Japanese scientists’ English for research publication purposes

Error Analysis was developed by early Second Language Acquisition (SLA) researchers as a way to investigate interlanguage and better understand the second-language learning processes. While it is no longer an active branch of research in SLA, it remains a useful tool for those concerned with accuracy in language use. Given the elevated importance of accuracy in English for Research and Publication Purposes (ERPP), this may be one area where Error Analysis may continue to inform research and praxis. The high degree of language precision demanded in ERPP contexts can be a source of frustration for many engaged in scholarly publication, especially those for whom English is an additional language, such as the Japanese scientists studied in this paper. For Japanese scientists and possibly others, an empirical profile of their most frequent error patterns may help them to better deal with accuracy in research writing. With this motivation, this study applies a corpus-assisted error analysis framework to quantify sentence-level grammar errors and identify the most frequent error patterns in the research article manuscripts of Japanese scientists. A corpus of 53 research article manuscripts with 4,495 errors comprises the primary data. Additionally, two raters and a comparison of errors from scientists from six different L1 backgrounds are employed to triangulate the data and investigate the reliability and generalizability of the findings. Findings reveal that the top ten most frequent errors comprise 52.9%, or half of all errors in the corpus, and are dominated by errors with determiners and prepositions.