Acquisition of *ABA paradigms in a child Artificial Language Learning experiment

This study is concerned with the learnability of a typologically unattested morphological pattern among related positive, comparative, and superlative adjective forms. The generalization is that the positive and superlative may not share a common stem (an “A” stem) if the comparative has a distinct stem (a "B” stem) (Bobaljik 2012 et seq.). The term “suppletion” is used to describe the replacement of the positive form’s stem with a different one. The pattern is illustrated in (1) in the attached .pdf file. Given a triplet of positive, comparative, and superlative forms, we find cases of no suppletion (“AAA” pattern), cases where the comparative and the superlative share one stem different from the positive’s stem (“ABB”), and cases where all three are different (“ABC”); however, there’s no attested case where the positive and the superlative share one stem to the exclusion of the comparative (“ABA”). The aim of this study is to determine the learnability of ABA patterns, and thus the exact nature of the *ABA paradigm gap. To test this, we design a child Artificial Language Learning experiment, where 6 to 7 years old children are trained to learn various suppletion paradigms (ABB, AAB, ABA).