A subliminal stimulus is, by deﬁnition, below the sensory threshold, and therefore cannot be perceived at all. The word is used by psychologists to refer instead to a stimulus that is presented in such a way that it is not consciously perceived, and few concepts in psychology have generated quite so much use in pseudoscience, as well as an enduring urban legend.
First, the legend. In 1957 a New Jersey movie theater ﬂashed messages reading “drink Coke” and “eat popcorn” on the screen for a single frame’s duration (1/24 of a second) during a feature presentation. The customers were unaware of having seen the messages, but popcorn sales increased by 58 percent, although Coca-Cola sales increased a comparatively disappointing 15 percent. Although the account of this phenomenon, provided by an advertising man (of course) turned out to be a hoax, the idea that such hidden messages can inﬂuence behavior is an enduring one. Numerous companies, for example, sell self-help audiotapes containing subliminal messages which are alleged to help the listener lose weight, quit smoking, become more conﬁdent, become a better lover, ﬁnd one’s mate, overcome stage fright, make smarter investments, and nearly anything else a person might wish to improve upon. Despite the nearly total lack of evidence that such a product could work as advertised, these tapes are a $50 million-a-year business. Furthermore, large retailers have shelled out hefty licensing fees for in-store music loaded with messages urging customers not to shoplift (and to buy something).
Belief in subliminal appeals to consumers also led to one of the more interesting conspiracy theories of our time in the late 1970s and early 1980s, with a series of books by Wilson Bryan Key (titles included Subliminal Seduction and The Clam-Plate Orgy) in which he claimed that advertisers were concealing graphic sexual images in all sorts of ads, to increase sales to people who had no idea they were being manipulated in this way. The title of The Clam-Plate Orgy, for example, refers to a photo on a restaurant menu in which fried clams were arranged to resemble a large-scale sexual orgy, involving numerous men and women, as well as a donkey. Oddly, nobody noticed this until Key suggested they look for it. He also saw the word “sex” in many places, including in the ice cubes in a glass of gin and in the salt crystals on the surface of a Ritz cracker.
Surely the most absurd manifestation of belief in subliminals, however, came in the late 1970s and early 1980s when rock musicians, especially heavy-metal bands, were accused of hiding paeans to sex, drugs, and Satan in the grooves of their vinyl record albums. Not only were the messages often subliminal, but also they were supposedly recorded backwards. They could be discovered only by the deliberate process of stopping the turntable and slowly turning it backwards by hand. How their discoverers happened to ﬁnd themselves doing this, especially with records they considered immoral and inappropriate to purchase or listen to in the ﬁrst place, usually remained unexplained. Many of the top names in rock were targeted, including Led Zeppelin (whose “Stairway to Heaven” allegedly says “a child is born for Satan, my sweet Satan”), Queen (“Another One Bites the Dust” backwards sounds, with a large helping of imagination, like it says “I want to smoke marijuana, marijuana”), and Ozzy Osbourne and Judas Priest (both of whom were accused of causing teenage suicides with their subliminal messages, though both seemed equally surprised to hear that their music contained them). The most absurd thing about the backward-masking frenzy is the suggestion, even if subliminal audio messages work, that they could work backwards. There is no evidence whatsoever that the human brain is capable of perceiving, consciously or otherwise, anything that is played backwards. Some bands, including Pink Floyd, Iron Maiden, and Electric Light Orchestra (ELO), responded to it all by actually placing clearly audible, humorous backward messages on their albums. ELO’s Fire on High features the funniest: “The music is reversible! Turn back, Turn back!” in a deep, serious voice. Still, like most pseudosciences, the widespread belief in the inﬂuence of subliminals begins with a small grain of truth.
Subliminal perception, in the sense of getting experimental subjects to perceive stimuli without any awareness of having done so, has actually been central to cognitive research for decades. In studies of priming, subjects sit at a computer screen and are shown images (pictures, letters, or words) for very brief durations, measured in tens of milliseconds, to determine how long an exposure is necessary for the subject to detect the image and identify it. A common ﬁnding is that subjects will respond more rapidly and accurately to stimuli they have already seen, even when they were unaware of having seen them the ﬁrst time. Clearly subliminal perception has a slight impact on memory.
Some researchers have taken things a step further, attempting to inﬂuence subjects’ emotional states with subliminal images. In one typical study, subjects were shown a series of photos of an ordinary woman going about her daily chores, always with a fairly neutral expression, and were asked immediately after the slide show to answer a series of questions about her, including whether she seems happy with her life. What they didn’t know was that they had also seen other pictures, presented for fractions of a second. Some subjects saw “happy” pictures, such as puppies and kittens. Other subjects saw scary pictures, including spiders, snakes, and skulls. The result was a small but signiﬁcant difference in how happy the experimental subjects judged the woman to be. When subjects were tested even a few minutes later, rather than immediately upon the end of the slide show, no effect was found. Subliminal images can have an effect, but the effect is very subtle and very ﬂeeting. Absent is any evidence that actual behavior can be impacted. These studies concern only visual subliminals, however.
Where subliminal audiotapes are concerned, no evidence exists to suggest they will work, and the combination of simple logic and the design of the visual research suggests strongly that they should not work. For a visual image to be presented subliminally, it must be presented very brieﬂy. For it to be perceived, even subliminally, however, it must be bright enough and clear enough for the eye to detect it—as bright and clear as the images the person is aware of seeing. The claim made by the marketers of audio subliminals is quite straightforward—the messages are recorded so quietly, relative to the other content of the tapes, that they are inaudible. Consider this in comparison to the research on visual subliminals—no researcher claims that a very dim or blurry image will result in anything being perceived at all. The messages on the tapes, on the other hand, are by deﬁnition too quiet to be heard. No evidence whatsoever exists to show that an unheard message can have any impact on behavior.
Finally, mention should be made of the fact that, in 1974, despite the absence of any empirical evidence supporting the notion that subliminal messages can change behavior, the Federal Communications Commission (FCC) issued a regulation speciﬁcally forbidding the broadcast of subliminal messages. This rule actually became relevant during the 2000 presidential campaign, when two Democratic senators complained to the FCC about a Republican television ad that ﬂashed the word “RATS” for 1/30 of a second. Despite the absence of evidence that such an ad could have any effect other than to waste the money of those who paid for it, our government clearly continues to take the threat of subliminal seduction very seriously.
- Pratkanis, A. R. “The Cargo-Cult Science of Subliminal Persuasion.” Skeptical Inquirer, 16(3) (1997): 260–272;
- Pratkanis, A. R., Eskenazi, J., Greenwald, A. G. “What You Expect Is What You Believe (But Not Necessarily What You Get): A Test of the Effectiveness of Subliminal Self-Help Audiotapes.” Basic & Applied Social Psychology, 15(3) (1994): 251–276.