Findings from a matched study of science writers and their scientist sources seem to portray two basic approaches to the science interview: 1) an asymmetric, carefully controlled interview that produces an overview or “wrap-up” of scientific inquiry within the context of an article generally known as interpretive or explanatory, written for readers of publications such as The New York Times, the Boston Globe, Wall Street Journal, the Chronicle of Higher Education, or Popular Science; and, 2) a symmetric, intense, less controlled interview situation that produces more technical articles of greater depth and breadth, perhaps of one scientist's research or of an important era in a scientific domain, written for readers of publications such as New Yorker, Psychology Today, Science, and Discover. The first type of interview emphasizes two language principles—singularity and parsimony—through use of items on a questionnaire representing tight organization, precise wording, unity, and others; the second type of interview emphasizes two other language principles—redundancy and complexity—through use of items such as clarification, figures of speech, and others. “Understanding” as an item was more highly correlated with use of figures of speech and other items represented in the second type of interview.