reinforcement.

Researchers watched Head Start kids describe toys. They counted how many describing words each child used.

Then an adult simply modeled richer language. While playing, the adult said things like 'red round ball' or 'tiny soft bear'. No praise, no tokens, no prompts—just the model.

The team tracked whether the children copied the adjectives in their own talk.

Adjective use doubled right away. Kids kept using more describing words even when the adult stopped modeling.

A control group that got no modeling showed no change.

Bennett et al. (1973) seems to disagree. They tried modeling alone to teach teens with ID to ask questions and saw no gain. Only when they added prompts and praise did the skill take off. The key difference: age, diagnosis, and target skill. Preschoolers with typical language may pick up new words just by hearing them, while older learners with ID need extra help.

Loughrey et al. (2014) and Ingersoll et al. (2006) extend the idea. Both used adult models without direct reinforcement and still saw new expressive language in children with autism. Their targets were category names and object imitation, not adjectives, but the pattern is the same: watch, then say.

Abadir et al. (2021) push the method into video. Four kids with ASD learned safety refusals after watching short clips. Modeling still worked; the screen just replaced the live adult.

If you serve preschoolers with solid basic language, try sprinkling in vivid models during play. Say 'striped shirt', 'loud drum', or 'smooth rock' and keep the flow natural. No need to stop and praise—just keep the conversation going. Track the child’s next five descriptions; you may see the same word pop out. It’s a zero-prep, low-intrusion way to grow vocabulary before kindergarten.