First and second language sublexical processing: a comparison of native and non-native readers of English

Abstract

This thesis examines reading performance in Filipino, a language that has received scant psycholinguistic attention, in an attempt to identify which linguistic properties influence sensitivity to particular sublexical units. In so doing, questions are addressed about the transfer of sublexical reading strategies from first language to second language. The presented stimuli were always physically broken up at different boundaries in order to determine sublexical preferences. Few differences were found between the monosyllabic English word processing of native Filipino and native English speakers. Instead, results from lexical decision experiments (LDT) pointed to an effect of onset sonority for both groups regardless of whether the onsets occurred in word or syllable initial position (Experiments 1, 2, 5, 6). This was interpreted as a phonological effect, particularly since it disappeared under concurrent articulation (Experiment 4). When a naming task was used (Experiment 3), results consistent with those of the lexical decision task were found only for nonword items, suggesting that sonority effects emerge at a sublexical level. Subsequent experiments focused on disyllabic word processing and found consistent language group differences. Lexical decision (7, 8, 9, 12, 15) and naming (10, 13) experiments found that a "maximal onset principle" of syllabification was more prominent in the Filipino and English reading strategies of Filipino speakers compared to a "maximal coda principle", though this preference was modulated by syllable stress (Experiments 7-10, 12, 13). Such effects disappeared when concurrent articulation was present (Experiments 11, 14), suggesting a phonological basis. Native English speakers consistently favoured a maximal coda analysis in their processing of English words (Experiments 8, 12, 15). No effects of syllable stress were found in LDT. They did emerge, however, in the naming task (Experiment 13), suggesting that native English speakers exploit phonological syllable units when overt pronunciation is required. Experiments 15, 16, and 17 investigated the sensitivity of Filipino speakers to variations in vowel quality, a feature that is found in English but not Filipino. It was concluded that the extent to which Filipino speakers are sensitive to a phonological property is largely determined by whether it features in their native language.