To a cognitive neuroscientist, success in life depends largely on how well your brain shapes actions to fit both your goals and the context in which you act. If you need to cross the street, first check the traffic. If you're feeding your new in-laws, serve better wine--and different jokes --than you do for your poker buddies.

Such context-driven "cognitive control," thought to reside in the brain's prefrontal cortex, has lately become one of neuroscience's hottest topics, inspiring hundreds of papers regarding its role in everything from academic and sports performance to depression and gambling. Yet it remains unclear just how the prefrontal cortex exerts this control. Researchers have explored models emphasizing (not necessarily to mutual exclusion) attentional control, conversations between the prefrontal cortex and limbic areas, and the prefrontal cortex's sensitivity to context.

On Monday at the Society for Neuroscience conference in Atlanta, David Badre, a post-doc at the University of California, Berkeley, described an experiment bolstering another emerging model, that of a "hierarchy" of control. In this view, Badre says, "we respond to increasing cognitive challenges not on a continuum, but in leaps as different areas of the prefrontal cortex activate." That is, we throw more light on a problem not by sliding a dimmer switch, but by flipping a series of toggles that successively activate chunks of the prefrontal cortex from back to front. Recognizing this is an important early step in understanding how cognition works, and it could help lead to better treatment for patients who have suffered strokes or other injury to the prefrontal cortex.

Badres experiment built on imaging studies in the late 1990s, which discovered that as test volunteers faced more variables in a given problem, they recruited first the premotor cortex (just behind the prefrontal cortex) and then, as context became more important, the pre-premotor cortex, a section of the prefrontal cortex directly in front of the premotor cortex. A 2003 Science paper showed that a third area, the lateral prefrontal cortex, would kick in if solving the puzzle required weighing past events or ongoing goals. Interestingly, the later, more "advanced" areas never lit unless the earlier ones preceding them had activated--but once on they stayed active longer, as if monitoring the more complex situation.

Badre, by confronting his volunteers subjects with a simple puzzle that he could make frighteningly complex, produced evidence that this hierarchy includes a fourth area, the frontal pole, at very front of the lateral frontal lobe. He asked subjects to classify paired objects shown in illustrations as either the same or different. Simple enough--until he told them they had to start making the same-different call based on particular characteristics (shape, size, texture or color) and that the characteristic of relevance depended on whether a square surrounding the objects was red, blue, yellow or green. By varying these rules he could create an increasingly difficult series of puzzles. The most complicated puzzles consistently lit up the frontmost part of the prefrontal cortex.

"This seems to nicely confirm this hierarchal sequence," Badre says. "Now we're studying stroke patients to see if lesions in the rearmost areas--the ones they use first--stop the ones further up from working." His early results suggest that is the case. If that holds up it could help explain how strokes affect mental powers. From there Badre hopes to explore how other neural systems might modulate this sequence. But that must wait for now, he says, because sometimes you have to do the simpler things first.