Doctor of Philosophy
Dr. Jody Culham
Tool use is essential and culturally universal to human life, common to hunter-gatherer and modern advanced societies alike. Although the neuroscience of simpler visuomotor behaviors like reaching and grasping have been studied extensively, relatively little is known about the brain mechanisms underlying learned tool use.
With learned tool use, stored knowledge of object function and use supervene requirements for action programming based on physical object properties. Contemporary models of tool use based primarily on evidence from the study of brain damaged individuals implicate a set of specialized brain areas underlying the planning and control of learned actions with objects, distinct from areas devoted to more basic aspects of visuomotor control. The findings from the current thesis build on these existing theoretical models and provide new insights into the neural and behavioural mechanisms of learned tool use.
In Project 1, I used fMRI to visualize brain activity in response to viewing tool use grasping. Grasping actions typical of how tools are normally grasped during use were found to preferentially activate occipitotemporal areas, including areas specialized for visual object recognition. The findings revealed sensitivity within this network to learned contextual associations tied to stored knowledge of tool-specific actions. The effects were seen to arise implicitly, in the absence of concurrent effects in visuomotor areas of parietofrontal cortex. These findings were taken to reflect the tuning of higher-order visual areas of occipitotemporal cortex to learned statistical regularities of the visual world, including the way in which tools are typically seen to be grasped and used. These areas are likely to represent an important source of inputs to visuomotor areas as to learned conceptual knowledge of tool use.
In Project 2, behavioural priming and the kinematics of real tool use grasping was explored. Behavioural priming provides an index into the planning stages of actions. Participants grasped tools to either move them, grasp-to-move (GTM), or to demonstrate their common use, grasp-to-use (GTU), and grasping actions were preceded by a visual preview (prime) of either the same (congruent) or different (incongruent) tool as that which was then acted with. Behavioural priming was revealed as a reaction time advantage for congruent trial types, thought to reflect the triggering of learned use-based motor plans by the viewing of tools at prime events. The findings from two separate experiments revealed differential sensitivity to priming according to task and task setting. When GTU and GTM tasks were presented separately, priming was specific to the GTU task. In contrast, when GTU and GTM tasks were presented in the same block of trials, in a mixed task setting, priming was evident for both tasks. Together the findings indicate the importance of both task and task setting in shaping effects of action priming, likely driven by differences in the allocation of attentional resources. Differences in attention to particular object features, in this case tool identity, modulate affordances driven by those features which in turn determines priming. Beyond the physical properties of objects, knowledge and intention of use provide a mechanism for which affordances and the priming of actions may operate.
Project 3 comprised a neuroimaging variant of the behavioural priming paradigm used in Project 2, with tools and tool use actions specially tailored for the fMRI environment. Preceding tool use with a visual preview of the tool to be used gave rise to reliable neural priming, measured as reduced BOLD activity. Neural priming of tool use was taken to reflect increased metabolic efficiency in the retrieval and implementation of stored tool use plans. To demonstrate specificity of priming for familiar tool use, a control task was used whereby actions with tools were determined not by tool identity but by arbitrarily learned associations with handle color. The findings revealed specificity for familiar tool-use priming in four distinct parietofrontal areas, including left inferior parietal cortex previously implicated in the storage of learned tool use plans. Specificity of priming for tool-action and not color-action associations provides compelling evidence for tool-use-experience-dependent plasticity within parietofrontal areas.
Valyear, Kenneth F., "Perception meets action: fMRI and behavioural investigations of human tool use" (2010). Electronic Thesis and Dissertation Repository. 45.