The effects of listener and speaker training on emergent relations in children with autism.

Early et al. (2012) worked with four children with autism aged 5–7.

They gave each child two kinds of lessons. First, listener training: when the teacher said a word, the child touched the matching picture or printed word. Second, speaker training: the child looked at a picture or word and said its name aloud.

No one taught the kids to match every possible pair. The team watched to see if the two kinds of training would create new, untaught connections on their own.

Three of the four children formed new, correct matches they had never been taught. For example, after learning to touch the written word “cat” when hearing “cat,” and after learning to say “cat” when seeing a cat picture, they could now touch the written word when they saw the picture—without direct practice.

The fourth child did not show these emergent relations.

Cicchetti et al. (2014) conceptually replicated this result while dropping the speaker lessons. They gave only listener training and still saw new tacting and sorting emerge. This suggests the speaker part may help, but it is not always required.

WMruzek et al. (2019) extended the idea to younger preschoolers. They added echoic and identity-match prompts and again saw untrained tacts appear, showing the approach works across ages.

Maddox et al. (2015) added a caution: only children who scored at ABLA-R Level 6 formed equivalence classes with letters. Kids at lower levels did not, even with the same training. This finding extends Early et al. (2012) by showing that learner readiness, not just the procedure, decides success.

If you run table-top or natural-environment trials, pair listener and speaker tasks in the same lesson block. The combo can spark new relations without extra teaching time. Probe for emergent skills after every few lessons, and check ABLA-R Level 6 before you start if you use printed words. This saves hours of direct instruction and keeps kids moving.