| Popis: |
The theme linking the various empirical and theoretical components of this thesis is the impact of semantic knowledge on representations and processes that are typically considered to precede access to the semantic system. The focus is on orthographic representations that are activated by presentation of a written word. The opening three chapters of this dissertation are theoretical. Chapter 1 describes the dominant view that repetition priming of visual words – facilitation in responding to a word that has been encountered recently – is mediated predominantly by orthographic form. The second chapter outlines why brain-damaged patients with a selective impairment to semantic knowledge can be used to address the question of whether repetition priming is influenced by conceptual factors. Other variables can modulate priming effects, and some of these are outlined in Chapter 3. Theoretical predictions arising from the research surveyed in the first three chapters were tested in the empirical Chapters 4, 5 and 6. Chapters 4 and 5 present results from semantic dementia (SD) patients in repetition priming experiments that employed perceptual identification and lexical decision tasks. Findings from three repetition priming experiments with normal controls are presented in Chapter 6. A striking finding from the priming studies was that SD patients demonstrated consistent baseline slowing effects, especially for ‘degraded’ words. Chapter 7 is a theoretical chapter that expands on this observation. Relative to controls, SD patients were slower both to report words in perceptual identification, and to make lexical decisions. A major reason for this was the slowness of their reading; indeed several SD patients showed signs of letter-by-letter (LBL) reading. Chapters 8 and 9 were aimed at resolving the question of whether LBL reading in SD can be attributed to the visual deficit that has been used to explain the phenomenon in pure alexia. Word length effects were calculated for word reading and for a non-visual identification from oral spelling task in SD and pure alexic patients. Performance on a set of visual tasks, incorporating linguistic and non-linguistic stimuli, was also assessed. A final experiment assessed the impact of lexicality on the length effect in the two patient groups. Reading RTs were compared for words and nonwords of different lengths, and indicated whether top-down lexico-semantic support moderated LBL reading. It is argued that semantic factors can influence both the initial activation of a written word and the subsequent pattern of repetition priming. |
