本研究旨在探討部件、部件位置處理對於中文字詞辨識的影響。經由文獻回顧，這個研究認為中文組字規則包含字的結構、部件本身的訊息、以及部件的位置，因此在一系列的行為實驗操弄部件位置，觀察實驗操弄在各種實驗情境對觸發效果的影響；除行為研究外，也進行電生理實驗，以形狀大小作業探討中文字詞辨識的電生理指標與拼音文字的異同。 實驗一操弄作業特性以及刺激起始時間差(SOA)。當觸發字為目標字的某一部件時，結果無論目標字字頻為高頻或低頻的情況下，各種實驗情境均無得到顯著的觸發效應。 實驗2-1的觸發字與目標字為相同結構，結果觸發字與目標字相同部件位於相同位置時得到顯著的促進觸發效果。而實驗2-2證實了這樣的觸發效果不會受到語音的影響。實驗2-3的觸發字改為接上低頻部件的非字，得到抑制觸發效果，故推論部件位置和部件之間接合機率均有可能會影響到觸發效應。 實驗3-1、3-2、3-3將增長SOA為60毫秒，並以唸名作業進行，觸發效應則反轉為抑制的觸發效果。顯示SOA或作業特性會產生不同的結果。 實驗四則探討觸發字與目標字為完全相同的部件時，操弄部件位置對於觸發效應的影響，結果發現促進的觸發效應。由於實驗四中是兩個部件相同，而前三個實驗僅一個部件相同，推測是因為部件在辨識初期相互限制所產生的增益效應。 實驗五則利用事件相關電位檢驗適用於代表組字規則處理歷程的成分。This research explored the role of radical and radical position in Chinses character recognition. The constituents of Chinese character orthography shall cover at least the character structure, the radicals, and the radical positions. Both behavioral and electrophisological approaches were adopted to tackle the orthographic processing issue. In experiment 1, different tasks and SOA were used to find the priming effects during the manipulations of radical positions. The prime is one of the radicals belong to target character, but there is no significant priming effect. When the prime’s structure is the same as the target’s structure, in experiment 2, the prime and the target shared one radical. There were facilitated priming effects when the same radical at the same position. If the prime is the radical combined with one low frequency radical, and then the subjects were interrupted to process the target character. So the combinability of radicals induced the inhibited priming effects. In experiment 3, through naming task and longer SOA, the facilitated priming effects turned into inhibited priming effects. When subjects process the characters longer, more information than radical position would be processed. In experiment 4, there would be repetition priming effect when the prime and the target were the same characters. If the prime’s radical position reversed, there were still facilitated priming effects of low frequency target characters. The current study shows the effect of radical position processing of Chinese character recognition in different experimental designs. According to the results, there would be more confident hypothesis of Chinsese orthographic rules.
