自1995年起，語言學家開啟了中文言談錯誤修正的研究。早期的學者皆著重於分析說話者與他人對話時所犯的語言錯誤，並沒有特別研究人們在獨白言談情境下所產生的語言問題。此外，在先前的研究中，研究人員主要致力於分析人們修正錯誤時所採用的句法機制，並沒有特別關注到引發人們修正錯誤的原因。然而，前人的分析方式不但無法清楚點出語言修正發生的起因，也造成了語料歸類不當的問題。因此，本研究的主要目的是觀察教室獨白中說話者會有哪些語言使用的錯誤，採用包爾葛瑞斯合作原則中的四項準則來歸類說話者修正的錯誤類型，同時也觀察說話者修正不同類別錯誤的句法機制並討論其語用功能。最後，在本研究的錯誤分析中發現，未修正與修正過的語言句法形式相當雷同，這顯示即使是需要修正的話語仍然遵循著語言生成機制運作而成。

Mandarin Chinese repair study was initiated in 1995. Researchers of earlier studies primarily focused their attention on conversational utterances. Monologue speech productions were not particularly studied. Also, in previous investigations, only the syntactic operations imposed on speaker’s erroneous utterances were analyzed. The incentive of each repair initiation, however, was not properly discussed. Thus, the
repairs incurred by different erroneous productions remained
indistinguishable and different problems were mistakenly categorized as the same type of error. In this investigation, a different approach is adopted to analyze Mandarin Chinese erroneous productions based on the data collected from academic monologue discourse. In the present attempt, the problem and the repair mechanism of each imperfect production
are analyzed. Gricean Maxims are employed to identify the reparandums and with regard to repair strategy, the syntactic operation and the pragmatic function of each repair device are examined. Finally, in the data of this monologue corpus, it is also observed that the repairable and the repaired constituent of the same utterance are structurally interrelated to a certain extent, thus arguing against the misconception that
speech blunders are unsystematic productions.