This response has been shaped by Wiese’s perspicuous focus on the philosophy of science and on the phenomenology of perceptual presence. My response to the first topic was to frame the Bayesian brain in terms of control-oriented abduction, where falsification is replaced by “inference to the best prediction” as a criterion for progress. I also reinforced the dependency between active inference and counterfactual processing, which underpins the important case of disambiguatory active inference in Bayesian model comparison. With respect to perceptual presence I proposed a distinction between world-revealing presence and phenomenal unreality (Seth 2015b). World-revealing presence corresponds to objecthood and is associated with hierarchical depth, expected counterfactual richness, and perspective invariance of perceptual hypotheses. Phenomenal unreality transpires when perceptual inference fails to unmix world-related from perceiver-related causes; this corresponds to a loss of “phenomenal transparency” (Metzinger 2003) and depends on violation of counterfactual sensorimotor predictions. Space constraints prevented me considering Wiese’s discussion of the “presence” of cognitive phenomenology, like abstract mathematical and philosophical thinking, in these terms. There is of course a rich literature in linking such phenomena to the body (Lakoff & Nunez 2001), and hence perhaps to active inference where the concept of a “mental action” becomes critical (O'Brien & Soteriou 2009). Space constraints also prevented Wiese from elaborating on interoception, which I consider the most interesting setting for control-oriented active inference, in virtue of the cybernetics-inspired emphasis on homeostasis and allostasis. Interesting questions emerge here about how counterfactual processing plays into the phenomenology of interoceptive experience.
Cognitive scientists have long argued for a continuity between perception and action (Dewey 1896). To close, I suggest thinking instead of a continuum between epistemic and instrumental active inference. This is simply the idea that active inference – a continuous process involving both perception and action – can be deployed with an emphasis on predictive control (instrumental), or on revealing the causes of sensory signals (epistemic). This process intertwines interoception, proprioception, and exteroception, and autonomic and motoric action, with the balance always delicately orchestrated by precision optimisation and counterfactual processing. Putting things this way provides a new way to link “life” and “mind” (Godfrey-Smith 1996) and may help reveal the biological imperatives underlying perception, emotion, and selfhood.
Acknowledgements
I am grateful to the Dr. Mortimer and Theresa Sackler Foundation, which support the work of the Sackler Centre for Consciousness Science. Many thanks to Thomas Metzinger, Jennifer Windt and the MIND group for inviting me to participate in this project, to Jakob Hohwy and Karl Friston for correspondence about abductive inference, and to Wanja Wiese for his excellent commentary.