Citation practices of expert French writers of English : issues of attribution and stance

In research articles (RA), reporting the results and claims of other authors is a crucial skill as it both demonstrates the writers' familiarity with the literature of the field and allows them to position their own findings and conclusions within the existing body of research, thus creating space to promote their work. Using citations effectively, however, demands considerable linguistic and rhetorical expertise. While several studies have shown that citation causes problems for novice researchers, the specific linguistic problems of expert non-English-speaking researchers have been little investigated. We hypothesize that their problems are unlikely to be the same as those of novices and that cultural and language factors may interfere when citing in a foreign language. To test this hypothesis, we collected a corpus comprising three subsets of articles: 40 pre-publication uncorrected draft manuscripts written in English by expert French researchers in engineering, science and computational linguistics; a comparable corpus of 40 published RAs by native English researchers in the same disciplines; and 40 published RAs written in French by French researchers. The drafts were first examined to detect potential problems with citation; we then checked whether these problems also occurred in the English RAs; if not, this was considered to indicate that it might be a problem specific to French researchers writing in English. The French RAs were then analysed to see which problems could be attributed to the influence of the French language or French citation conventions. The concordancer AntConc 3.2.1. was used for quantitative searches in the corpus. The results revealed that four features related to issues of attribution and stance were particularly problematic for the expert French writers: the use of reporting verbs, of according to, of the would conditional, and concessive if-clauses. The French writers of English used reporting that-clauses far less than the English writers, and with a more restricted range of verbs, a profile of use reflecting that of the French RA subset. The other three problems relate to the different spectrum of values that the English expressions and their French equivalents can take: according to and selon express different degrees of writer commitment to the cited source; the French conditional is widely used to express lack of commitment or distance, a value that is not directly transposable into the English would-conditional; si clauses are frequently used to express concession, unlike if-clauses. French writers of English tended to import all these features specific to French into their English drafts, resulting in many cases of ambiguity as to the writer's position towards the cited source. This cross-linguistic study shows that citing in English is far from straightforward for writers of other languages, and that citation practices are neither language nor culture-free. The influence of the writer's native language and of French academic citing conventions can be clearly perceived in the citing structures and strategies adopted, often leading to a lack of clarity in this respect and thus significantly weakening the strength of the argument