I have tried to show that useful analogies between PP accounts and classical ideas in theory of science run deeper than portrayed in Seth’s target paper. Based on such analogies, I have argued that a proper treatment of active inference needs to be more sophisticated than Seth’s threefold distinction. In particular, Seth blurs a whole range of ways in which models can be falsified.
Furthermore, I have suggested that Seth’s predictive processing account of perceptual presence may profit from taking not just the counterfactual richness of generative models, but also their degree of perspective-dependence and their causal encapsulation into account (as mentioned above, this suggestion is inspired by Jakob Hohwy’s work). I have proposed a way in which examples of possible combinations of these features can be explored, which may serve as a useful tool for future research.
Thomas Kuhn (1962, p. 88) writes that “normal science usually holds creative philosophy at arm's length, and probably for good reasons”. I thus hope that research on predictive processing and consciousness has not yet reached the phase of normal science, so that this commentary can still make a humble contribution.
Acknowledgments
I am grateful to two anonymous reviewers, and to Jennifer Windt and Thomas Metzinger especially for providing a vast number of comments and remarks, which helped tremendously in revising the first draft of this paper. This comment was written with support by a scholarship from the Barbara Wengeler foundation.