Attention and cognition are interesting to me. See Skeptical Animus (hyperactive "oh shit circuit" and a hair trigger delete key to erase sensations from memory) and Unicycling Clowns (people who were so distracted by their cell phone use that they failed to see the bizarre occurrence of a unicycling clown passing them on the street) for earlier discussions. Individual variations also seem related to working memory capacities.

What Miller called the informational bottleneck has been recognized as a profound constraint on human cognition. Crudely speaking, there are two ways to manage its effects. One is to "chunk" information so that you can, in effect, pack more material into one of those seven units. As Miller put it, "A man just beginning to learn radiotelegraphic code hears each dit and dash as a separate chunk. Soon he is able to organize these sounds into letters, and then he can deal with the letters as chunks. Then the letters organize themselves as words, which are still larger chunks, and he begins to hear whole phrases." That sort of process is obviously central to many kinds of learning.

The second method for managing the bottleneck—and the one that concerns us here—is to manage attention so that unwanted stimuli do not crowd the working memory. That might sound simple. But as the Swedish neuroscientist Torkel Klingberg explains in his recent book The Overflowing Brain: Information Overload and the Limits of Working Memory (Oxford University Press), scholars are far from agreement about how to describe the relationship between attention and working memory. Does a poor attention system cause poor working-memory performance, or does the causation sometimes work in the other direction?

One common metaphor is that controlled attention acts as a "nightclub bouncer," preventing irrelevant stuff from getting into working memory. A few years ago, Klingberg and a colleague conducted brain-imaging experiments that suggested that a region known as the globus pallidus seems to be highly active when people successfully fend off distraction.

"Why is it that some people seem to reason well and others don't?" asks Michael J. Kane, an associate professor of psychology at the University of North Carolina at Greensboro. "Variability in working-memory capacity accounts for about half the variability in novel reasoning and reading comprehension. There's disagreement about what to make of that relationship. But there are a number of mechanisms that seem to be candidates for part of the story."

One of those seems to be attentional, Kane says. "The view that my colleagues and I are putting forward is that part of the reason that people who differ in working-memory capacity differ in other things is that higher-working-memory-capacity people are simply better able to control their attention."

In other words—to borrow a metaphor from other scholars—people with strong working-memory capacities don't have a larger nightclub in their brains. They just have better bouncers working the velvet rope outside. Strong attentional abilities produce stronger fluid intelligence, Kane and others believe.

That sounds like the anterior cingulate cortex (ACC) and dorsolateral prefrontal cortex (DLPFC) functions (The oh shit! circuit and the delete key) from the Skeptical Animus discussion. The difference here is that such mental editing is being linked to stronger fluid intelligence rather than selective blindness to observational data. It's a conflict since too much of either degrades cognition. You have to pay attention to significant inputs while not overlooking unusual insights. The definition of significant changes with learning.