During SC's time at USC, he was fortunate to take a class on Brain Theory from a legend, Michael Arbib. Prof. Arbib's research focuses on "mirror neurons", a part of the brain which helps us imitate actions that we see performed by other people. It happens that they may also play a role in explaining how human speech evolved, which we might discuss some other time.
Because the action of mirror neurons depends crucially on interaction with the human visual system, we spent a lot of time learning about approaches to simulating vision. One key insight of computational research on vision is the idea that we don't perform spatial computations on the entire glorious mosaic of visual information, but rather pick out particular surfaces and edges which provide opportunities for interaction, called affordances. While not all neuroscientists accept the existence of affordances, theorists who work with them argue that they provide a solid explanation of how we manage to sometimes do very mistaken things in our interactions with objects in the world. For example, a product liability expert suggests that affordances explain how a person might misuse a child's play table as a stepping stool, erroneously extracting the affordance for stepping from the fact that it was a low, flat object. Considerations of mass and rigidity, not to mention actual identity, sometimes fail to enter into the computations that precede a decision to act.
At least, that better be this woman's defense.
Comments