Very recent publicity includes popular magazine articles mixing correct information on subliminal perception with wide-ranging statements regarding the ubiquitousness of subliminal advertising (Jonsson, 1998). Even some journals targeting scientists publish articles with similar content (Frydman, 1996). A much-publicized criminal case in the US involved the parents of two young men who attempted suicide, one of them successfully, suing the rock group Judas Priest for having included in a song, at a low, subliminal level, the words "Do it!".

 Initial reports that movie-goers increased their consumption of pop corn and Coke, when subliminal verbal messages encouraged them to buy these products, were eventually exposed as a hoax. In the suicide case, the jury couldn't agree on whether the alleged message was actually there, except by some accidental mixture of sounds (Jonsson, 1998). The judge, however, eventually concluded that the words could be heard, but that there was no evidence that they would have been capable of triggering someone's suicide without the aid of the suicidal ruminations that the young man had previously engaged in (Loftus and Klinger, 1992). Subliminal "self-help" tapes were investigated by Greenwald (1991) who found them to be without effect. 

 In a scholarly review, Bornstein (1989) addressed the reality of claims that subliminal  techniques can be used as tools in advertising and propaganda. In brief summary, there is a difference between stimuli that are subliminal in the sense of not being detectable, and stimuli that are sometimes erroneously called "subliminal", in the sense that they are not logical and obvious (e.g. a strongly positive affective tone in a commercial that sells a product that in and of itself has very little emotional value). The latter is something that can easily be discovered by the audience, if educated and encountering the message in a situation where sufficient time and attention can be allocated to the analytic task. Truly subliminal stimuli, however, cannot be subjected to conscious analysis and control, and therefore remain a more provocative  possibility for mass influence.

 Nevertheless, the studies reviewed by Bornstein show that while some subliminal techniques can theoretically be used for such purposes, the logistic difficulties in translating complicated laboratory procedures into something that could work by way of the mass media are immense. Most importantly, the extremely variable, uncontrolled conditions under which people might encounter such attempts to influence them subliminally makes it highly questionable that effects would ever be greater, or more cost-effective, than that achieved by accepted, well-known means of advertising and propaganda. In that sense then, mass influence via subliminal procedures is an issue for scientific investigation (cf Underwood, 1994), but not for commercial or political exploitation.

 Regardless of reviews such as Bornstein's, it is unlikely that subliminal research will cease anytime soon to be met with fascination and fear. Our answer to those asking why we address such a possibly dangerous subject has always been very simple. If subliminal influence, propaganda and commercials do not work, real scientists doing real science can help expose the charlatans and con-men who make money saying that it does. 
If on the other hand it does work, only publicly available data from scientists working in an open society can provide the means of dealing with such a situation. Further, only scientists working in an open society are motivated to investigate whether a potentially destructive phenomenon could, under some circumstances, be used for constructive, positive purposes, e.g. in teaching and therapeutics. At the present state of knowledge then, the truly unethical position of the scientific community would be not to research subliminal phenomena. In pursuing this line of work we accept our role as ethically responsible scientists in psychology.

 As a first attempt at dealing with these issues, we have performed two experiments testing whether debriefing does what it is intended to do, namely "normalize" participants. The question was: does the information to participants given at the end of an experiment help remove or prevent any lasting effects of the subliminal stimulation? The intriguing results implied that some small amount of information was better at preventing long-term effects than detailed, in-depth information about the stimulus and its effects. This is quite surprising and unexpected when compared with other studies testing the effects of debriefing after entirely conscious stimulation (e.g. Ross, Lepper & Hubbard, 1975). The abstract to the article can be found here.

 Of course there was no subliminal message saying EAT POP CORN and DRINK COKE flashing in the background as this page was loading on your computer screen. If you are a person who believes that subliminal commercials work, and also believed that we did flash you with such messages now, your behavior may nevertheless be affected by that very belief. Beliefs are among the most powerful forces discovered by psychology.

Bornstein, R. F. (1989). Subliminal Techniques as Propaganda Tools: Review and Critique. The Journal of Mind and Behavior, 10(3), 231-262.

Frydman, M. (1996). Subliminal Manipulation of Smoking. Journal of Environmental Pathology, Toxicology and Oncology, 15(2-4), 173-176. 

Greenwald, A. G., Spangenberg, E. R., Pratkanis, A. R., & Eskenazi, J. (1991). Double Blind Tests of Subliminal Self-help Audiotapes. Psychological Science, 2, 119-122. 

Jonsson, T. (1998). You Can Never Escape. Darling: Det Paranoida Numret, 1/1998, 42-46. 
(In Swedish). 

Loftus, E. F., & Klinger, M. R. (1992). Is the Unconscious Smart or Dumb?
American Psychologist, 47(6), 761-765. 

Ross, L., Lepper, M.R. & Hubbard, M. (1975). Perseverance in self-perception and social perception: Biased attributional processes in the debriefing paradigm. Journal of Personality and Social Psychology, 32(5), 880-892.

Underwood, G. (1994). Subliminal Perception on TV. Nature, 370(14 July), 103.