有關漢字辨識的研究,目前集中於探討部件扮演的角色,其中Zhou與Marslen-Wilson(1999b)根據進行唸名作業實驗的結果,宣稱部件處理與整字處理相同,都會牽涉語意處理,本研究企圖以聲旁部件為對象來檢驗此說法。在本研究的實驗一、二中,發現在唸名作業下並沒有發現如同Zhou與Marslen-Wilson(1999b)所觀察到的顯著聲旁語意促發效果;反而是在字彙判斷作業下可以觀察到顯著的語意促發效果,本文作者判斷這是由字彙判斷作業的作業要求造成。實驗一、二的結果也同時顯示,在促發情境呈現聲旁單獨成字的狀況,會使聲旁語意促發效果較強。實驗三進一步操弄聲旁單獨成字的字頻(以下簡稱部件字頻)與部件符頻(token frequency),並排除部首部件可能造成的影響,結果只在部件字頻高的情況下有機會對低頻目標字產生聲旁語意促發效果。綜合實驗結果推論,在某些情況下,低頻漢字的辨識過程裡,縱使牽涉部件層次的處理,也通常不會涉及語意處理,只有當部件容易同時被當成整字來處理時,才可能觀察到語意促發效果,一般典型的日常情況下,卻不是這樣的。
Based on finding from primed character naming experiments, Zhou and Marslen-Wilson (1999b) argued that there are simultaneous processes of a whole character and its embeding radicals. The present studies examined their argument. Three experiments were conducted. In Experiment 1, the same design as their experiments was adopted, using not only a naming task but also a lexical decision task. In Experiment 2, a similar design as Experiment 1 was adopted, except that a primed condition of phonetic radical priming was excluded. The results of these two experiments showed that there was no facilitatory semantic priming effect of phonetic radical observed in the naming task, but instead, some facilitatory semantic priming effects was observed in the LDT task. In Experiment 3, token frequency of a radical, and character frequency of a radical on its own right as a character, were further included into orthogonal manipulation. It was found that only when low frequency target chatacter preceded by a primed character with a low token frequency phonetic radical, which also could stand alone as a high frequency character, facilitatory priming effect could be observed in lexical decision. It was elaborated that semantic activation of phonetic radical could only be observed under some circumstances which can induce subject’s coping strategies. The proposal of Zhou and Marslen-Wilson was thus not supported.