1 For a sociology of translator training Anthony Pym Universitat Rovira i Virgili The University of Melbourne Email: anthony.pym@unimelb.edu.au A talk given at the inaugural conference of the World Interpreter and Translator Training Association, Guangzhou, China, November 19, 2016. Version 1.1. May 2017 Abstract. Translator training first depends on specific local circumstances, primarily the distribution of languages and the intensity (frequency and invested social effort) of information flows between languages. The conceptual space thus opened allows us to map each training situation in terms of local objectives, questioning ideologies of universal best practices. It is proposed that, with this wider social view, translator trainers should not neglect the role of translation within foreign-language learning, and they should accept the social responsibility of training people for short-term work in minority languages. The need for such training is clear from the many situations where, especially thanks to free online technologies, whole sections of society think they can translate, while whole generations of language teachers think they know, with equal aplomb, that translation does not concern them. Finally, perhaps paradoxically, the conceptual mapping of training situations can identify instances where we should be training for an international professional community, with its own pressing demands for certification. 1 Why a sociological approach? International cooperation in translator training is much needed and is always welcome. We all benefit from the experiences of others; we all have our favorite tricks that we are sometimes willing to share. There is, however, a basic question to be formulated and repeated in most such circumstances, at least to plant an implicit doubt: To what extent are there best practices that can be disseminated? It is common enough in all kinds of education to invent a timeline that leads from benightedness to enlightenment: what we did in the past was bad, what we do now is much better, and so we are on the road to perfection. Language education, especially, is full of invented historical linearity, as one approach or another gains adepts with an enthusiasm worthy of football fans: transmission in the past, collaboration in the future; structure in the past, communication in the future; grammar translation in the past, collaborative communicative translation in the future, perhaps? Anyone wary of reductive binarisms should take heed and pause to reflect on an obverse model: translator training, like all education, is contextual, situated, grounded, embedded, or whatever other methodological word one might use to represent practical aspects of the social. There are always specific collectivities involved in the production, dissemination, and use of the information subject to translation, just as there are other specific collectivities to be found in the groups that pay to be trained, are paid to give training, and are certainly paid for organizing the paying. That is, to the extent that