Conclusion
The epistemological notion of realism and objectivity, when applied to formal sciences, poses many epistemological problems. We have analyzed the problems that the realist and objectivist encounter in the theories of cognitive science.
We have argued that it is not in regard to the concept of theory, be it understood in the syntactic or semantic manner, that the realist and objectivist positions can be best discussed. We preferred the pragmatic approach for it understands a theory as a set of interacting models.We have also argued that there exist at least three different classes of models that interact constantly in cognitive science. As Braillard and Malaterre (2015) remind us, this interaction has to be better understood. It lacks precision. It seems interesting to see that it is this interaction of models that allows the emergence of an objective apprehension of reality.
The realist and objectivist problem may pose the question of reality. But as Schiffrin expresses it, it is not the role of models to pose ontological questions. Their main function is cognitive, they “enable progress in our understanding of this vastly complex system” (Schiffrin 2009:736). Still, we have argued that the realist and objectivist thesis are impaired in each type of model. Each one is theory laden and allows many epistemic biases to interfere. Formal models are robust only regarding decidability. Their semantic association with physical models is fragile. Experiments have to be repeated in order to ensure some stability of the apprehend of reality. Most of all, conceptual models, expressing conceptual spaces, are dependent upon viewpoints, on culture, if not on pure subjectivity and situations.
Still, we believe, the question of realism and objectivity in cognitive science invites us to see these terms as not only predicable about theory, but also about models. This seems an important change in the exploration of the realist and objectivist questions.