Differences between natural and unnatural prehension are not inevitable if calibration is allowed

Abstract

 

Reach to grasp movements (prehension) constitute one of the most ubiquitous actions performed by humans. Virtual reality (VR) systems present computer graphics within a viewer-centred frame of reference in order to recreate the visual array. VR systems are potentially powerful tools for studying perception in prehension as they afford precise control over all visual information and allow one to perturb the information in ways that are restricted by Newtonian laws within the physical world. Nonetheless, VR systems do not appear to have fully realised their potential. One major reason for the limitations of VR in prehension research is that people do not appear to show the same levels of performance when reaching-to-grasp virtual items as they evidence when grasping physical objects. One possible explanation for reduced performance in unnatural prehension tasks hinges around the need for the nervous system to calibrate itself. We decided to test this idea directly by asking participants to reach and grasp in four conditions: (A) normal prehension; (B) virtual grasping with no feedback (visual open-loop prehension to a virtual object); (C) virtual grasping with haptic feedback (visual open-loop prehension to a real object); (D) a random mixture of (A) and (B). We used a 45o mirror in order to produce our virtual objects with practically no decrement in the visual information available (apart from the obvious and crucial fact that participants could not see their hand). The normal decrements in performance were observed in condition (B) but not in the identical trials that occurred in condition (D). These findings support the notion that participants can produce normal visual open-loop reaches to virtual objects when calibration is allowed. The results are important because they provide a powerful means of exploring the information used for supporting skilled prehension. The results also have important ramifications for the design and use of virtual reality displays.

 

Footnotes

 MMW and RC acknowledge the support of the Wellcome Trust