Electronic Thesis and Dissertation Repository

Degree

Doctor of Philosophy

Program

Psychology

Supervisor

Dr. Debra Jared

Abstract

Three ERP experiments examined the role of syllables during English visual word recognition. A colour congruency paradigm (Carreiras, Vergara, & Barber, 2005) was used in which disyllabic words were presented in two colours that divided each item either at the syllable boundary (congruent condition), or one letter away from the syllable boundary (incongruent condition). Experiment 1 investigated syllable congruency effects for words that either were presented with an orthotactically illegal segment in the incongruent condition (e.g., whi-mper, comr-ade), or were presented with orthotactically legal segments in the incongruent condition (e.g., whi-sper, cont-act). A syllable congruency effect was observed in the ERP data, but only for words presented with an orthotactically illegal segment in the incongruent condition. Experiment 2 contrasted the phonological syllable with the Basic Orthographic Syllabic Structure (Taft, 1979), and the Maximal Onset Principle. Behavioural and ERP results did not offer any evidence in support of the BOSS, and provided mixed evidence for the MOP. Although phonological syllable effects were found in both behavioural and ERP data, the advantage for a syllable division appeared to occur primarily when the initial segment in alternative divisions was pronounced differently in isolation than in the context of the word (e.g., pi-cnic but not pla-ster). Experiment 3 investigated syllable congruency effects for phonologically confounded and phonologically unconfounded words. For phonologically confounded words, pronunciation of the initial segment in isolation matched that of the whole word in the congruent condition, but did not match in the incongruent condition (e.g., po-ny vs pon-y; pon-der vs po-nder). For phonologically unconfounded words, the pronunciation of the initial segment in isolation matched that of the whole word in both congruent and incongruent conditions (e.g., cab-in vs ca-bin), or mismatched in both iv congruent and incongruent conditions (e.g., ca-ble vs cab-le). A syllable congruency effect was found in the ERP data, but only for phonologically confounded words. These data suggest that readers of English do not parse words into syllables during silent reading. Implications for theories and computational models of English word recognition are discussed.


Share

COinS