Géraldine Hilaire, Harriet Jisa and Valérie Régol

Dynamique du langage (CNRS UMR 5596 & Université Lyon2)

Geraldine.hilaire@ish-lyon.cnrs.fr

Much research has been devoted to understanding the role that phonological cues play in the child’s acquisition of closed class items (cf., volume Morgan & Demuth 1996). Language input contains information sufficient to permit syntactically and semantically naïve learners to assign words to two major categories, corresponding closely to content words and function words (Morgan, Shi, Allopenna 1996). Function words, in opposition to content words, tend to be productively and perceptually minimal.

Languages differ in their organisation of segmental and suprasegmental phonology and these differences have a profound effect on early acquisition (Peters & Menn, 1990) Function words have been shown to play an important role in the early acquisition of French (Bassano, Maillochon & Eme 1998, Veneziano & Sinclair 2000). In contrast to English-acquiring children, where nouns clearly dominate in early lexical inventories (Bates, Bretherton & Synder, 1988), French-acquiring children show more grammatical morphemes (e.g., pronouns and determiners) in their early lexicons.

Acquisition of determiners is an important developmental step in spoken French as they control marking of gender and number. Despite the fact that French determiners are prosodified with the following noun and it is this second element that carries phrase final length, normally developing children acquire these minimal phonetic morphemes early on. The goal of this study is to trace the development of these grammatical morphemes in spoken French of implanted children.

Two explanations for omission in early language production are particularly pertinent for this project. The "rythmic production" account argues that English-speaking children’s early words and utterances conform to an alternative Strong-Weak (stressed-unstressed) pattern, where unfooted syllables/morphemes will be omitted (Allen & Hawkins 1978, 1980, Gerken 1991, 1994, 1996). It is possible to transpose this rythmic account to French and predict that children will omit determiners in prosodic units consisting of DET + N because of a production constraint which requires vowel lengthening on the last syllable of a dysllabic utterance. The "perceptual account" predicts that children tend to include only "perceptually salient" stressed and final lengthened syllables in their early word production (Echols 1993, Echols & Newport 1992). While both of these accounts can explain omissions, they suffer from a lack of a developmental perspective. The project we will present details the longitudinal profile of three implanted children, beginning at the time where the vast majority of determiners are omitted and ending at the point in time where the majority of determiners are present.

Three prelingual profoundly deaf females received cochlear implants at the age of 1;11, 2;4 and 3;1 respectively. They were recorded every six months during natural interaction with their mothers. Our analyses begin at 10 months post-implant and continue until 36 months post-implant. Our analysis examines the children’s productions with particular attention being paid to filler syllables in determiner position. All common nouns (accompanied or not by a determiner) were extracted from the corpus. The determiner slot was examined in order to ascertain: 1) the rate of omission errors over the time period investigated; 2) the phonetic content and stability of the child’s filler syllables and; 3) the context in which the form was produced (for example, characteristics of the following noun such as vowel in initial position, monosyllabic or multisyllabic word). Of particular interest in our study is the distribution of segments making up determiners. For instance, the vowel [E ]) is present in a number of word final positions (bain, pain, faim), however its use in the indefinite article un is absent in the beginning of the study and very unstable across development.