If arguments are to generate public knowledge, as in the sciences, then they must travel, finding acceptance across a range of local contexts. But not all good arguments travel, whereas some bad arguments do. Under what conditions may we regard the capacity of an argument to travel as a sign of its cogency or public merits? This question is especially interesting for a contextualist approach that wants to remain critically robust: if standards of cogency are bound to local contexts of (...) evaluation, then how may arguments legitimately travel at all? The key to a contextualist conception of cogent travel, I argue, lies in the way local contexts are linked to broader contexts of evaluation by relations of relevance. The burden of the article is to elaborate the different forms these relations can take in the travel of scientific arguments. (shrink)