The effect of phonological neighbors and homophones on spoken word recognition in Mandarin Chinese

Abstract

This dissertation examined the effects of phonological neighborhood and homophone mates on spoken word recognition in Mandarin. Phonological neighbors are words that sound similar, and homophones are words with exactly the same pronunciation. Neighborhood density (number of phonological neighbors) and neighbor frequency (average frequency of the neighbors) are the main measures of a phonological neighborhood. Homophone density (number of homophones) and homophone frequency (average frequency of the homophone mates) are the main measure of a homophone family. A lot of research has been devoted to phonological neighborhood effects in English and other European languages, and found overall inhibitory effects of phonological neighborhoods. Words with many neighbors and high frequency neighbors are responded with more errors and longer response time as compared to words with fewer neighbors and lower-frequency neighbors, respectively. But there has not been much research on tone languages. On the other hand, much less is known about the effects of homophone mates on spoken word recognition, probably because most languages like English do not have a large number of homophones in the lexicon. In this thesis, the language under investigation is Mandarin, a tonal language with a high density of homophone mates. The main goal of the current research is to compare the roles of phonological neighbors and homophone mates in the process of Mandarin spoken word recognition. I conducted two experiments: an auditory lexical decision experiment and an auditory naming experiment. Mixed-effects regression analyses of auditory lexical decision results showed facilitatory effects of phonological neighbor frequency, homophone density, and homophone frequency for real monosyllables and inhibitory effects of neighborhood density for pseudo-syllables. In addition, the models also showed a significant interaction of homophone density and homophone frequency for real monosyllables. The results from the auditory naming experiment showed no significant effects of either phonological neighborhoods or homophone mates. Taken together, the current research showed both similarity and differences between phonological neighbors and homophone mates in the processing of spoken Chinese words. Implications for models of Mandarin lexicon are discussed.