Young L2 learners' performance on a novel morpheme task.

The teaching of an invented language rule has been proposed as a possible non-biased, language-independent assessment technique useful in differentiating young L2 learners with specific language impairment from their typically developing peers. The current study explores these notions by testing typically developing sequential bilingual children's ability to learn an invented language rule in either L1 (Spanish) or L2 (English). Participants were 20 children, age 3:6-5:8, who attended a Head Start programme. For all children, Spanish was the primary language spoken in the home and English was the primary language of instruction. Children were randomly assigned two groups. Group L1 was taught the novel language rule in Spanish; Group L2 was taught the novel language rule in English. Performance was better for the L1 group than for the L2 group. Moreover, not all of these typical language learners were able to learn the new rule, even in their strongest language. These findings suggest that even for typically developing children, specific language proficiency as well as individual differences are closely linked to performance on this novel morpheme learning task.