Active inference through self-touch and movement resolves body ownership ambiguity in the rubber hand illusion
Year: 2025
Type: preprint
Abstract: The famous rubber hand illusion (RHI) is evoked by brush-stroking a participant’s real hidden hand in synchrony with the rubber hand that is positioned on a table in front of the participant. Typically, participants report perceiving touch sensations where they visually see the touch by the brush (i.e., on the rubber hand) and that the rubber hand seems to belong to their body. These phenomena are often referred to as referral of touch and body ownership over the rubber hand. While many participants vividly perceive referral of touch and ownership, it is at the same time obvious to them that this is an illusion because they clearly know that the rubber hand is factually not their hand, that it cannot really feel touch, and that it does not really belong to their body. They are therefore in a state of ambiguity in body perception arising from a discrepancy between the feeling of body ownership and the knowledge that they are looking at a rubber hand that cannot belong to their body. Humans can, however, resolve such ambiguities by using active inference, i.e., changing the perceptual state by means of their own actions and, through this process, testing internal hypotheses about possible world states. There are two obvious options available to participants in this scenario to resolve the ambiguity in the RHI by producing counter-evidence against the hypothesis that the rubber hand is part of one’s own body: 1. Actively touch the rubber hand: If touching the rubber hand does not lead to a touch sensation, it is probably not part of the body. 2. Move the hidden real hand: If the rubber hand does not visually move at the same time, it is probably not part of the body. In this series of experiments, we will test the hypothesis that the RHI is reduced or abolished when participants touch the rubber hand with their unaffected hand or when they move their left hand after induction of the RHI. Furthermore, we will test whether the RHI motivates participants to use either of these two strategies spontaneously. This would suggest that humans attempt to resolve body-related perceptual ambiguity by these readily available active inference strategies. We will conduct a series of five experiments that address different aspects of the hypothesis. In all experiments we will use the classical RHI, induced manually by the experimenter with two soft brushes. We will always use the participants left hand to induce the illusion. The rubber hands will be prosthetic gloves in 3 sizes of which we will choose the one that matches best the participant’s hand. Experiment 1 will induce the RHI and assess referral of touch and body ownership over the rubber hand. Then, participants will be instructed to either touch the rubber hand, move the affected hand, or perform (control) actions that is not suitable to disambiguate the RHI (move their unaffected right hand, or touch an object placed next to the rubber hand with their right hand) or to not perform any action. This experiment will follow a within-subject design. The experiment will take place under controlled laboratory conditions. Experiments 2–5 will follow a two-group design, with one group receiving synchronous and the other receiving asynchronous stimulation. These experiments are designed as ‘micro experiments’ meaning that they only take 5-10 minutes per participant. As participants, we will recruit university students on the corridors of the university by direct verbal contact. Participation will be compensated with sweets (such as chocolate bars). To allow for efficient recruiting and testing, Experiment 2-5 will take place in a quiet corner of the main entrance hall at the University of Salzburg. For Experiment 2-5 two experimenters will work collaboratively, with one of them stroking and the other asking the questions and recording the responses. Each of Experiment 2-5 assesses referral of touch and body ownership of the rubber hand, but each experiment implements a specific manipulation or measurement: Experiment 2: Participants are asked in an open format about behavioral inclinations during stimulation with the brushes. Experiment 3: Participants rate on a numerical rating scale several options to express how much they felt inclined to perform each action. The list of actions also includes touching the rubber hand and moving the left (affected) hand. The full list is provided in the section on measured variables. Experiment 4: Participants are allowed to perform one of several possible actions during stimulation. The list of actions is the same as in Experiment 3. Experiment 5: Participants are encouraged to "experiment" and perform actions of their own during stimulation with the brushes. This experiment therefore includes the participants into the research process.
Source: OSF Preprints (OSF Preprints)
Author: Xaver Fuchs
Language: English
Subfield: Human-Computer Interaction
Field: Computer Science
Domain: Physical Sciences
Open Access status: green
OpenAlex ID: W7113709136
