本文探討語境可用性以及語音相似度兩個因素對台閩語子音音段 語誤的影響。結果顯示大部分的語誤都是受到語境當中的的語音所影 響，而且語誤的數量與目標音跟來源音之間的距離成反比關係；語境 可用性範圍侷限在目標音七個音節內的距離。此現象對於探討大腦處 理語言訊息時所運用的語言單位之相關研究具有顯著意義。語音相似 度分析顯示語誤數量與目標音跟來源音兩者的語音相似度成正比關 係，兩個語音的語音相似度越高，越容易發生語誤。總言之，本文的 分析結果與文獻上其他語言之語誤研究結果一致，顯示語誤有跨語言 的普遍性特徵；同時，本文也進一步發現語音相似度的影響會隨研究 所採取的不同語音徵性系統而有程度上的差別。

This study explores the effects of contextual availability and phonetic similarity on speech errors of Taiwan Southern Min. The results show that most consonant errors are attributed to the influence of surrounding segments. The number of error tokens and the distance of contextual availability reveal an inverse relationship. The limited distance of 7 syllables as a domain of the contextual-availability effect is of significance concerning the linguistic unit in language processing. On the other hand, the results show that the number of error tokens is directly proportional to the degree of phonetic similarity shared by the target and the intruder. In summary, the results support the findings shown in previous studies of other languages, implying some universal properties of speech errors. Furthermore, this study shows that the effect of phonetic similarity may vary with the featural system used as an index of phonetic similarity.