Electronic Thesis and Dissertation Repository


Doctor of Philosophy




Dr. Melvyn A Goodale


The ability to adeptly interact with a cluttered and dynamic world requires that the brain simultaneously encode multiple objects. Theoretical frameworks of selective visuomotor attention provide evidence for parallel encoding (Baldauf & Deubel, 2010; Cisek & Kalaska, 2010; Duncan, 2006) where concurrent object processing results in neural competition. Since the end goal of object representation is usually action, these frameworks argue that the competitive activity is best characterized as the development of visuomotor biases. While some behavioural and neural evidence has been accumulated in favour of this explanation, one of the most striking, yet deceptively common, demonstrations of this capacity is often overlooked; the movement of the arm away from an obstacle while reaching for a target object is definitive proof that both objects are encoded and affect behaviour. In the current thesis, I discuss three experiments exploring obstacle avoidance.

While some previous studies have shown how visuomotor biases develop prior to movement onset, the dynamics of the bias during movement remains largely unexplored. In the first experiment I use the availability and predictability of vision during movement as a means of exploring whether obstacle representations might change during a reach (Chapter 2, Chapman & Goodale, 2010b). While the visuomotor system seems optimized to use vision, I found no difference between reaching with and without vision, providing no evidence that obstacle representations were altered. To more directly test this question, in the second experiment participants made reaches to a target that sometimes changed position during the reach (Chapter 3, Chapman & Goodale, 2010a). The automatic online corrections to the new target location were sometimes interfered with by an obstacle. Using this more direct approach we found definitive evidence that obstacle representations were accessed or updated during movement.

In the third experiment, I directly tested the neural encoding of obstacles using functional magnetic resonance imaging (Chapter 4, Chapman, Gallivan, Culham, & Goodale, 2010). When participants planned a grasp movement that was interfered with by an obstacle versus when the grasp was not interfered with, one area in the left posterior intraparietal sulcus was activated. This activity was concurrent with a suppression of early visual areas that were responsive to the position of the obstacle. This study confirmed that the PPC was involved with the encoding of obstacles, and demonstrated that one effect of interference was the suppression of the visual cortical signal associated with the obstacle.

These findings extend our understanding of competitive visuomotor biases. Critically, in a world filled with potential action targets, the selection of one target necessarily means all other objects in the workspace are potential obstacles. My results indicate that the visuomotor biasing signal to inhibit obstacle activity is putatively provided by the PPC, which in turn causes the visual cortical representation of the obstacle to be suppressed. The behavioural result of biasing the visual input is the propagation of this suppression to the motor output - ultimately resulting in a reach which intelligently deviates away from potential obstacles.