Linguistic trash

Abstract

Linguistic descriptions tend to present language as structure, with data as accessible and transparent as possible. This involves what one might term ‘data scrubbing’; i.e. the deliberate editing out of all elements that don’t fall within the focus of the descriptive effort. Elimination of such ‘noise’ reflects, we propose, at least to some extent, ideological concepts in linguistics. This is also reflected in the reactions of participants, who correct data, keep linguistic specimen pure from loans, etc. Our paper deals with the outtakes of linguistic fieldwork and analysis: noise, languaging that appears to make no sense, mistakes and stammering. By turning our gaze to the unpleasant and incomprehensible, language without meaning that is – as data – usually thrown away, we demonstrate that these parts of communicative practice are meaningful in postcolonial contexts of data collection.