Do Subliminal Audio Tapes Work? by Todd Stark

Do Subliminal Audio Tapes Work?^*
by Todd I. Stark

The theoretical potential for subliminal audio is not able to be established from the published scientific data at this time.
Clearly, the published empirical data show that subliminal audio influence is either extremely weak or non-existent. And most of the theoretical rationales for the effect are questionable at best. By far, the strongest effects found have been those attributed to the belief or expectation of the listener that they were being influenced.
But ultimately, it is an empirical question, not a theoretical one. The most interesting and exciting aspect of science is the discovery of new effects that were not predicted by previous models.
Unfortunately, there is very little research on the kind of masking techniques used in commercial subliminal audio tapes, perhaps a handful of studies. For lack of interesting results, the putative effect has not been of significant theoretical interest, except insofar as public concern has made it an issue.
This is in striking contrast to the large amount of data available on dichotic listening and tachistoscopic "subliminal" techniques. The contrast makes it hard to believe the claims of a conspiracy to suppress positive results.
Many experiments performed to test the effectiveness of subliminal self-help tapes have failed to show any useful effect, except that attributable to expectancy and belief. (Beyerstein,1993). Proponents of the tapes often claim that subliminal perception is a secret, that they have the only technology that works, and that their special technology has not yet been tested. (On the contrary, subliminal perception has been well studied: see Controversy)
But what of the remarkable claims for subliminal audiotapes and even backwards messages in music ?
In one of the few experiments which appeared to show an interesting subliminal audio effect, Henley and Dixon ( 1974) examined laterality differences as a contributing variable in auditory subliminal perception. Masking their subliminal messages with music, they found that subjects in the experimental groups reported cognition related to subliminal words significantly more often than control subjects did. The experimental groups also discriminated between the subliminal words and other words presented on a checklist.
The Henley and Dixon result was replicated by Mykel and Daves (1979), once with music as a mask, and once without music. The results without music were equivocal, there were more hits in the subliminal group than the control, but not when reports were grouped by "blind" judges.
This led researchers to suspect that the type of music used as the mask might be influencing responses more than the subliminal messages. Benes and colleagues (1990) discovered that using mellow masking music produced more target imagery than frenetic music, but that the imagery did not turn out to be related to the subliminal words used. That is, the control group with music but no subliminal messages produced more imagery for "water" and "family" words than did the subliminal group !
These kind of results have unfortunately left the field under-researched, providing the opportunity for advertisers to claim that the research has not disproven their claims of effectiveness.
There is far less research into acoustic masking for subliminal messages, because it has not yet been found to be a reliable way of delivering subliminal stimuli. Also, the practical difficulties of setting the appropriate loudness threshold in a generic way, and getting sufficient fidelity on mass produced audio tapes can be significant obstacles.
Finally, some proponents of audio subliminal influence claim that the effect only seems to work when the recipient in a particular state of mind. Hypnosis, for example, biases our information processing to favor non-salient cues. There is no good research at this point to support or disconfirm this claim. The visual subliminal effects do not appear to have this requirement, so it's importance is questionable. It is likely, however, that whatever is perceived is processed differently in different mindsets, and that this differs between subliminal and supraliminal stimuli.
A reliable, useful form of audio subliminal delivery may someday be available, or one may exist today and not yet be known through the published literature, but these possibilities are difficult to evaluate and seem unlikely at this point.