French-English bilingual acquisition of phonology: One production system or two?

Abstract

We examine the onset, atrophy and possible interaction of a set of patterns in the speech of our child(Tom) acquiring French and English, from first words(1;8.0—1;10.0) until 2;6.0. The patterns are: (a) consonant harmony (CH); (b) reduplication(REDUP), (c) an iambic stress contour(IAMBIC) for all words, whether French or English, up until the age of 2;2.21,(d) the over generalization of word-final[t](WFT), operational until around 2;5.14, (e) avoidance of word-initial voiceless fricatives(AVOIDVF), operational until the age of 2;6.00, (f) the emergence of a preponderance of English monosyllabic words ending in a consonant(MONOC) from 2;5.00 onwards. Our aim is twofold. We ask how our data bear on the question of whether the bilingual child has two distinct production phonologies from the earliest stage(1;8.0—1;10.0). We also use our data to test recent claims, (a) that placelessness plays a role in CH patterns(Harrison,1999) and(b) that CH is driven by foot structure, such that the differing foot structures of English and French give rise to different patterns of CH in children acquiring English and French(Rose,2000). With respect to the first aim, one of our main findings is that two segmental phenomena(CH and reduplication) in Tom's speech showed clear differentiation between English and French, with CH occurring(we claim) exclusively in Tom's English lexis, and reduplication exclusively in Tom's French lexis. We cannot, however, show that this differentiation was present from the earliest stage, since Tom's word productions at that stage are(we claim) exclusively French¹. Unlike CH and reduplication, other segmental patterns showed “leakage” of a pattern from one language to another. We consider a nonsegmental pattern(Tom's word stress pattern), showing that this exhibited differentiation only at a later stage. Tom's segmental patterns and processes showed complex patterns of interaction, while there was no interaction between his segmental patterns and his word stress constraints. With respect to our second aim, we argue that appeals to “placelessness”in describing CH lack explanatory value. We also show that our data do not support the claims(a) that CH is driven by foot structure and (b)that the distinct foot structures of English and French lead to distinct patterns of CH in children acquiring those languages, whether monolingually or bilingually.