Grant: Operating Grant

The world constantly bombards us with sensory inputs about many possible motor actions. Imagine walking along a breakfast buffet table. Sometimes, the input is clear and the action is obvious - you see a glass of orange juice, reach to it and put it on your tray. Other times, the situation is uncertain or even conflicting - you see pancakes and waffles and like both; which should you take? The 1st aim of this proposal is to study how premotor and prefrontal cortical areas use conflicting sensory inputs about different actions to select one action to make, using neural recordings in non-human primates that perform conditional motor tasks. We do not respond slavishly to every sensory input. Instead, voluntary behaviour is flexible and often involves arbitrary stimulus-response (SR) rules that we learn by experience. Normal interactions with the world involve selective and often conditional context-dependent responses to only some stimuli, such as the sequences of arm and leg movements elicited by green, yellow and red traffic lights while driving to work. The 2nd aim is to study how neurons in those cortical areas apply arbitrary SR rules to control action. Failure to impose rule-based control on behaviour leads to problems for both the individual and society, including impulsive, inappropriate and even criminal behaviour. We also often use rules to assess the "rightness" or "wrongness" of observed events. We know we should stop when the light is red, so when we see someone drive through an intersection on a red light, we know that what they did was wrong. The 3rd aim is to use MEG imagining methods to determine whether premotor cortical areas in humans can use SR rules to help assess the "correctness/wrongness" of observed sensory events that they watch on a monitor while a computer performs conditional tasks correctly or incorrectly. This would implicate nominally motor cortical circuits in higher-order cognitive processes not associated with action planning.