中文字形與音韻周邊預視效益：於較佳閱讀能力失聰者之證據

The Parafoveal Preview Benefits of Chinese Orthography and Phonology among Deaf Readers with High Reading Abilities

失聰者之閱讀能力往往較低，本研究欲瞭解教育程度較高之失聰者在受到不同語言訓練（口語、非口語）後，如何影響中文閱讀時周邊視野詞彙的字形與音韻之運作。實驗使用眼動邊界典範與眼動誘發呈現改變技術（Rayner, 1975），研究假設受試者若可在周邊視野擷取到詞彙相關訊息，則當眼睛通過設定邊界使得預視詞立即取代為目標詞時，眼睛凝視目標詞的處理便會受到影響。實驗操弄目標詞與預視詞的字形與音韻相關性，受試者必須閱讀句子並判斷合理性。透過中文字形及音韻可分離的特性，可釐清失聰者詞彙中音韻的激發情形。結果是：字形預視效果在聽人、失聰口語組（第一凝視時間、凝視時間與總凝視時間）與失聰非口語組（總凝視時間）中發現；音韻助益在聽人（第一凝視時間、凝視時間）與口語失聰組（凝視時間）中發現，但在非口語失聰組中未出現；將前凝視點位置區分為遠近時，字形效果可發現於聽人與失聰非口語組的遠、近凝視點之眼動指標，以及失聰口語組的遠凝視點指標；音韻效果發現於聽人近凝視點（第一凝視時間）與遠凝視點（為凝視比例），以及失聰口語組遠凝視點（第一凝視時間），而失聰非口語組則未有音韻效果。由此可見，失聰口語組於遠凝視點發現音韻效果，聽人則於近凝視點發現，反映出失聰口語組與聽人在周邊視野中擷取語言表徵歷程有所差異；失聰非口語組未激發音韻表徵，反映出失聰者語言背景會影響其詞彙表徵激發；兩組失聰族群皆能正確地判斷句子語意，則顯示音韻表徵之激發與閱讀理解之間並非必然的對應關係。

Purpose: Deaf people frequently have poor reading skills. This study consisted of deaf people with high reading abilities with Chinese oral training (DO) and non-oral training (DnO), and investigated orthographic and phonological preview benefits while reading Chinese. The dissociation of orthography and phonology in Chinese characters provides a useful tool to examine phonological processing in deaf people. This study used a boundary and display change technique (Rayner, 1975) in which a preview word was replaced by a target word when the reader’s eyes crossed an invisible boundary. We hypothesized that, if readers obtain partial lexical information from a preview word in the parafoveal area, they acquire preview benefits when fixing the target word. Methods: This study recruited 30 hearing-enabled students, 28 DO participants, and 19 DnO participants. This experiment manipulated the orthography and phonology similarities between the target and preview words. The experiment contained 60 reasonable sentences (preview words), 20 identical reasonable sentences (unchanged words), and 40 unreasonable sentences (unchanged words). Each sentence was composed of 14 to 20 Chinese characters, and the target words were located between the third and eighth characters. The participants had to decide whether a sentence was reasonable (yes response) or unreasonable (no response) based on their comprehension of the entire sentence. Findings: By considering the previous fixation location, preview benefits for orthography were located in first fixation duration (FFD), gaze duration (GD), and total viewing duration (TVD) for both the hearingenabled and DO participants, and in the TVD for the DnO group. The phonological effects were shown in FFD and GD for the hearing-enabled participants and in the GD for the DO group, but not for the DnO group. Considering the possibility of an enhanced perceptual span for deaf readers, the pre-fixation locations were divided into near and far fixations. The results showed that orthography effects were found in the near pre-fixation and far pre-fixation for the hearing-enabled and DnO groups, and in the far pre-fixation location (FFD, GD) for the DO group. The phonological preview benefit was evident in the far pre-fixation location (FFD) for the DO group, whereas it was located in near prefixation location (FFD) for the hearing-enabled group. No phonological effect was found for the DnO group. Conclusions/Implications: The results demonstrated that linguistic presentation processing in the parafoveal area may be different between hearing-enabled and deaf people. Although phonological preview benefits were only shown the OD group, both deaf groups correctly comprehend and assessed the sentences. The findings suggest that the differing language backgrounds of deaf people entail varying representation activations, and that the activation of spoken language-based phonological codes is sufficient, but not required, for reading comprehension.