選擇性詞彙缺失一直是失語症研究中受到矚目的一個議題。文獻報告指出不同類型的失語症患者對於名詞和動詞會有不同的表現:在一般口語以及圖片命名作業中,非流利型和失文法型的病人相對於名詞而言傾向於省略動詞,而流利型的病人,特別是Wernicke’s病人以及忘名症的病人(Anomic aphasia),則會有相反的表現。 由於之前的研究重點多將焦點擺在印歐語系的語言上,我們目前對於漢語失語症患者在這方面的表現還是有限。有鑑於此,本研究欲藉由探討漢語失語症患者的雙字複合詞處理歷程,而對於名詞/動詞選擇性缺失的這個議題有更進一步的認識。 本研究包含了兩個子實驗—圖片命名作業以及詞類判斷作業。我們操弄了複合詞的詞性以及複合詞內部的構詞結構。 這樣子的操弄使我們可以同時看到病人在整字層次以及次詞彙層次的表現。 本研究欲回答的問題有 (1) 名詞/動詞選擇性缺失是否只出現在口語表達的情況下? (2) 名詞/動詞選擇性缺失是否也同樣存在於次詞彙的層次? (3) 參考Levelt’s 的詞彙處理模式, 這樣的現象應該是屬於哪一層次的詞彙缺失? 我們的資料顯示名詞/動詞選擇性缺失,即非流利型病人的動詞缺失以及流利型病人的名詞缺失,同時存在於口語表達以及語言理解的情況。此外,我們在次詞彙層次中並沒有發現名詞/動詞選擇性缺失。相反的,我們發現病人還保有對於整體構詞結構的概念。 基於這些結果,我們建議名詞/動詞選擇性缺失應該是屬於語意(而非詞類)上的缺陷。此外,我們的資料顯示: 複合詞的構詞結構越典型(如),病人的表現就越好,表示詞素的排列組合必然是心理詞彙的規範之一。這個現象也許跟構詞結構的內隱學習有關。未來後續研究可從雙字複合詞的語意透明度或鄰項數目大小切入。
Category specific naming deficits have been closely attended to in studies of aphasia. It has been reported that main verbs and object names can be selectively dropped in different types of aphasics. Nonfluent and agrammatic aphasic patients tend to produce far fewer verbs than nouns in connected speech, while fluent patients, particularly Wernicke’s aphasics and Anomics, demonstrate the opposite pattern. Similar phenomena have also been demonstrated in naming-to-confrontation tasks. Due to skewed research emphasis on certain languages, particularly English, our knowledge concerning this issue is still too limited. In view of this, this study takes advantage of the specialty of Mandarin Chinese by exploring the processing of disyllabic compound words by Mandarin aphasia. With this investigation, we hope we can open another window on the NOUN/VERB issue. The present study comprises two experiments. One uses a picture confrontation naming task and the other a linguistic category judgment task. By manipulating the morphological structure of the target compound words, we analyzed participants’ word processing at the whole word as well as the sub-lexical level. Three questions were addressed in this study: (1)Does the NOUN/VERB double dissociation occur in production tests only? (2) Does a similar phenomenon also exist at the ‘sub-lexical’ level? (3) Based on Levelt’s lexical production model, where may the possible functional locus of the deficit be? Our data indicate that NOUN/VERB lexical impairments are not exclusive to production tests, but occur in both perception and production tasks. This became apparent when we increased task requirements, causing the deficit in question to appear in the comprehension tests, too. In addition, we did not find the breakdown of NOUN/VERB lexical retrieval penetrating into the sub-lexical level. On the contrary, the global morphological structure seems to be relatively preserved in those the aphasic participants in our study. Based on the above two findings and the structural features of our target words in relation to Levelt’s (1999) model of lexical retrieval, we suggest that the selective lexical deficit may be located at the semantic/conceptual level. In addition to the above findings, our research reveals another interesting aspect of lexical retrieval. Our data shows that the more typical the compounds are, the better the participants performed, indicating that morphology—the patterns of morpheme composition—must be listed in the mental lexicon. This phenomenon may be related to implicit learning of morphological structure, and future studies could focus on the neighborhood size effect of different compound words.