Unconscious integration of multisensory bodily inputs in the peripersonal space shapes bodily self-consciousness

Abstract

Recent studies have highlighted the role of multisensory integration as a key mechanism of self-consciousness. In particular, integration of bodily signals within the peripersonal space (PPS) underlies the experience of the self in a body we own (self-identification) and that is experienced as occupying a specific location in space (self-location), two main components of bodily self-consciousness (BSC). Experiments investigating the effects of multisensory integration on BSC have typically employed supra-threshold sensory stimuli, neglecting the role of unconscious sensory signals in BSC, as tested in other consciousness research. Here, we used psychophysical techniques to test whether multisensory integration of bodily stimuli underlying BSC also occurs for multisensory inputs presented below the threshold of conscious perception. Our results indicate that visual stimuli rendered invisible through continuous flash suppression boost processing of tactile stimuli on the body (Exp. 1), and enhance the perception of near-threshold tactile stimuli (Exp. 2), only once they entered PPS. We then employed unconscious multisensory stimulation to manipulate BSC. Participants were presented with tactile stimulation on their body and with visual stimuli on a virtual body, seen at a distance, which were either visible or rendered invisible. We found that participants reported higher self-identification with the virtual body in the synchronous visuo-tactile stimulation (as compared to asynchronous stimulation; Exp. 3), and shifted their self-location toward the virtual body (Exp.4), even if stimuli were fully invisible. Our results indicate that multisensory inputs, even outside of awareness, are integrated and affect the phenomenological content of self-consciousness, grounding BSC firmly in the field of psychophysical consciousness studies.

Introduction

Based on clinical and experimental research in humans, it has been proposed that multisensory integration is a key mechanism for self-consciousness. In particular, bodily self-consciousness (BSC) has been shown to depend on the integration of multisensory bodily stimuli (Blanke, 2012, Blanke et al., 2015, Ehrsson, 2012a, Tsakiris, 2010). Research has focused on two central aspects of BSC: people normally self-identify with a given body, which they perceive as their own (self-identification) and they experience their self at the location of their body (self-location) (Blanke, 2012, Blanke and Metzinger, 2009). The notion that BSC depends on multisensory integration of bodily inputs is evidenced by neurological patients who present deficits in multisensory integration together with an altered perception of their own body (Blanke et al., 2004, Blanke et al., 2002), and by experimental manipulations of BSC in healthy subjects using multisensory conflicts (Ionta et al., 2011, Lenggenhager et al., 2007, Petkova and Ehrsson, 2008, Petkova et al., 2011, Salomon et al., 2013). For example, in the full body illusion, viewing an avatar’s body being stroked, while concurrently receiving the same tactile stimulation on one’s own body, makes participants self-identify with the avatar (Ehrsson, 2007, Petkova and Ehrsson, 2008) and induces changes in self-location such that subjects perceive themselves closer to the avatar’s position (Ionta et al., 2011, Lenggenhager et al., 2007).

Under normal conditions, multisensory body-related stimuli occur within a limited distance from the body, which defines the peripersonal space (PPS Serino et al., 2015). Accordingly, neuronal populations have been described both in monkeys and in humans integrating somatosensory stimulation on the body with visual and/or auditory stimuli specifically when presented close to the body (Graziano and Cooke, 2006, Ladavas and Serino, 2008, Rizzolatti et al., 1997). PPS and BSC are thought to involve common neural structures in premotor, posterior parietal, and temporo-parietal cortex (Blanke et al., 2015, Grivaz et al., 2017, Makin et al., 2008) and it has recently been shown that the full body illusion leads to a shift in PPS from the physical body toward the virtual body that participants identify with (Noel, Pfeiffer, Blanke, & Serino, 2015), compatible with an extension of the PPS boundary (Serino, Canzoneri, Marzolla, di Pellegrino, & Magosso, 2015). These data link processing and integration of multisensory stimuli within PPS to self-consciousness, and to BSC in particular (Blanke et al., 2015, Noel et al., 2016).

Conscious experience has also been related to the integration of sensory information in the brain by other authors (Dehaene and Naccache, 2001, Mudrik et al., 2014, Tononi, 2008). Indeed, consciousness is characterized by a unity of experience in which information from multiple sensory modalities is integrated and bound together (Bayne, 2002, James et al., 1981). Recent experimental work has shown that non-visual stimuli that are consciously perceived may be integrated with stimuli rendered invisible through various masking paradigms (i.e. auditory (Alsius and Munhall, 2013, Lunghi et al., 2014), tactile (Lunghi and Alais, 2013, Lunghi et al., 2010, Salomon et al., 2015), olfactory (Zhou, Jiang, He, & Chen, 2010), proprioceptive (Salomon, Lim, Herbelin, Hesselmann, & Blanke, 2013) and vestibular (Salomon, Kaliuzhna, Herbelin, & Blanke, 2015)). It was further shown that even a subliminal auditory and a subliminal visual stimulus can be integrated despite unawareness (Faivre et al., 2014, Noel et al., 2015). It is unknown, however, whether integration of unconscious multisensory events affects self-consciousness, and BSC in particular, which is often considered a distinct and specific form of conscious content (Dehaene and Changeux, 2011, Faivre et al., 2015, Gallagher, 2000).

Previous research on the multisensory basis of BSC focused on the integration of sensory inputs that are presented above the visual and tactile thresholds for conscious access. Yet as it has been argued that BSC is based on low-level and pre-reflexive brain mechanisms, it is possible that the sensory events shaping the experience of the self need not be consciously perceived. While there is no experimental evidence suggesting that the multisensory integration processes of BSC do not require conscious awareness of the multisensory stimuli, interactions between unconscious multimodal stimuli have been shown in humans (see above) (Faivre et al., 2014, Salomon et al., 2015, Salomon et al., 2013) and at the neuronal level in anesthetized animals (Graziano et al., 1997, Meredith and Stein, 1986, Stein and Stanford, 2008). Here, in a series of four experiments, we tested for the first time whether multisensory integration of bodily stimuli underlying BSC also occurs for signals presented below the threshold of conscious perception. We first asked whether tactile stimuli on the body are preferentially integrated with visual stimuli presented within; as compared to outside the PPS, when visual inputs were subliminal and tactile inputs supraliminal (Exp. 1) or when visual were subliminal and tactile inputs were near-threshold (Exp. 2). Next, we investigated whether it is possible to manipulate BSC by using visuo-tactile stimulation administered below the threshold for conscious access. To this aim, we coupled tactile stimulation on the body with invisible synchronous visual stimuli on a virtual body to induce the full body illusion (Lenggenhager et al., 2007) and tested whether this would affect self-identification, as assessed by questionnaires (Exp.3) and self-location, as assessed by the location of PPS boundaries (Exp. 4).