Chetail, F. 1 , Colin, C. 2 & Content, A. 1
1 LCLD, Université Libre de Bruxelles, Belgium
2 UNESCOG, Université Libre de Bruxelles, Belgium
Carreiras, Alvarez, and de Vega (1993) reported that written words with a high-frequency first syllable (HFS) were recognized more slowly than words with a low-frequency first syllable (LFS). This syllable frequency effect is generally attributed to lexical competition between syllabic neighbours, and constitutes one major source of evidence that syllables are functional units in visual word recognition. Recently, further evidence has been provided through ERPs both in Spanish (Barber, Vergara, & Carreiras, 2004) and German (Hutzler, Bergmann, Conrad, Kronbichler, Stenneken, & Jacobs, 2004). The first aim of the present research was to test whether the syllable frequency effect on ERPs could be replicated in French, using a design similar to that used in the previous studies. Eighty pairs of words and 80 pairs of pseudowords were selected, each containing one item with a HFS and one with a LFS. Participants performed a lexical decision task while ERPs were recorded using a 32-channel cap. In addition, to try to disentangle the respective influence of syllabic neighbourhood and syllable frequency, we contrasted a subset of pairs in which the number of higher frequency syllabic neighbours covaried with syllable frequency with another subset for which the number of higher frequency neighbours was held constant. Overall, LFS words and nonwords elicited a more positive P200 component than HFS items, and HFS items elicited a more negative N400, which was slightly delayed for pseudowords. In addition, the effects of syllable frequency were modulated by the number of higher frequency syllabic neighbours in both early and late temporal windows. The findings replicate and extent previous ERP reports and lend further support to the hypothesis of complementary facilitatory and inhibitory processes respectively related to syllable frequency and number of competitors.