The term "minimal phenomenal selfhood" describes the basic, pre-reflective experience of being a self (Blanke & Metzinger, 2009). Theoretical accounts of the minimal self have long recognized the importance and the ambivalence of the body as both part of the physical world, and the enabling condition for being in this world (Gallagher, 2005; Grafton, 2009). A recent account of minimal phenomenal selfhood (MPS, Metzinger, 2004a) centers on the consideration that minimal selfhood emerges as the result of basic self-modeling mechanisms, thereby being founded on pre-reflective bodily processes. The free energy principle (FEP, Friston, 2010) is a novel unified theory of cortical function that builds upon the imperative that self-organizing systems entail hierarchical generative models of the causes of their sensory input, which are optimized by minimizing free energy as an approximation of the log-likelihood of the model. The implementation of the FEP via predictive coding mechanisms and in particular the active inference principle emphasizes the role of embodiment for predictive self-modeling, which has been appreciated in recent publications. In this review, we provide an overview of these conceptions and illustrate thereby the potential power of the FEP in explaining the mechanisms underlying minimal selfhood and its key constituents, multisensory integration, interoception, agency, perspective, and the experience of mineness. We conclude that the conceptualization of MPS can be well mapped onto a hierarchical generative model furnished by the free energy principle and may constitute the basis for higher-level, cognitive forms of self-referral, as well as the understanding of other minds.

The brain constructs a flexible representation of the body from multisensory information. Previous work on monkeys suggests that the posterior parietal cortex (PPC) and ventral premotor cortex (PMv) represent the position of the upper limbs based on visual and proprioceptive information. Human experiments on the rubber hand illusion implicate similar regions, but since such experiments rely on additional visuo-tactile interactions, they cannot isolate visuo-proprioceptive integration. Here, we independently manipulated the position (palm or back facing) of passive human participants' unseen arm and of a photorealistic virtual 3D arm. Functional magnetic resonance imaging (fMRI) revealed that matching visual and proprioceptive information about arm position engaged the PPC, PMv, and the body-selective extrastriate body area (EBA); activity in the PMv moreover reflected interindividual differences in congruent arm ownership. Further, the PPC, PMv, and EBA increased their coupling with the primary visual cortex during congruent visuo-proprioceptive position information. These results suggest that human PPC, PMv, and EBA evaluate visual and proprioceptive position information and, under sufficient cross-modal congruence, integrate it into a multisensory representation of the upper limb in space.SIGNIFICANCE STATEMENTThe position of our limbs in space constantly changes, yet the brain manages to represent limb position accurately by combining information from vision and proprioception. Electrophysiological recordings in monkeys have revealed neurons in the posterior parietal and premotor cortices that seem to implement and update such a multisensory limb representation, but this has been difficult to demonstrate in humans. Our fMRI experiment shows that human posterior parietal, premotor, and body-selective visual brain areas respond preferentially to a virtual arm seen in a position corresponding to one's unseen hidden arm, while increasing their communication with regions conveying visual information. These brain areas thus likely integrate visual and proprioceptive information into a flexible multisensory body representation.

One of the central claims of the Self-model Theory of Subjectivity is that the experience of being someone – even in a minimal form – arises through a transparent phenomenal self-model, which itself can in principle be reduced to brain processes. Here, we consider whether it is possible to distinguish between phenomenally transparent and opaque states in terms of active inference. We propose a relationship of phenomenal opacity to expected uncertainty or precision; i.e., the capacity for introspective attention and implicit mental action. Thus we associate introspective attention with the deployment of ‘precision’ that may render the perceptual evidence (for action) opaque, while treating transparency as a necessary aspect of beliefs about action, i.e., ‘what I am’ doing. We conclude by proposing how we may have to nuance our conception of minimal phenomenal selfhood and agency in light of this active inference conception of transparency-opacity.

To accurately guide one's actions online, the brain predicts sensory action feedback ahead of time based on internal models, which can be updated by sensory prediction errors. The underlying operations can be experimentally investigated in sensorimotor adaptation tasks, in which moving under perturbed sensory action feedback requires internal model updates. Here we altered healthy participants' visual hand movement feedback in a virtual reality setup, while assessing brain activity with functional magnetic resonance imaging (fMRI). Participants tracked a continually moving virtual target object with a photorealistic, three-dimensional (3D) virtual hand controlled online via a data glove. During the continuous tracking task, the virtual hand's movements (i.e., visual movement feedback) were repeatedly periodically delayed, which participants had to compensate for to maintain accurate tracking. This realistic task design allowed us to simultaneously investigate processes likely operating at several levels of the brain's motor control hierarchy. FMRI revealed that the length of visual feedback delay was parametrically reflected by activity in the inferior parietal cortex and posterior temporal cortex. Unpredicted changes in visuomotor mapping (at transitions from synchronous to delayed visual feedback periods or vice versa) activated biological motion-sensitive regions in the lateral occipitotemporal cortex (LOTC). Activity in the posterior parietal cortex (PPC), focused on the contralateral anterior intraparietal sulcus (aIPS), correlated with tracking error, whereby this correlation was stronger in participants with higher tracking performance. Our results are in line with recent proposals of a wide-spread cortical motor control hierarchy, where temporoparietal regions seem to evaluate visuomotor congruence and thus possibly ground a self-attribution of movements, the LOTC likely processes early visual prediction errors, and the aIPS computes action goal errors and possibly corresponding motor corrections.