Self-repair is commonly produced by speakers of any languages; both native speakers and non-native speakers. This study focuses on self-repair by non-native speakers of English (English L2 speakers). In speech, L2 speakers might produce different types of self-repair, which might also need editing terms to fill the pause before a new utterance is produced. This study aims to explore the types of self-repair
that occurred in the speech production of English L2 learners based on the self-repair classification systems proposed by Levelt (1983), van Hest (1996), and Kormos (2000), as well as the relation between the use of the editing terms and the types of self-repair as proposed by Levelt (1983) and van Hest (1996). The research was conducted in Al-Izhar Islamic High School with 23 research subjects who were randomly chosen. The instrument used in this study is a comic strip to allow the subjects to produce the L2 naturally. The result of this study shows that there are six types of self-repair that occured in the data; namely: covert, difference, appropriateness, error, rephrasing, and rest-category repairs; and there is a relation between the types of self-repair and the editing terms, in which ‘um’ and ‘er’, functioning as conversation fillers, were found only in the covert, different, appropriateness, and error repairs, while several L1 typical terms indicating troubles in conversation were found in the different, error, and rephrasing repairs. |