Visual codes or biomechanical constraints: A test of two hypotheses regarding the utility of weber's law in grasping

  • Joseph Manzone School of Kinesiology, Western University
  • Shirin Davarpanah Jazi School of Kinesiology, Western University
  • Robert Whitwell Department of Psychology, The University of British Columbia
  • Matthew Heath School of Kinesiology, Western University

Abstract

Variance in grasping provides a measure of just-noticeable-difference (JND) values that can be used to address whether actions adhere to the relative psychophysical properties of Weber's law. Previous work has shown that naturalistic and pantomime-grasping respectively violate and adhere to Weber's law – a finding taken to evince action control mediated via distinct visual cues. Notably, Utz et al. (2015: Vis Res) proposed JND values do not reflect distinct visual codes; rather, the measure is reflective of reduced biomechanical freedom when the effectors (i.e., thumb and forefinger) reach their anatomical limits for aperture separation. The goal of the present study was to empirically test Utz et al's predictions. Hence, participants performed naturalistic and pantomime-grasping to differently sized objects matched to individual participants' maximal aperture separation (i.e., 50, 60, 70, and 80% of maximal aperture separation). JND values for naturalistic grasping showed an inverse relationship to object size, whereas pantomime-grasping showed a null scaling. Thus, results for naturalistic grasping – but not pantomime-grasping – are consistent with the biomechanical constraint hypothesis. Notably, however, the range of physical object sizes used here (i.e., 55-120mm) were much larger than that used in previous work (i.e., range: 20-70mm) and as a result the current findings may represent an artifact associated with very large (and possibly uncomfortable) grasp postures. Consequently, our current work is exploring grasping responses associated with more 'comfortable' object sizes (i.e., 10-40% of maximal aperture separation) to ascribe whether JND values reflect a measure of visual object properties or biomechanical constraints of aperture shaping.

Acknowledgments: Natural Sciences and Engineering Research Council of Canada