Conjecturas instruídas e juízos factuais: Um elemento da fundamentação de propostas de políticas públicas

The present text deals with the importance of methodological debates in research organizations that formulate studies, proposals and appraisals of public policies. They should be open to diversity and innovation of conceptions, interpretations and methods. The text also expounds this methodological vision: to a large extent, scientific knowledge consists of conjectures and educated guesses, connected overall in a uncertain and fallible web. The connection is often made by means of nondemonstrative logic. Some subjective elements enter into nearly all arguments, inferences and reasonings. Hence, appraisals of almost all of them are personal. In science, alongside abundant disagreements, there are agreements and settlements; but there are no factual truths that may be definitely established by holders of a rank or position in any organization.