Teaching mands by manipulating conditioned establishing operations.

The team asked a simple question: can you teach someone to ask for something if they can already name it? They started with tact training — teaching labels like “cookie.” Then they tried to turn those labels into mands — requests like “cookie, please.”

When tact training alone failed, they switched tactics. They used imitative and tact prompts, then faded the prompts while tying the word to a real need — hunger for the cookie. This shift is called transferring stimulus control to a conditioned establishing operation.

Tact training by itself never produced requests. Only after they added direct mand training — linking the word to the moment the person wanted the item — did asking appear. Prompt fading from imitation and tact cues to the natural need was the key step.

Belisle et al. (2019) extends the idea. They used discrete-trial training to teach kids with autism and Down syndrome to label abstract feels like “wet” and “hard.” The kids could then use those tacts in new places, showing tact training can generalize when the goal is naming, not asking.

Mace et al. (1990) used the same prompt-fading tool, but for teaching discriminations like “same vs. different.” Their task-demo model cut errors in half, proving fading works across very different skills.

Bruns et al. (2004) looks like a clash at first — they found listener training alone was not enough for toddlers to sort new pictures. But the gap is about purpose: listener and tact training help kids understand categories, while G et al. shows you still need direct mand training to create requests.

If your learner can name “juice” but never asks, stop adding more tact trials. Instead, set up a moment of need — a small sip left in the cup — and prompt the mand. Fade your prompt as the thirst grows. This study tells you to train the request directly, not hope it leaks out of labeling practice.