Abstract :
This paper investigates the acquisition of extrasyllabic consonants in Polish on the basis of the body of longitudinal data from a Polish speaking child, 3;8–5;1. The data reveal several strategies adopted by the child to deal with the problematic consonants: deletion, vowel insertion and adjunction, each characterising a different stage of development. Shifting from one strategy to another seems to reflect a slow and gradual process of learning the intricate relationship between extrasyllabicity, prosodic affiliation and transparency vis-à-vis voicing processes found in adult Polish. The discernible reorganisation in the childʹs grammar seems to affect both underlying representations and phonological processes. The childʹs underlying representations are argued to be essentially adult-like in the sense that they are based on alternations rather than directly on adult surface forms. The shift in the processing strategies, analysed within Optimality Theory [henceforth, OT; Prince, A.S., Smolensky, P., 1993. Optimality Theory: constraint interaction in generative grammar. Unpublished manuscript, University of Rutgers, Rutgers, University of Colorado, Boulder; McCarthy, J., Prince, A.S., 1995. Faithfulness and reduplicative identity. Unpublished manuscript, University of Massachusetts, Amherst, Rutgers University, Rutgers], is argued to be best motivated by developing units of prosodic organisation. The claim is that at the stages of development under discussion, the prosodic hierarchy is not fully activated in the childʹs output structure. Specifically, the childʹs outputs are organised into syllables but not into prosodic words. This explains not only why, in contradistinction to adult Polish, extrasyllabic segments exhibit no transparency effects at the adjunction stage but also why extrasyllabicity is initially resolved by deletion.
