Linguistes de terrain en quête d’éthique : de l’archivage numérique insouciant à la réflexivité permanente

Abstract

This text is an attempt to look back on two decades of linguistic fieldwork on endangered languages in Yunnan (China) and of activism in favour of opening up fieldwork data. The time frame coincides with two decades of spectacular progress in computer science, and in particular in natural language processing. Looking back, it seems clear that the consistent adoption of open science principles implies that the ethical and socio-political issues raised by the creation, electronic publication and exploitation of multimedia fieldwork corpora must be given a central place (rather than hoping for ready-made solutions to deflect concerns and fend off trouble). These topics call for constant collective reflection involving specialists in natural language processing.

Keywords

• open data
• natural language processing
• archives
• ethics and deontology
• linguistics and language sciences