Is it possible to explain students' conceptions of natural phenomena purely in terms of the interactions between two people and the language they use during an interview? I argue that this hypothesis cannot be accepted on several grounds. First, contextual factors prior to the interview influence the course of its events, and that these in turn influence future events. Second, people have agency over their interactions and the ability to use language creatively in ways that a strong version of language preordination inherent in this hypothesis would not permit. Third, people bring language fluency and ideas to an interview that allow them to grapple with phenomena and issues they might not have previously considered, and formulate conceptions that they can and do use in future interactions. In addition, I argue that the field of science education is able to consider curricular and instructional issues relating to students' conceptions without resorting to the extremes of cultural relativism or intellectual imperialism, and that conceptual change theory addresses both the processes and outcomes of students' interactions.