The problem of fossilized errors has been a problematic issue with EFL researchers because that they indicate that traditional methods of instruction are not effective. This issue was, therefore, examined with university-level first-year Japanese EFL students to better understand the context in which they are occurring and their frequency over the course of an academic year. Data was collected from two corpora, the Monologic and Dialogic Corpus (MDC) 2019 has 20,368 words, and 42 subjects, and the second corpus MDC2020, which has 16,997 words and 29 participants. Errors in the 2019/2020 corpora were identified and then coded for frequency; results showed the following fossilized errors: articles deletions (92/94), prepositions (39/43) plurals (54/55), subject-verb agreement (85/46), and general wording (60/69). However, in looking at clauses with errors/100 words, there were 5.29 errors in the 2019 corpus, whereas in the 2020 corpus, there was a slight improvement of 3.35 errors/100 words, indicating that marginal progress was made. These results show many of these errors are interlingual and that students are unaware of their errors that they are making in their spontaneous speech. Alternative methods of instruction are thus needed in EFL education to highlight awareness and self-editing skills.
Robert Long, Kyushu Institute of Technology, Japan
Hiroaki Watanabe, Kyushu Institute of Technology, Japan
Stream: Learning Experiences
This paper is part of the IICEHawaii2021 Conference Proceedings (View)
View / Download the full paper in a new tab/window