O tratamento da prova em vídeo no processo penal e os vieses cognitivos: um alerta sobre os potenciais impactos da visão em túnel na aplicação do standard de prova quanto à autoria

Abstract

The accelerated development of technology, the surveillance society and its consequences in everyday life have been the subject of several academic studies around the world, in numerous scientific fields. Inserted in this scenario, this research is dedicated to investigating some developments in the treatment given to the video as judicial evidence of authorship in the Brazilian criminal justice system, always starting from a guarantor and, therefore, rationalist understanding of the national criminal procedural structure. Adopting a correspondent understanding of truth, the study focused on empirical and epistemological research methods to analyze whether the impacts of the cognitive bias of tunnel vision from a video evidence have the capacity to influence the adequate application of the evidentiary standard imposed for conviction in Brazil. The theoretical frameworks necessary for the development and understanding of empirical research are exposed, including the treatment of constant surveillance of the social body as a public security tool, as well as the possible mental traps to which the interpreter of information transmitted via video is subject, which is done mainly based on the conceptualization of naive realism as treated by Neal Feigenson. These mental traps are addressed especially in terms of their relationship with the cognitive bias of tunnel vision based on the studies by Findley and Scott, to analyze their interaction with the evidentiary standard imposed for conviction in Brazil: proof beyond reasonable doubt. Having established these theoretical premises, the research sought to meet the objective of analyzing whether the video has the capacity to compromise or bias the application of this standard regarding proof of authorship. Using quantitative and qualitative research and analysis tools, without prejudice to their exploratory characteristics, the research focused on criminal judgments handed down throughout 2020 by the TJMG that had at least one video evidence in their evidentiary framework for analyze the phenomenon under study. In preliminary analyses, it was possible to point out data that confirm previous studies on video evidence, such as the evidence that a large portion of judges do not directly watch the video evidence found by Riccio, Silva, Guedes and Mattos, and Fardim's notes and Guedes that the production and valuation of video evidence occurs indirectly in most cases. Regarding the central objective of the work, several variables were applied in different filtering stages, allowing the identification of eight judgments in which the search for criminal authorship started from a video, this being the only evidence available in the records. It was concluded that the particular characteristics of video evidence, intrinsically related to the cognitive bias of tunnel vision, can directly and negatively influence the judge's decision-making process, notably when applying the standard of evidence regarding criminal authorship. It was also possible to point out that the adequate observation of the evidentiary standard can represent an important tool for weakening this bias, confirming the relevance of the topic and the need for further academic studies in this sense.