RESEARCH ARTICLE Effects of visual uncertainty on grasping movements Erik J. Schlicht Paul R. Schrater Received: 28 June 2006 / Accepted: 13 April 2007 Ó Springer-Verlag 2007 Abstract To successfully lift an object, a person’s fingers must be moved to locations where forces can be applied that are sufficient for maintaining contact and that allow for easy object manipulation. Obtaining such finger positions becomes more difficult when there is perceptual uncer- tainty about the location of the hand and object. However, knowledge about the amount of uncertainty could be incorporated into grasp plans to mitigate its effect. For example, during peripheral viewing the fingers could open wider to avoid colliding with or missing the object. The goal of this study is to determine the degree to which people incorporate their understanding of visual uncer- tainty when making a precision grasp. To investigate, subjects reached to a spatially fixed object whose retinal location was varied by fixating points 0–80° to the left of the object. This manipulation controlled the visual uncer- tainty of the hand and target without affecting the kine- matic demands of the task. We found that people systematically changed their grasping behavior as a func- tion of the amount of visual uncertainty in the task. Spe- cifically, subjects’ maximum grip aperture increased linearly with target eccentricity. Moreover, the effect of visual uncertainty on finger trajectories could be captured by a single dimension of change along an axis. Together, these findings suggest that the sensorimotor system esti- mates visual uncertainty and behaviorally adjusts for it during grasping movements. Keywords Visual uncertainty Reach and grasp Maximum grip aperture Principal components analysis Displacement vector Introduction Several studies in motor control have demonstrated that the sensorimotor system incorporates estimates of motor uncertainty to improve the success of point-to-point reaching movements. For example, previous research has shown that people use knowledge about their motor uncertainty to minimize end-point variance (Harris and Wolpert 1998), avoid obstacles (Sabes and Jordan 1997; Hamilton and Wolpert 2002), and maximize the expected payoff of a reach (Trommershauser et al. 2003, 2005). Other studies suggest that the sensorimotor system also uses estimates of visual uncertainty and optimally com- bines them with motor uncertainty, both across reach trials (Kording and Wolpert 2004; Cheng and Sabes 2006) and during feedback control of a reach (Wolpert et al. 1995; Saunders and Knill 2004). However, these studies do not address how visual uncertainty is used in grasping move- ments, where the goal of the task is to manipulate an ob- ject. The purpose of this paper is to investigate the effects of visual uncertainty on grasping behavior. The consequences of perceptual uncertainty for produc- ing a successful grasp are complex and currently unknown. In contrast to point-to-point tasks where the objective of the movement can be expressed in terms of the final state of a E. J. Schlicht P. R. Schrater Department of Psychology, University of Minnesota, Minneapolis, MN, USA P. R. Schrater Department of Computer Science and Engineering, University of Minnesota, Minneapolis, MN, USA E. J. Schlicht (&) N218 Elliott Hall, 75 East River Road, Minneapolis, MN 55455, USA e-mail: schl0360@umn.edu 123 Exp Brain Res DOI 10.1007/s00221-007-0970-8