Revisiting a translation efect in an oral language* Karsten Koch Zentrum für Allgemeine Sprachwissenschat (Berlin) his paper examines previous claims that subject-initial word order in the oral Salish languages is induced via translation from English. he analysis concentrates on new ieldwork data from Nłe˚kepmxcin. By taking prosody as its starting point, this study ofers a new look at this issue, and uniquely combines detailed phonetic with syntactic, pragmatic and historical-comparative analysis. Subject-initial forms are, at irst glance, at odds with the basic verb-initial structure underlying all 23 Salish languages. However, examination of the deeper prosodic and syntactic properties of subject-initial forms in Salish suggests that they are in fact native forms induced by pragmatic context and not translation efects. he study therefore cautions against reliance on (written) surface word order in translation studies. . Introduction he research question addressed in this paper is a very simple one, namely whether English word order afects Salish word order in translation tasks. Because Salish languages are oral and have no written form, the study addresses the more general question of to what extent multilingual discourse efects occur in oral transmis- sion, as opposed to translation efects mediated by written form (e.g. Koller 1998). Ater a detailed prosodic and syntactic examination of original ieldwork data from * I am indebted to language consultants Flora Ehrhardt and Patricia McKay, without whom this research would not be possible. Many thanks for helpful commentary go to audiences at the Workshop on Multilingual Discourse Production in Hamburg, and at the Syntax-Phonology Group at ZAS (Berlin). I am particularly grateful to Laura Downing, Caterina Petrone, and Hubert Truckenbrodt for valuable feedback. he inal version proited from the careful work of two anonymous reviewers, whom I thank for thoughtful comments and suggestions. his work has been supported by Jacobs and Kinkade Research Grants from the Whatcom Museum Foundation (Bellingham, Washington); a Social Sciences and Humanities Research Council of Canada Research Fellowship; and two Research Fellowships from the Deutscher Akademischer Austausch Dienst. All errors are my own.