The birth of subliminal advertising reads almost like a script from a television show. In this real-life story, the spotlight falls on James M. Vicary, an independent marketing researcher.

On September 12, 1957, Vicary called a press conference to announce the results of an unusual experiment. Over the course of six weeks during the preceding summer, he had arranged to have slogans—specifically, “Eat popcorn” and “Drink Coca-Cola”—flashed for three milliseconds, every five seconds, onto a movie screen in Fort Lee, N.J., while patrons watched Picnic. Vicary argued that these messages were too fast for filmgoers to read but salient enough for the audience to register their meaning subconsciously. As proof, he presented data indicating that the messages had increased soda sales at the theater by 18 percent and popcorn sales by 58 percent.

The public reacted with fury. Vicary’s findings played directly into a popular fear at the time that Madison Avenue could manipulate consumers like mindless puppets. The idea that ads might be broadcast subliminally, below the threshold of conscious awareness, seemed akin to brainwashing. On October 5, 1957, some three weeks after Vicary’s event, Norman Cousins, editor in chief of the Saturday Review, wrote an article called “Smudging the Subconscious,” in which he lambasted ad campaigns designed to “break into the deepest and most private parts of the human mind and leave all sorts of scratch marks.” The Central Intelligence Agency soon issued a report on the operational potential of subliminal perception. Vance Packard’s book The Hidden Persuaders—which described Vicary’s claims in detail—became an overnight best seller. As public pressure mounted in response, the U.K., Australia and the National Association of Broadcasters in America all banned subliminal advertising sight unseen.

There was a glitch, however. Researchers tried to replicate Vicary’s findings during this time, but none succeeded. After five years Vicary confessed that his so-called experiment was “a gimmick.” His admission garnered far less attention than his initial publicity stunt. Many in the U.S. and Europe continued to believe that subliminal advertising could shape consumer choice despite all the evidence to the contrary.

Recently, though, psychologists have begun to discover that subliminal messages can sometimes redirect our decisions, but not at all in the way Vicary had proposed. Subliminal messaging cannot override our intentions or commandeer our will. On the contrary, it seems that we are susceptible to these extremely brief suggestions only under special, somewhat limited circumstances. Because these subconscious hints streak through our memory almost as fleetingly as they flash on a screen, they hold no power unless they happen to relate to our immediate goals or natural proclivities.

Backlash and gniksamkcaB
In the decades after Vicary’s experiment, marketers, politicians, film directors and even law-enforcement agencies tried to harness the powers of subliminal persuasion without measurable success. Their intimation tactics typically followed Vicary’s lead, embedding millisecond flashes of words or images in other film clips. For example, in 1978 a Wichita, Kan., TV station received permission from the police to show a glimpse of the sentence “Now call the chief” during a report on the “BTK” serial killer, hoping he might then feel compelled to turn himself in. Unfortunately, the man they were after, Dennis Rader, eluded capture until 2005.

In 2000 subliminal messaging entered the U.S. presidential race. One Republican campaign spot spliced the word “rats” into a segment about Democratic candidate Al Gore. Although “rats” was part of a clearly visible line, “bureaucrats decide,” the less than flattering four letters appeared on screen 30 milliseconds before the rest. Republican candidate George W. Bush claimed it was an accident, but television affiliates quickly pulled the commercial from the airwaves.

Other controversial campaigns have involved “backmasking,” or backward masking—a technique in which audio engineers record spoken words backward onto a track. Proponents claimed that the reversed messages acted subliminally on listeners. In the 1980s religious groups in the U.S. feared that some rock bands used backmasking to convey satanic teachings. Two sets of parents sued British musician Ozzy Osbourne, claiming that backmasked phrases in his songs had prompted their children to commit suicide. The courts dismissed these cases—as they did similar suits brought against rock band Judas Priest—because they found insufficient evidence that backmasking worked. Researchers repeatedly demonstrated that backmasking left no measurable traces in memory. Even so, the uproar led to public record burnings, and in 1983 California restricted the practice.

Also during the 1980s a flourishing trade arose around self-help cassette tapes that claimed to employ subliminally perceptible messages recorded in the correct direction. In 1991, though, Anthony G. Greenwald of the University of Washington and his colleagues proved that these recordings were also ineffective. Greenwald and his team gave 237 test subjects classical music cassettes that held subliminal tips to boost either self-confidence or memory. Unbeknownst to the study participants, who listened to the tapes daily for five weeks, half of the cassettes were deliberately mislabeled. The researchers found that the cassettes had no effect on self-confidence or memory. The participants, however, had a different experience: those who believed that their cassette would increase self-confidence perceived an improvement, as did listeners who expected supercharged memories.

For many scientists this experiment closed the books on subliminal messaging. In 1992 Anthony R. Pratkanis, a psychologist at the University of California, Santa Cruz, and one of the co-authors of the cassette study, wrote that belief in the efficacy of subliminal persuasion offered an example of what physicist Richard Feynman called a cargo-cult science, in reference to the phenomenon in which a tribal society encounters “cargo” from a technologically advanced culture and designs rituals around it. By Feynman’s definition, given as part of a commencement speech at the California Institute of Technology in 1974, cargo-cult science appears to have all the trappings of real science—seeming objectivity and apparently careful experimentation—but is missing something fundamental: its practitioners lack skepticism. Throughout the 1990s subliminal messaging as a research field fell silent, relegated to the realm of reflexology, ESP and other dubious disciplines.

During the past decade, though, psychologists have taken a renewed interest in the topic, and their work has produced some intriguing results. In 2001 Ap Dijksterhuis of Radboud University Nijmegen in the Netherlands, then working with colleagues at the University of Amsterdam, gave students a computerized attention test. Throughout the test he flashed either nonsense syllables or “cola” and “drink” on the screen. Afterward he asked the participants if they would like a cola or a mineral water. The subjects who watched the subliminal messages were more likely to ask for a drink. They did not, however, ask for cola more often. A year later Joel and Grant Cooper of Princeton University replicated the finding, planting subliminal suggestions—the word “thirsty” and images of cola cans—in an episode of The Simpsons. Again the people they subjected to the subliminal messages felt parched by comparison to those who watched unaltered shows.

Drinking the Kool-Aid
To understand why the subliminally cued participants in these studies felt thirstier but not necessarily more inclined to drink cola, consider what happens when you enter a convenience store in search of a drink. First you have to be able to retrieve from memory the name of a beverage. Chances are you will select whatever brand comes to mind fastest. If you drink Coca-Cola all the time, you are probably impervious to any subliminal suggestion to buy another brand. If, however, you sometimes drink Lipton iced tea, messaging that you experience below the threshold of consciousness might sway your choice, making that brand name at least temporarily more accessible in your memory.

We decided to test the theory that Coca-Cola as a brand name may be too deeply imprinted in most people’s memories for subliminal stimuli to have any effect. Working with Jasper Claus at Utrecht University, John Karremans of Radboud and I conducted an initial study in 2006 in which we asked volunteers to perform a computerized attention task. We repeatedly bombarded half of our participants with 23-millisecond flashes of the words “Lipton Ice,” a brand of iced tea. Based on a questionnaire, we determined that Lipton Ice was well suited to our purposes: it is a good thirst quencher but not most people’s first choice. The other half of our subjects saw 23-millisecond flashes of nonsense syllables. After the test, participants had to choose a beverage, either Lipton Ice or mineral water. As expected, the Lipton Ice group chose that brand far more often than the control group did. Again, as in the studies described above, only thirsty subjects reacted this way. Unless you are thirsty, it doesn’t matter which drink brand is foremost in your mind.

In a second study, we used some pretense to give salt drops to half of our volunteers in the hopes of making them thirsty before we showed them the subliminal advertisement. In this scenario, more than 80 percent of the thirsty subjects—and about half of those who said they were not thirsty—chose Lipton Ice. Without subliminal messages, only 30 percent of the thirsty crew and 20 percent of our well-hydrated subjects took the iced tea. In 2011, working with our colleagues Thijs Verwijmeren and Danil Wigboldus, Karremans and I refined these results and demonstrated that the subliminal priming worked only in thirsty test subjects who liked Lipton iced tea but did not drink it regularly. We could not influence people who said that Lipton iced tea was their favorite beverage. This finding might explain at least in part why earlier investigations, which typically involved Coca-Cola, failed to demonstrate a subliminal effect on brand choice. For decades Coke has been the favorite drink among university students, from whom researchers typically recruit their test subjects. Also, these studies did not take into account different levels of thirst.

Other researchers have observed a similar weakness to subliminal persuasion among tired, as opposed to thirsty, individuals. In 2009 Christina Bermeitinger of the University of Hildesheim in Germany, then working at Saarland University, with colleagues at the University of Western Australia, told the subjects that she and others planned to examine the effects of dextrose pills on concentration. They devised two fictitious brands of these pills and designed logos, each of which they presented subliminally to half of the participants while they played a computer game. During breaks the test subjects were offered dextrose pills labeled with the phony brand names. In the end, the more fatigued the participants said they were, the more they gravitated toward the brand they had seen flashed subliminally on the screen.

From these investigations it is clear that an individual’s vulnerability to subliminal suggestion depends on a number of variables, including his or her physical needs and habits. A related effect, subliminal revulsion, also can be triggered under particular conditions. We showed this effect in a more recent study in which we subliminally projected the words “Lipton Ice” during two film clips: a funny sequence from the animated film Madagascar and a disturbing scene from a film about heroin addicts, Trainspotting. After the screening we offered participants Lipton Ice or mineral water. Compared with a control group, who were not subliminally primed, those who saw the brand embedded in Madagascar wanted more Lipton Ice. Those who watched Trainspotting, however, chose it less often. Once again, the subliminal messaging influenced only thirsty test subjects.

Brainwashing at the Supermarket
The idea of subliminal advertising still terrifies many people. Research in the area remains somewhat taboo, and funding is scarce. Programming the Nation, a documentary film released in October 2011, sensationally asks, “Are we all brainwashed? Or have we lost our minds?” Such levels of fear simply are not justified. Certainly no one likes to feel manipulated, but the fact is that our surroundings color our choices all the time, without us consciously realizing it. The aroma of coffee escaping from a bakery can make us crave an espresso; the scent of grilled meat from a restaurant can set our stomach growling. Our research to date indicates that subliminal messages hold sway over our behavior in the same way as these environmental cues do. The thirsty test subject is more receptive to a subliminal hint about a drink just as the hungry shopper is more likely to overfill his or her cart at the supermarket.

To test the potency of everyday hidden persuaders, in 2005 Rob Holland and his colleagues at Radboud devised a clever experiment. The team asked 56 students to list five activities they hoped to undertake during the next few days. Half of the participants encountered the citrus smell of an all-purpose cleaner in the lab, whereas the other half worked in a scent-free room. The first group did not report noticing any odor. Even so, 36 percent of them wrote that they planned to clean their apartments. By comparison only 11 percent of the subjects who worked in the odor-free setting considered cleaning. Holland and his colleagues concluded that the citrus scent had increased the cognitive accessibility of the goal of cleaning. They did not find out how many of the would-be cleaners completed the task, however. Those good intentions may well have disappeared down the memory hole as soon as other, more urgent matters—such as studying for exams—came to the fore.

Indeed, such hints do not last long in our memory. Environmental triggers appear to be most potent in scenarios where we can act on them immediately, a fact that makes them useful in certain commercial settings. When department stores play Christmas music, it is meant to put us in a gift-giving mood and increase sales. In 1993 economists Charles Areni and David Kim of Texas Technical University revealed another way in which music can alter behavior. During several weeks at a wine store they played a variety of music, alternating between classical tracks, such as Antonio Vivaldi’s “The Four Seasons,” and popular tunes, including songs by Fleetwood Mac. They found that the musical selection had no bearing on the total number of bottles sold. Customers listening to classical selections, however, bought more expensive wine than did those listening to pop.

The spending habits of restaurant patrons appear to vary in response to musical cues as well. Adrian North, then at the University of Leicester in England, and his colleagues spent three weeks varying the music in a restaurant dining room from classical to pop to no music at all. When the background track was classical, guests spent an average of $45. By comparison they spent $40 when listening to pop songs and only $39 when there was no music at all.

In some cases, background music can even influence what types of products customers choose. In another experiment, North and his colleagues put a selection of four German or four French wines, equally priced, on display in a British supermarket. On some days the market played German brass band tracks, on others, French accordion music. When interviewed later, very few shoppers could say if they had heard any music. Those customers who heard French tunes, though, more frequently chose the French wines and vice versa.

We have every reason to believe that, just like the music in these examples, subliminal advertising could be used successfully in immediate, day-to-day situations. To have any genuine effect, however, subliminal slogans would have to be short, delivered near the time of a decision, and relate to a person’s immediate intentions or habits. Given such constraints, it is unlikely that subliminal television ads could ever compel consumers days later to buy one brand or another on a weekly shopping trip.

Our work reveals that, in practice, subliminal messaging is far less potent or terrifying than it was first believed to be. It might even be put to good use. A handful of studies have shown that millisecond exposures to the words “angry” or “relax” can have definite, if short-lived, effects on a person’s heart rate and blood pressure. Our subconscious registers many different kinds of suggestions, not just the ones advertisers may be aiming for.

This article was published in print "The Subtle Power of Hidden Messages."