Images not consciously perceived can affect behavior. A recent experiment showed that subliminally presented happy faces caused thirsty participants to pour more of a beverage and to consume more. Happy faces also increased participants willingness to pay for the beverage. Angry faces presented subliminally had the opposite effects. The effects were not just statistically significant but also economically significant: thirsty participants poured more than twice as much of the beverage after being presented subliminally with happy faces than after being presented subliminally with angry faces. Despite these effects, participants were not aware of any influence on their any influence on their behavior, nor of any change in their mood.[*]
Hoping that subliminal messaging could be used to improve fans’ attitudes toward referees (and public attitudes toward regulation more generally), I attempted to add subliminal messages to a basketball video. At 5 seconds, 41 seconds, and 52 seconds into the video below, I superimposed for one frame the word “happy”, a smiling face, and the word “good.” One frame of Flash video has a duration of 33 milliseconds. The primes in the above experiment were flashed for 40 milliseconds. However, they were also masked with similar, neutral images. This clearly makes a difference, because the messages I inserted don’t quite pass subliminally. Subliminal messaging isn’t as easy as I thought it would be.
Subliminal messaging is a misleading concept. Human beings always process a huge amount of sensory information subliminally. What persons perceive consciously depends largely on the focus of their attention. Subliminal sensory processing that detects a threat to a person interrupts consciousness and directs attention to the threat. Eliminating subliminal messaging cannot be done without imposing death on the recipients of concern.
Persons might reasonably be concerned about unethical attempts to influence behavior through media manipulations that cannot be readily perceived and discussed. Flashing the text “Drink Brand-X Soda” in a video in a way that cannot be consciously perceived in normal viewing is an influential example of such action. With decentralized production and sharing of video, government regulation of unethical influencing is more difficult. But if such unethical influencing became a major concern or actually significant, video players might include detectors or filters for unusual, short-duration text insertions or frame patterns.
Note:
[*] See Winkielman, Piotr, Kent C. Berridge, and Julia L. Wilbarger, “Unconscious Affective Reactions to Masked Happy Versus Angry Faces Influence Consumption Behavior and Judgments of Value,” Personality and Social Psychology Bulletin, v. 31, n. 1, January 2005, pp. 121-135.