Smell and Learning

We've all heard about the unique roll played by olfaction in the panoply of human sensory experience (!). Smells trigger the most enduring and vivid memories, humans are better at discriminating subtly different smells than other sensory inputs, etc. This article in the New York Times retreads a lot of this ground, but one thing stood out to me in particular:

"In one experiment, Dr. Gottfried said, subjects exposed to a single floral scent for just three and a half minutes markedly improved their ability to discriminate among whole families of flower odors. In another, participants soon learned to distinguish normally undetectable differences between one herbal smell and its mirror-image molecular twin if they were given mild electric shocks every time they guessed wrong."

I don't think this study is interesting because of what it tells us about smell, but because of what it tells us about learning. Presumably, participants could not learn to discriminate between the scents under normal conditions. When a shock was applied, however, their learning ability increased. I wonder if subjects merely concentrated more because of the shock threat, or if there's some other mechanism at work. Would they learn at all (or as well) if there were a set number of learning trials before the shocks began? Presumably participants would have the same incentive to learn if they knew shocks were imminent. Obviously, incentives can change learning outcomes. But this study seems to show that incentives can change the capacity to learn, which is surprising to me. Anyone know the mechanism at work, or any other research along this avenue?