采用詞匯判斷任務和ERP技術,操縱義符和聲符的家族大小,考察形聲字識別中義符和聲符的家族效應。結果發現:(1)當義符為大家族時,大家族聲符字比小家族聲符字誘發了更小的P200波幅;當義符為小家族時,大家族聲符字和小家族聲符字誘發的P200波幅差異不顯著。(2)大家族聲符字比小家族聲符字誘發了更大的N400波幅,但當義符為大家族時,大、小家族聲符字之間的差異比義符為小家族時更加顯著。這說明,義符和聲符的家族大小共同影響形聲字的識別,聲符的家族效應受義符的家族大小調節。研究表明,在詞匯識別的不同階段,義符和聲符對資源的獲取處于變化狀態,這取決于兩者的呈現率差異。整個研究表明,在形聲字的識別中,義符與聲符的作用存在著博弈,二者作用的此消彼長與加工階段和家族大小有關。

Phonograms are comprised of a semantic radical and a phonetic radical. The semantic radical usually implies the meaning of a character, while the phonetic radical offers a phonetic clue for the pronunciation of a character. Prior research has indicated that semantic and phonetic radicals each play an important role in phonogram cognition. However, the way in which the 2 kinds of radicals combine to influence both phonogram recognition and the activation process remains a controversial topic. Neighborhood size is one of the most important characteristics of radicals. The differences induced by the neighborhood size of radicals are called neighborhood effects. Prior studies employed various methods to investigate the roles of semantic and phonetic radicals on phonogram recognition and discovered that phonetic radicals have an advantage in phonogram processing. They also found that the neighborhood size of semantic radicals regulated the relative contributions of semantic and phonetic radicals. However, most of these researchers investigated the roles of semantic and phonetic radicals respectively. Only a few investigators have focused on the comprehensive role of semantic and phonetic radicals in character processing. This study attempted to investigate the comprehensive effect of semantic and phonetic radicals on phonogram recognition.Participants’ brain responses were recorded using event-related brain potentials. Four types of characters were selected: HH(phonogram comprised of a high-frequency semantic radical and a high-frequency phonetic radical), HL(phonogram comprised of a high-frequency semantic radical and a low-frequency phonetic radical),LL(phonogram comprised of a low-frequency semantic radical and a low-frequency phonetic radical), and LH(phonogram comprised of a low-frequency semantic radical and a high-frequency phonetic radical). A lexical decision task was adopted; the frequency of the entire character and the number of strokes were equivalent for all groups.The results showed that, for characters with high-frequency semantic radicals(HH and HL), the characters with high-frequency phonetic radicals(HH) elicited a larger P200 waveform component than the characters with low-frequency phonetic radicals(HL). However, for characters with low-frequency semantic radicals(LH and LL), there was no significant difference between the characters with high-frequency phonetic radicals(LH) and those with low-frequency phonetic radicals(LL). Characters with high-frequency phonetic radicals(HH and LH)elicited a larger N400 component than those with low-frequency phonetic radicals(HL and LL), and the N400 differences induced by the neighborhood size of phonetic radicals with high-frequency semantic radicals(HH and HL) were larger than those induced by characters with low-frequency semantic radicals(LH and LL). These results indicate that in the early stage of phonogram recognition, the neighborhood effect of phonetic radicals is regulated by the neighborhood size of semantic radicals. In the late stage of phonogram recognition, the higher the frequency of phonetic radicals, the stronger the semantic activation degree of the character. In general, the neighborhood size of phonetic radicals was found to affect vocabulary accessibility. However, the effect of the neighborhood size of phonetic radicals is regulated by the neighborhood size of semantic radicals.