This study focuses on outcome justifications during the preparation of project applications at the Swedish Research Council during the period of 2002–2004. The study is based on theoretical ideas derived from Michel Foucault’s placement of reproductive processes in the use of language, also inspired by John L. Austin’s pragmatic speech act theory and Judith Butler’s idea of power in authoritarian language, and combines it with modes of representativity used in the methodological approach of qualitative content analysis of institutionally oriented communication. Three analyses are made that aims to clarify the argumentative content used in the language for the preparation of project applications. The result is that assessment varies without obvious reason; contextual and functional representativity is included within the rhetoric’s in assessments as a message within the practice itself; and the overall communication of representativity lacks factuality. The following conclusions are made. Sanctions are justified even if they only express the principled positions or speculation. Assessment does not need to give any information of value beyond what motivates the impression of the sanction itself. Information may be missing in the assessment without prejudice to recommendations for further preparation. Experiences of this might have repercussions for the applicants and for the reviewers. It is pointless to engage in a more honest form of text-processing. One of the merits of the study is to learn how to deal with the practice of academic writing in such a way that it ceases to impress. By this anyone involved in producing research and evaluations can re-evaluate their understanding of their practice. The purpose of this study has taken on a wider significance beyond these examples of language use. As further research is proposed, the broader issue of how research is conducted with the primary interest of strengthening researcher’s own institutional importance.