Using proprioception to get a better grasp on embodiment

We interact with, interpret and understand the world around us through our senses. We see our environment, touch the things in it, feel the ground beneath our feet, and know how we move within our surroundings. This rich multisensory information stream not only forms the basis of our impression of the world, but also establishes our intrinsic sense of self within that world. By integrating these sensory streams we perceptually separate ourselves from the environment around us: an ‘I’ among a world of ‘other’. For instance, when we see and feel a touch to our hand we know that hand is a part of our body and it belongs to us. When we see another person being touched, without feeling that same touch, we know that other body is not ours. Although we take for granted the simple and basic sense that our arms, legs and bodies belong to us, this innate perception of ownership or embodiment is much more malleable than we might ever have imagined.

Humans have long reflected on the idea of the self, the sense that you observe the world from an internal perspective that is distinct from others and separate from the external (Reid, 2002). Until recently, investigation of the concept of self has been confined to the realm of philosophy. However, in 1998 Botvinick and Cohen demonstrated that through a simple manipulation of seen and felt touch, able‐bodied individuals could be fooled into perceiving that a fake hand was their own hand (Botvinick & Cohen, 1998). This phenomenon of inducing cognitive embodiment of a hand that is not actually part of the body is known as the ‘rubber hand illusion’, and it is simple to generate. A participant is seated with his or her hands outstretched comfortably in front on a table top, with one hand hidden from view behind a screen. A rubber hand is positioned in place of the hidden hand in the same orientation and within the participant's correct visual frame of reference. An experimenter sits across from the participant and simultaneously strokes the rubber hand (that the participant can see) and the participant's hidden hand (that the participant cannot see) in the same way with the same timing. After a few seconds most participants will feel the experimenter's touch as though it was occurring on the rubber hand; as though it was their own hand. During this transition the participant cognitively incorporates the fake hand into his or her self‐image (Botvinick & Cohen, 1998), while simultaneously, dis‐incorporating his or her actual hand (Newport & Preston, 2011). This embodiment illusion has opened the door to experimental approaches that are now beginning to tease apart the brain circuits and cognitive/perceptual mechanisms that underlie the sense of self (Blanke, 2012).

In this issue of The Journal of Physiology, Héroux and colleagues (2018) present their work providing insight into the mechanisms of self‐attribution. Using their ‘grasp illusion’, a new variant of the rubber hand illusion, this team separated out individual components of the sensory streams to explore how the brain integrates information about relevant features to generate the sense of embodiment. The grasp illusion arises when a participant, who cannot see his or her hands, grasps a hidden fake finger positioned such that it appears to be in the same orientation as his or her other actual finger. Simultaneously, the tip of the participant's actual finger, which is sitting at a distance directly below the fake finger, is lightly clamped so as to feel as though it is being grasped. When this is done, the participant perceives that his or her two hands are aligned more closely in vertical space than they actually are.

In this study, Héroux and colleagues found that the perception that the fingers are more closely aligned occurs very rapidly after the proprioceptive input from the grasp, and that a reported sense of ownership over the fake finger also occurred at the initial grasp and grew stronger over the 3 min of the experiment. This is interesting because it suggests that body ownership is not necessarily tied solely to vision and touch, and that the brain likely uses relevant sensory inputs such as proprioceptive position sense from the skin to generate a feeling of self‐attribution. The authors also investigate the influence of physical characteristics of fake fingers on ownership and perceived location. Participants grasped fake fingers that were hot or cold, pliable or hard, rough or smooth, and rectangular or oddly shaped. The characteristics of the fake finger did have an effect on the sense of embodiment and perceived alignment in vertical space even though the fake finger was never actually seen during the experiments. It just happened to be ‘like’ a finger, and perceived as if it were a finger. Most interestingly, grasping a cold finger, a rough finger, or an oddly shaped or rectangular finger all reduced embodiment. This is a particularly compelling finding because it suggests that the brain has an innate representation or pre‐supposition of what a finger should feel like. Thus the sensory integration system likely weights levels of embodiment within the context of physical relevance. This is important because there is debate as to how flexible our sense of ownership is and this study provides evidence that in order to be incorporated into the self‐image, the fake body part should feel, at least nominally, like an actual body part (Hohwy & Paton, 2010). In addition to exploring new methods for understanding embodiment, Héroux and colleagues also stepped outside of the traditional P‐value approach for interpreting data. Human perception is complex and restricting analyses to results of yes or no risks missing underlying nuances of experience. The authors achieved an unprecedented level of transparency by presenting individual participant data overlaid with the summaries and by providing raw data and their analysis code in an online repository.

Additional information

Competing interests

None declared.