| old_uid | 7731 |
|---|
| title | Recognizing liaison-initial words in native and non-native French: Phonemic effects and processing cost |
|---|
| start_date | 2009/12/01 |
|---|
| schedule | 13h30 |
|---|
| online | no |
|---|
| location_info | bât. BSHM, salle 11 |
|---|
| summary | Cross-linguistic research with native listeners has shown that words are recognized more slowly when their onsets do not align with syllable onsets (e.g., Content, Kearns, & Frauenfelder, 2001; Dumay, Frauenfelder, & Content, 2002; McQueen, 1998; Vroomen & de Gelder, 1997, 1999; Weber, 2001). On the basis of this research, it has been proposed that the word-recognition system of adult listeners uses syllable onsets as triggers for lexical access, also referred to as the Syllable Onset Segmentation Heuristic (SOSH; Dumay et al., 2002).
An interesting test case for SOSH is liaison in French, a process whereby a word-final, otherwise latent consonant is resyllabified as the onset of the following vowel-initial word (e.g., –s in sans elle). Because liaison creates a misalignment of the word and syllable boundaries, vowel-initial words preceded by a liaison consonant (henceforth liaison-initial words) should be more difficult to recognize than corresponding consonant-initial words (e.g., sans zèle).
Yet, recent cross-modal priming studies have shown that liaison does not appear to incur a processing cost for native French listeners (e.g., Gaskell, Spinelli, & Meunier, 2002; Spinelli, Cutler, & McQueen, 2002; Spinelli, McQueen, & Cutler, 2003). These findings were attributed in part to acoustic differences between liaison and word-onset consonants, with the former being on average 15% shorter than the latter (cf. Nguyen, Wauquier-Gravelines, Lancia, & Tuller, 2007). This suggests that acoustic information helps native French listeners compensate for the difficulty that liaison might otherwise pose for the word-recognition system.
Little research exists on the recognition of liaison-initial words by non-native listeners. Dejean de la Bâtie and Bradley (1995) report that second-year English-speaking learners of French have more difficulty than native French speakers in determining whether a consonant-initial word is present in sequences where the critical consonant (e.g., t) is either a liaison (e.g., grand éléphant) or a word onset (e.g., grand théâtre). These results indicate that the presence of a latent consonant in potential liaison contexts creates word-recognition difficulties for native English listeners, who do not have a phonological resyllabification process such as liaison in their native language.
Using a more sensitive experimental method, namely eye tracking within the visual world paradigm, we report data from native and non-native speakers of French that shed new light on the recognition of liaison-initial words. The first set of data comes from native French speakers and English-speaking learners of French in the United States (Tremblay, 2009, under review). The second set of data comes from native French speakers and English-speaking learners of French in France.
The participants heard auditory stimuli containing adjective-noun sequences where the noun was ambiguous between a liaison- and consonant-initial word until the vowel in its second syllable (e.g., fameux érable vs. fameux zéro). In one condition, the visual display contained the target word (e.g., érable or zéro) and three distracter words; in another condition, the visual display contained the target word (e.g., érable or zéro), a competitor word (e.g., respectively, zéro or érable), and two distracter words. The experiments that the two sets of participants completed differed in the pivotal consonants that the stimuli contained (/z/ vs. /z/, /t/, and /n/) and in the visual presentation of the words in the display (orthographically vs. with images).
The results show that the recognition of liaison-initial words varies as a function of the liaison phoneme: for /t/ and /n/, liaison-initial words are recognized more slowly than consonant-initial words by both native and non-native speakers, whether or not a lexical competitor is present in the display; in contrast, for /z/, liaison-initial words are recognized more rapidly than consonant initial words for both native and non-native speakers (second set of data), but only in the absence of a lexical competitor.
The more sensitive experimental method used in this study suggests that liaison does incur a processing cost for both native French speakers and English-speaking learners of French, but only for those consonants that often occur in word-onset position (i.e., /t/ and /n/). These findings provide further support for SOSH, but they also suggest that modeling the recognition of liaison-initial words would require weighing the relative frequency of pivotal consonants in liaison and word-onset contexts. |
|---|
| responsibles | Kandel |
|---|
| |