Cognition in eco, cognition in vitro: measurement, explanation, and contextuality in cognitive sciences

  Cognitive psychology's explanation of human behavior is implicitly based on the hypothesis of functional individuation of cognitive architecture, which induces a direct link between behavioral regularities measured in vitro and those expressed in eco. Cognitive neuroscience, on the other hand, relies on explicit causal modeling of the dynamics underlying cognition, and views human behavior as a contextual expression of its cognitive architecture. Here, we will outline the fundamental methodological divergence between these two approaches, which respectively encourage a reduction of human behavior to a set of individualized psychological effects, and an integration of cognition into its biological and cultural context. We will show how the physical nature of cognitive dynamics induces a contextuality incompatible with the existence of well-defined stimulus-response laws. We will also propose the formal and conceptual framework of active inference as a candidate for studying human cognition in a way consistent with its dynamic and contextual character. In this way, we hope to highlight the necessity of a naturalistic and integrative approach in the study of cognition, and more generally in the explanation of human behavior.