Normative Practices of Linguistic Correction on Hatsugen Komachi: A Corpus-Assisted Approach to (Meta)discourses around Linguistic 'Mistakes'

Linguistic ‘mistakes’ (quotes intentional) is a negatively evaluated concept constantly reinforced through social institutions such as the educational system. The act of labelling something as a ‘mistake’ is one of the most visible manifestations of language ideologies and, as such, provides a window into the values we live by. In this contribution, I propose a corpus-assisted analysis of what users of Hatsugen Komachi, a text-based online question-and-answer forum addressed mainly to the female public, frame as linguistic ‘mistakes’. I focus on the search item <間違*> machiga* (‘mistake’), and more specifically on what lexical items it co-occurs with and the (evaluative) meanings it acquires in context. Specific attention is given to the overt expression of certain emotions, which are taken to be culturally conditioned and normative (hence ideological) reactions to what is perceived as linguistically inadequate.