Finding translations. On the use of bibliographical databases in translation history

In any study of translations one must first decide what is to be counted as a "translation" and how such things are to be found, usually through recourse to bibliographical databases. We propose that, starting from the maximalist view that translations are potentially everywhere, various distribution processes impose a series of selective filters thanks to which some translations are more easily identified and accessible than others. The study of translation must be aware of these prior filters, and must know how to account for them, and sometimes how to overcome them. Research processes then necessarily impose their own selective filters, which may reduce or extend the number and kinds of translations given by prior filters. We present three research projects where the play of prior and research filters is very different. For one-off large-scale relational hypotheses, the Index Translationum is found to be relatively cost-efficient. For more detailed objects such as translation flows from Spanish into French in a specific period, a book-industry database offers significant advantages. And for a study marked by a paucity of texts, as is the case of translation from Korean into English following the Korean War, a combination of databases is necessary, the most useful turning out to be Amazon.