David Oates, an Australian writer who spent much of the 1990s in California, claims to have discovered Reverse Speech (henceforth RS), a previously unreported human language phenomenon. He believes that as the brain is constructing and delivering the sounds of speech, two messages (normally in the same language) are communicated simultaneously: the normal forward message, which is what everyone hears and responds to consciously, and a second one in reverse, which people hear and respond to unconsciously. RS can be heard as clear, grammatical statements (usually brief) which are mixed in amongst some gibberish (though in Oates’s latest work, he suggests that there really is no gibberish, only messages which we cannot yet recognize). The reversals are accessed by recording a section of forward speech (henceforth FS) and playing the recording in reverse.
The content of reversals is nearly always related to the equivalent FS dialogue: RS often gives additional information to accentuate or strengthen the FS speech. RS also tends to reveal an individual’s unspoken thoughts, which may be in total contradiction to their conscious FS. Therefore, RS can be used as an effective tool by psychological counselors, legal professionals, parents, teachers, politicians, etc. to discover unspoken truths. However, many less transparent RS sequences involve metaphors, which require elucidation by RS analysts. According to Oates, very young children begin to produce coherent RS (in the form of reversals of babbling, etc.) well before they produce normal FS in their first language (as early as midway through their first year).
Oates’s organization offers teaching materials, courses, counseling, etc., and its practitioners give advice based on RS-without necessarily having had any other relevant training (phonetics, psychology, etc.). On the other hand, they are involved with Neuro-linguistic Programming-a prominent, recent outgrowth of Korzybski’s General Semantics, which owes only a little to linguistics proper-and with various other, more obviously New Age, ideas. All this is rather alarming! Indeed, we know of several individual cases that are very scary indeed. For instance, we heard from a man in the U.S., who reported that after attending an RS course, his wife analyzed their infant daughter’s “speech” and decided that the child was saying (backwards) that her father had sexually molested her. The mother reported her husband to the relevant child-protection agency, and when that failed, she sought custody of the child and tried to have him banned from having any access to her.
If RS really existed, the consequences for our view of human linguistic and mental activity would (as Oates himself says) be very major. However, Jane Curtain and I examined Oates’s claims and found that they were implausible and not supported by the empirical evidence. There are major methodological and theoretical problems. Notably, Oates makes a misbegotten attempt to distinguish between “genuine” RS and phonetic coincidence, the accidental occurrence of very short sequences which are (almost) the same in FS and RS (i.e., phonological palindromes, e.g., dad) or where the reversal of the FS sequence yields another equally possible sequence (so that there is a pair of corresponding forms, each of which is (approximately) the reversal of the other (e.g., say/yes). These phenomena (both types) are labeled constants, and Oates does not regard them as genuine RS. It is actually very important for him to exclude such sequences, because his theory implies that different speakers (even with the same accent) may produce different reversals of the very same utterances, depending on their often covert attitudes, etc. This is also very convenient for Oates, in that it reduces the reproducibility of his investigations! Now, Oates is not actually consistent as to which sequences do and do not count as coincidental reversals; but, more importantly, the distinction between “genuine” RS and coincidental reversals is simply incoherent.
There are several other major problems for Oates’s theory, mostly involving his lack of familiarity with linguistics. All of his criteria for identifying “good” reversals run aground on these.
As well as considering the theory of RS on these fronts, Curtain and I replicated (with refinements) Oates’s initial experiment, which he says showed that RS could be readily heard by naïve listeners. Because Oates continually prompts listeners with the RS sequences he says they should hear, we set out to establish how far the sequences can be heard without prompting. We used Oates’s own favorite examples, as they appear on his own tapes. We were able to show not only that unprompted listeners cannot generally hear the RS sequences but also that it is quite possible to induce them to hear any of a range of different sequences in the same reversed material, as long as the sounds and especially the vowels in the successive syllables are similar.
More recently, Oates has tried to rebut our criticisms, but his remarks are incoherent. There are some further tests which could be done, one of which could be decisive: if Oates is right, it should be possible to obtain otherwise unknown, specific information from RS data alone. But we believe that the basic case for RS is so weak that the onus to demonstrate a case lies with its advocates. (We have offered to advise them.)