ABSTRACT
This paper reviews the literature on health and female homosexuality in Brazil and, along the way, outlines an alternative approach to reviewing academic literature. Rather than summarising the contents of previously published papers, we relate to these publications primarily as partakers in the creation of knowledge. Inspired by Actor-Network Theory (ANT), we apply ethnographic methods to understand the papers as study participants endowed with action. We also draw on the notions of inscription and intertextuality to trace the complex relationship between the findings in the articles and the realities outside of them. We claim that 'evidence' is the product of translational processes in which original events, such as experiments, blood tests and interviews, are changed into textual entities. In addition, text production is seen as an absorption of everything else surrounding its creation. When events are turned into articles, the text incorporates the political environment to which original events once belonged. We thus observe a political text inscribed into the written evidence of sexually transmitted infections, and the practice of publishing about scientific vulnerabilities emerges as political action. In contrast with traditional ways of reviewing literature in medical scholarship, this article offers a reminder that although there is a connection between textual evidence and the reality outside publications, these dimensions are not neutrally interchangeable.