CONSTITUENT SYLLABLE EFFECTS IN A NONSENSE-WORD REPETITION TASK

Multisyllabic nonsense-word repetition tasks have been used to provide evidence on the phonological processing operations of children with language impairment, independent of their lexical knowledge (Gathercole & Baddeley, 1990; Kamhi, Catts, Mauer, Apel, & Gentry, 1988). However, recent evidence (Gathercole, Willis, Emslie, & Baddeley, 1991) and speculation (Snowling, Chiat, & Hulme, 1991) suggest that the nonsense words employed in such tasks may not be equally ''nonsensical.'' The present investigation directly tested the effect on repetition performance of one previously uncontrolled characteristic of multisyllabic nonsense words: the lexical status (word or nonword) of their stressed syllables. Normally achieving school-age boys repeated nonsense words with lexical stressed syllables significantly more accurately than nonsense words with nonlexical stressed syllables. These results suggest the need to control, at a minimum, the lexical status of constituent syllables in constructing nonsense-word stimuli.