Reinforcement and generalization of productive plural allomorphs in two retarded children.
Reinforcing one plural sound can make kids use the other plural sound with new words.
01Research in Context
What this study did
Two children with intellectual disability took part.
The goal was to teach when to add /s/ or /z/ sound to make a plural.
Each correct plural got praise or a small treat.
Sessions used short discrete trials: hear the word, say the word, get the reinforcer.
What they found
Both kids learned the trained plural sound quickly.
They also used the untrained plural sound with new words.
Reinforcement of one rule spilled over to the other rule.
How this fits with other research
Sanders et al. (1971) ran a near-copy study the same year.
They taught adjective endings (big, bigger, biggest) instead of plurals.
Both papers show that shaping one morphological rule creates broad language gains.
Meyer et al. (1987) later used matrix training to build whole phrases, not just word endings.
That study extends the 1971 idea: teach a few pieces, get many new sentences for free.
Marya et al. (2021) repeated the matrix plan with kids with autism who use talkers.
The chain of studies keeps the core finding alive across decades, populations, and tools.
Why it matters
You can teach grammar rules directly and still get free generalization.
Pick one clear contrast, like /s/ vs /z/, and reinforce it hard.
Then probe new words before you teach them; you may find you do not need extra trials.
This saves table time and gives the learner a bigger language set faster.
Want CEUs on This Topic?
The ABA Clubhouse has 60+ free CEUs — live every Wednesday. Ethics, supervision & clinical topics.
Join Free →Run ten trials on /s/ plurals, then test five untrained /z/ words and score generalization.
02At a glance
03Original abstract
Morphological dictates of English usage call for the unvoiced allomorph /-s/ to form the plural of singular nouns with unvoiced endings (e.g., cups). Conversely, the voiced allomorph /-z/ is required to form the plural of nouns with voiced endings (e.g., tree). The study sought to determine the extent to which differential reinforcement could control the acquisition of plural allomorphs in two retarded subjects. In Condition 1, one subject was trained with reinforcement procedures on a list of words calling for the /-s/ allomorph. She was then given unreinforced probe items to determine the extent of generalization to words calling for the /-z/ allomorph. In Condition 2, the procedures were reversed and this subject was trained on a /-z/ list and probed for generalization of /-z/ to words calling for /-s/. A second subject was exposed to the same conditions in the opposite order. The results for the two subjects lent unequivocal support for the hypothesis of generalized training effects. It was concluded that appropriate usage of the linguistic response class "plurals" is susceptible to generalized training effects of differential reinforcement.
Journal of applied behavior analysis, 1971 · doi:10.1901/jaba.1971.4-305