At a brain-imaging lab in Pasadena last week, a dozen volunteers lay down one after another inside a banging, clanging magnetic scanner to watch movie trailers.
As the two-minute clips unfolded, Caltech professor Steven Quartz and two technicians examined the viewers' brains.
Which film clips best activated areas that indicate memories are being encoded, hinting that this is a trailer people will remember?
Which movie previews made the biggest splash in parts of the brain associated with anticipation of pleasure, suggesting how eagerly this film might be awaited?
Quartz and a Los Angeles marketing company are putting the final touches on a brain scanning service that will be offered to film studios this spring to help them evaluate which trailers might attract the most moviegoers.
This is "neuromarketing" - the emerging field of studying the brain to help advertisers tap into people's unarticulated needs, drives and desires.
It is a field whose existence is one of the most vivid displays of how quickly understanding of the human brain is evolving with the use of functional magnetic resonance imaging, or fMRI.
Physicians and psychologists, philosophers and economists all have become fascinated by activity that can be mapped in the brain as volunteers placed within scanners undergo stress, look at beautiful faces, are trusted or betrayed in economic games, or taste squirts of Coca-Cola and Pepsi.
Scanning centers are cropping up or expanding at more universities, among them the University of California, Davis, which hopes to become an internationally known center for neuroscience, according to the head of its research imaging lab.
Researchers are still refining their understanding of how much these snapshots of the brain at work can reveal about human actions and emotions. No one can look at a brain scan and tell what someone is thinking. But knowledge of how the brain goes about its thinking has grown vastly since the development of fMRIs a decade ago.
"It's pretty darn stunning how far we've come," said Read Montague, a professor at Baylor College of Medicine in Houston, where he directs the human neuroimaging lab and the Center for Theoretical Neuroscience.
Using an ultra-powerful magnet, fMRIs track oxygen-rich and oxygen-depleted hemoglobin molecules in the brain, giving researchers a moment-by-moment portrait of where blood is flowing, and by inference, where large groups of nerve cells are active.
Doctors believe the scans will help them better understand addictive behavior and mental disorders. Other researchers expect brain imaging to offer insights into the physical processes underlying love and bonding, punishment and reward, happiness and decision-making.
Amid this exploration attention has also turned, perhaps inevitably, to one of the most fundamental activities in American culture - buying.
From there, it is a quick and easy leap to both neuromarketing and the consumer group trying to eradicate it.
Both have become active well before anyone can provably say that neuromarketing delivers the goods any better than more conventional research into consumer behavior.
The quest for deeper understanding of what makes us buy entrances businesses, because traditional market research has well-known flaws.
A single, strong personality can dominate a focus group, distorting reactions. People lie to pollsters. Even those who want to be completely open about what they prefer, and why, may not know - unable to penetrate the complex brew of thoughts, emotions and instincts that shapes their desires.
As hard as merchandisers try to fathom customers, behavior remains so difficult for businesses to predict that "from their point of view, consumers are like some kind of random, finicky cat," said Colin Camerer, a professor of business economics at the California Institute of Technology.
"Neuromarketing is kind of like interviewing the brain," he said. "Instead of just asking people what they want, you go right to the brain process."
That is a chilling prospect to some of the ad industry's biggest critics.
A group called Commercial Alert, which campaigns against everything from brand names on sports stadiums to junk food in schools, has tried to shut down neuromarketing research at Emory University in Atlanta.
"Any small increase in the effectiveness of advertising can cause tremendous disease, death and human suffering," said Gary Ruskin, executive director of the group he co-founded with Ralph Nader.
Ruskin doesn't want it to be any easier to push unhealthy products that could contribute to obesity, alcoholism or violence. He doesn't want children to become more vulnerable advertising targets.
It is no secret that some ad campaigns prey on the less-loved elements of our own psyches - fear and greed, lust and gluttony - to promote consumption that our better judgment might condemn.
Ruskin argues that neuromarketing is unethical precisely because it could facilitate that process, aiding sales of potentially harmful products.
Those doing the work disagree, saying they've addressed ethical issues by following standard research practices - working with willing volunteers, full disclosure and a technology that does not harm the subject.
They understand the anxiety, though.
"This starts to touch on the part of our social selves that we most treasure. People are going to get sparky about it," said Montague, who is organizing an April neuromarketing conference that will cover issues such as product development and brand identity.
Rather than trying to limit research into the brain's role in human behavior, he said, consumer groups and others should be engaging in serious debate about how the findings should be used.
That debate will go far beyond advertising.
"I could imagine pollsters using this" and politicians, said Brian Knutson, a Stanford neuroscientist studying anticipation of pleasure.
Montague suggests that lawmakers someday may have to decide whether businesses should be allowed to use brain scans in job screening, just as they now use psychological tests in an effort to weed out untrustworthy employees.
Courts might have to consider whether neuroimaging studies of a suspect's intent could be admitted as evidence.
And what about neural profiling? Could the technology, which already is shedding more light on risk-taking behavior and impulse control, one day identify potential criminals before they act?
Less Orwellian possibilities also abound.
Montague thinks it's likely that researchers will continue to demonstrate that children's brains absorb information much less critically than adults' brains - building a stronger scientific case for tougher regulation of ads aimed at kids.
Caltech's Camerer foresees less waste if products can be tailored more closely to consumers' desires.
"Once you start to think about the goal being to predict when people will really do something better than their words (predict), you can start to look at all kinds of behavior," from voting to personal commitment, he said.
When your sweetheart says he'll buy that engagement ring, Camerer asks only half facetiously, is the fear circuit in his amygdala screaming no?
Questions like these presume knowledge that, so far, we do not have, the nation's leading brain researchers agree. It has not been demonstrated that anyone can predict behavior based on what they see in brain scans.
A handful of marketers have started trying, though.
Quartz, who directs the Social Cognitive Neuroscience Laboratory at the California Institute of Technology, hopes to build a database of neural responses to movie trailers that could help separate box office bombs from megahits.
Gerald Zaltman, a Harvard emeritus business professor and author of "How Customers Think," has used brain scanning to augment other research that his marketing firm, Olson Zaltman, has done for several consumer products companies.
BrightHouse, an Atlanta marketing strategies company, has helped fund research into the neural basis of satisfaction and plans two or three more brain imaging research projects soon, said company scientist Justine Meaux.
All stress that even if research could give them a better idea of what customers want, marketers still would be light-years away from convincing someone to buy.
"We'll never find this 'buy button' you might hear about," some imagined spot in the brain that lights up when every customer is ready to spend money, said Clint Kilts, the Emory neuroscientist targeted by Commercial Alert for his consumer preference work. "We're not that good, to be perfectly honest."
But if you build a better mousetrap because you understand everything that people want a mousetrap to do - or if you tap into the brain to figure out why someone who's never seen a mouse might like that mousetrap - have you found something as good as the "buy button"?
Neuromarketers say no. Commercial Alert says yes.