Abstract
It is widely accepted that the integration of visual and tactile information is a necessity to induce ownership over a rubber hand. This idea has recently been challenged by Ferri et al. (Proc R Soc B 280:1-7, 2013), as they found that sense of ownership was evident by mere expectation of touch. In our study, we aimed to further investigate this finding, by studying whether the mere potential for touch yields a sense of ownership similar in magnitude to that resulting from actually being touched. We conducted two experiments. In the first experiment, our set-up was the classical horizontal set-up (similar to Botvinick and Cohen, Nature 391:756, 1998). Sixty-three individuals were included and performed the classical conditions (synchronous, asynchronous), an approached but not touched (potential for touch), and a 'visual only' condition. In the second experiment, we controlled for differences between the current set-up and the vertical set-up used by Ferri et al. (Proc R Soc B 280:1-7, 2013). Fifteen individuals were included and performed a synchronous and various approaching conditions [i.e., vertical approach, horizontal approach, and a control approach (no hands)]. In our first experiment, we found that approaching the rubber hand neither induced a larger proprioceptive drift nor a stronger subjective sense of ownership than asynchronous stimulation did. Generally, our participants gained most sense of ownership in the synchronous condition, followed by the visual only condition. When using a vertical set-up (second experiment), we confirmed previous suggestions that tactile expectation was able to induce embodiment over a foreign hand, similar in magnitude to actual touch, but only when the real and rubber hand were aligned on the vertical axis, thus along the trajectory of the approaching stimulus. These results indicate that our brain uses bottom-up sensory information, as well as top-down predictions for building a representation of our body.