neural process control as internal foraging

Discussing work on the neurobiology of decision mechanisms (in which they find that the signal used to integrate the evidence in favour of a particular action is represented in the same area as is used to represent the action itself), Shalden & Gold speculate:

This intention-based architecture seems to take the hard work of consciousness away from the homunculus. However, another, equally mysterious mechanism seems to be required. If sensory information flows to circuits where it can exert leverage on intentions, plans, and rules, what controls the flow? Which intentions, plans, and rules are under consideration at any moment? The need for a homunculus has apparently been replaced by the need for a traffic cop.

We speculate that, unlike for the homunculus, we already have insights into the brain mechanisms that serve as traffic cop. These are the same mechanisms that allow an animal to explore its environment; that is, to forage. Foraging is about connecting data in the environment to a prediction of reward through complex behavior (Gallistel, 2000). However, in principle, the mechanisms of foraging, like the mechanisms of decision-making, do not need to be tied to overt behaviors. The same principles that apply to visits to flowers could direct the parietal lobe to query the visual cortex for evidence needed to answer a question about motion. More generally, foraging might be related to the leaps our brains make to replace one percept with another (e.g., binocular rivalry), to escape one behavioral context for another, or to explore new ideas. For cognitive neuroscientists, these ideas inspire research on how reward expectation influences sensory-motor and higher processing in association areas of the brain. For the philosopher of mind, these ideas provide an inkling of how properties of the brain give rise to agency and, perhaps, free will.

Shadlen MN, Gold JI (2004) The neurophysiology of decision-making as a window on cognition. In: The Cognitive Neurosciences, 3rd edition. (Gazzaniga MS, ed): MIT Press. [Preprint][Proofs with color figs in back]