Abstract
Scientists can choose different claims as interpretations of the results of their research. Scientific rhetoric is understood as the attempt to make those claims most beneficial for the scientists' interests. A rational choice, game-theoretic model is developed to analyze how this choice can be made and to assess it from a normative point of view. The main conclusion is that `social' interests (pursuit of recognition) may conflict with `cognitive' ones when no constraints are put on the choices of the authors of scientific papers, as in an `ideal free speech situation'. Scientific institutions may help to solve this conflict. Lastly, some empirical predictions are offered that can inspire future social research of the refereeing process.