Over the past 30 years the study of politics has imitated the discourse of the natural sciences. Articles in this mode of discourse can be characterized as opening with a problem expressed through a review of the literature; continuing with an hypothesis, a methodology, and hypothesis testing; and ending with only narrow conclusions, limited to what can be documented by the mathematical argument. An examination of political science articles from the 1979 "American Political Science Review," however, reveals difficulties and complexities in this straightforward aspiration of political scientists to a scientific ideal. Arguments do not fit together as crisply as the ideal would have it, and the political scientists as authors inevitably find themselves in explanatory, justificatory, reconciliatory, and persuasive tasks that are not part of the idealized version of the scientific report. Problems arise in nomenclature, discussion of the literature, problem definition and significance, interpretation of data, authorial role, and relation to the audience. This rhetorical variety suggests that political science has yet to forge a consistent rhetoric. (Author/HOD)