中文的「詞優」與「字優」效果

WORD AND CHARACTER SUPERIORITY EFFECTS IN CHINESE

本研究分別以四個實驗在中文的「詞」與「字」的層次，採用Prinzmetal（1991）的實驗模式，在沒有「形態遮蔽」（pattern masking）與加上形態遮蔽各情況下，探討中文的「詞優效果」與「字優效果」並且研究「詞－字」與「字－部件」現象。實驗一與實驗二得到顯著的「詞－非詞」效果，實驗三與實驗四也得到顯著的「字－非字」效果。但是實驗一、二的「詞－字」現象與實驗三、四的「字－部件」現象與西方文獻之「詞－字母」效果不相同。本研究推測在雙字詞內辨認單字是把單字當成「單字詞」進行辨識，因而偵測單獨呈現的單字與偵測「真詞」裡的單字之正確辨認率相差不大。而在中文字內辨認部件則會同時受到「半字字優效果」與「側抑制」的影響，造成偵測單獨呈現的部件之正確率低於偵測「真字」裡的部件。這些現象同時反映辨認中文刺激中「位置」、「頻率」與「筆劃數」各變項影響的特性。

Using Prinzmetal's (1991) experimental paradigm, this study investigated the word and character superiority effect. A word in this study refers to a two-character word. Experiments 1 and 2 obtained the 'word-nonword' effect, and Experiments 3 and 4 found the 'character-noncharacter' effect. With the simultaneous pattern-mask procedure, Experiment 2 did not obtain the 'word-character' phenomenon, perhaps because a character is itself a single-character word. Experiments 3 and 4 obtained the 'partial-character' superiority effect, indicating that Chinese readers are able to perceive orthographic regularity from character parts. The effects of the two variables (constituent position and stroke member of a character) that give rise to specific features of the word and character superiority effects in Chinese were discussed.