Robert Whitwell¹, Alice Tan¹, Ana Victoria de Meira¹, Michelle Wong¹; ¹The University of Western Ontario

We used a grasp-adaptation paradigm to test two models of grasp planning: a ‘classic’ model in which the target is coded separately as size and position, and an integrated model that postulates the target is coded as a set of egocentric ‘grasp points’. Participants (N=144) reached, without visual feedback, for virtual targets and grasped real (haptic) ones whose sizes were either the same, larger, or smaller. Two adaptation sequences were administered, each comprised of a series of baseline, adaptation, and then washout trials. On baseline and washout trials, the virtual and haptic objects were identical. On adaptation trials, the target's haptic size was either always larger or always smaller than its virtual size. Consistent with prior work, we observed strong aftereffects in the first (control) sequence. We administered a second sequence to test if the aftereffect would generalize to a novel target orientation and/or position while preserving target size. Notably, the classic model implies that the aftereffect should generalize under such circumstances. In contrast, the integrated model implies the aftereffect should fail to generalize, because a novel target position and/or orientation requires a novel set of grasp points. Surprisingly, generalization depended on the direction of the grasp aperture’s induced adjustment and, therefore, neither model can fully accommodate our findings. Specifically, the aftereffect induced by the larger haptic object generalized to novel target position and/or orientation, whereas the aftereffect induced by the smaller haptic object did not. In the latter condition, novel target orientations resulted in the strongest attenuation of aftereffects. Considering the asymmetric consequences of under- vs. over-sizing grasp aperture, our findings suggest grasp planning prioritizes haptic size when compensating for an under-sized grasp aperture. When under-sizing is not applicable, however, haptic integration is subsumed under de novo coding of the orientation of the hand opposition space.

Acknowledgements: This work was supported by an NSERC Discovery Grant to RLW (RGPIN-2022-05050).