You're shopping online and happen upon that perfect dress. It's so nice you cough up an extra $10 for overnight delivery. But then you learn it is out of stock and won't be shipped until next month.

So do you still choose overnight delivery? Probably not. But if getting the dress a day or two early was worth the extra dough today, why wouldn't that earlier gratification be worth it next month?

The decisions are inconsistent, almost as if they were made by two different people. And that's not surprising, say researchers in the new field known as neuroeconomics, because the decisions really are being made by two minds -- two minds inside the same head.

At least that's what researchers saw when they performed brain imaging studies of 14 Princeton University students choosing between immediate or delayed gratification. Each was presented with choices between an immediate prize -- a $5 to $40 Amazon.com gift certificate -- and larger rewards that could be had only weeks later.

As scientists from Princeton, Harvard and Carnegie Mellon report in today's issue of the journal Science, the images showed that two areas of the brain -- the limbic system, which is associated with emotion, and the lateral prefrontal cortex, the center for abstract reasoning -- were both active when the students considered the choice before them.

"The outcome seems to be determined, at least in part, by a competition between these areas," said Dr. Jonathan Cohen, director of Princeton's Center for the Study of Brain, Mind and Behavior and also is a faculty member at the University of Pittsburgh.

The rational brain was engaged in all of the decisions, the researchers noted. But when immediate gratification was a possibility, the emotional brain also became activated.

When the students opted to delay gratification, the rational mind was more active than the emotional mind, as shown in functional magnetic resonance imaging scans. When they chose the immediate reward, the fMRI scans showed that the activation of both areas was comparable, with perhaps slightly more activity in the limbic system.

This tension between immediate and delayed gratification is well-recognized -- Aesop addressed it with his fable about the hard-working ant that stored food for the winter and the lazy grasshopper that lolled in the sun.

And it plays a role in such everyday choices as spending or saving money, having dessert or sticking to a diet, watching TV or doing homework, said George Loewenstein, a professor of economics and psychology at CMU and one of the study co-authors.

Economists have long tried to explain the inconsistencies. In one classic problem, people choose $10 today instead of $11 tomorrow. But when offered $10 one year from today and $11 one year and one day from today, they choose the $11 reward.

For a long time, Loewenstein said, some economists have argued that the human brain "is hard-wired to have a particular attitude toward time delay." Called exponential discounting, it suggests that people have a predictable response as a reward becomes more imminent or as the size of a delayed reward increases.

This phenomenon seems to hold true in nonhuman species, noted George Ainslie of the Coatesville VA Medical Center and John Monterosso of UCLA in a commentary also published in Science. But humans are far more inconsistent, with most people likely to change their preferences if the conditions are right.

"The study [by Cohen, Loewenstein and the others] is the first step in an important direction," Ainslie and Monterosso wrote, "but is not yet enough to specify the mechanism of preference reversal."

"I wouldn't pretend that we have any answers here," Cohen said yesterday. But if the study has identified a general principle -- that decisions are made by many competing minds inside each person -- then it will be important for neuroscientists to develop an understanding of each of those brain areas.

"It has implications for all sorts of behaviors," he added, including addictive behaviors.

Loewenstein has specialized in behavioral economics, using psychology to understand economic behavior. With neuroscience, including brain imaging, now on the cutting edge of psychology, it was only logical that those tools be brought to bear on economics, he said. The first workshop on neuroeconomics was hosted by CMU in 1997.

Originally pursued by just a handful of researchers, neuroeconomics "is becoming very, very hot," with the numbers of researchers growing almost by the day, he said.

Just what it all means is still unclear. It's possible that neuroscience may result in incremental advances in economic theory, such as refining the equations used to explain economic behavior. But some researchers suspect it could result in radical changes; "we may need to rethink the whole notion of rationality," he noted.

In the debate between these incremental and radical approaches, Loewenstein admits to being ... of two minds.

"I'm trying to keep one foot in each camp," he said.