There is a growing perception that English used as a lingua franca does not need to resemble English as a native language, but can include a great deal of lexicogrammatical variation. However, a more fundamental matter is whether research articles written in English need to conform to the dominant linear, deductive, ‘Anglo’ pattern of text organization, or whether non-native English speakers should be free to transfer rhetorical patterns from their L1s into English, such as, e.g., an inductive, indirect, end-weighted form of argumentation, perhaps with a less-assertive conclusion. Hinds (1987) describes the latter style of writing as “reader-responsible,” as opposed to the “writer-responsible” Anglo-American style, arguing that it requires a great deal more inferential work on the part of the reader. Yet from a relevance theory perspective it appears unlikely that a culture would choose to impose unnecessary additional processing effort on readers. The perception of difficulty is an etic perspective: analysts from “writer-responsible” cultures imagine the processing effort that would be necessary in their culture to make sense of “reader-responsible” texts. Indirect, inductive rhetorical styles do not cause problems for readers accustomed to them. Given that most academic writing in English is for an international audience, non-native English-speaking researchers – and indeed native English speakers too – should be free to adopt a range of styles, or some sort of heterogeneous hybrid, depending on their perceptions of their readers’ expectations. A further issue is whether researchers who have reservations about ‘Anglo’ styles of writing, e.g. in the social sciences or literary and cultural theory, should encourage non-native English speakers to imitate the noun-heavy, nominalized, passivized, syntactically-complex style dominant in these fields.