From Abracadabra to Zombies
by Dean Radin
Radin starts chapter 7, “Perception through time,” with a quote from Albert Einstein that says physicists know that the distinction between past, present, and future is “only a stubbornly persistent illusion.” Einstein apparently also once said that reality is “merely an illusion, albeit a persistent one.” So is Radin trying to get us to think that Einstein is on his side regarding psi research? It doesn’t matter what Einstein might have thought of all this. What matters is the data. Radin tells us that the data are such that “we must think of psi perception as a general ability to gain information from a distance, unbound by the usual limitations of both space and time” (p. 111). He might add a few other things that psi seems not to be bound by. For instance, psi information transfer isn’t bound by the usual process of the sender or receiver of the information knowing which particular bit of information has been sent or received. What other kind of information transfer consists of a statistical probability measured against chance odds?
Radin puts it this way: “Based on the experimental evidence, it is by no means clear that pure telepathy exists per se, nor is it certain that real-time clairvoyance exists.” The evidence, he says, “can all be accommodated by various forms of precognition.” I suppose this means that maybe there’s an Akashic record that the mind is tapping into when we think we’re reading another’s mind or seeing across space or time.
For whatever reason, Radin tells a few more anecdotes of premonitions that seemed to come true, but concludes that while “stories of accurate prophecies leave us in awe, we still need hard evidence that such things are really possible and not just fairy tales. This brings us back to the laboratory” (p. 113). The lab in this case is another meta-analysis of all published reports in English of “forced-choice” experiments on precognition conducted between 1935 and 1987 done by Charles Honorton and Diane Ferrari. They found 309 studies in 113 articles by 62 investigators: nearly 2 million trials involving more than 50,000 subjects. There was very little uniformity in these tests. They involved different kinds of guesses (ESP cards, die face, symbols, etc.), different methods of randomization, different size samples, different time intervals between guessing and unveiling, etc. Only 23 of the 62 investigators (37%) got positive results. He doesn’t tell us what percentage of the studies got positive results. Yet, Radin proudly proclaims that the odds against chance of the results were on the order of 1025 to one. “This eliminated chance as a viable explanation,” he says (p. 114). Yes, it certainly does! He also eliminates the file-drawer problem as an explanation because he used some sort of statistical formula (not revealed here) to arrive at 14,268 as the number of papers that would have to be in the drawer to tip these odds back to chance. Radin concludes from this that “the precognition effect had been successfully replicated across many different experimenters.” No other science would tolerate such reasoning.
Radin notes that Honorton and Ferrari dumped the outliers, the top and bottom 10% (highest and lowest results). They then produced odds against chance of a billion to one. Radin calls this “effectively the same” as the result for all the studies. I don’t think ten million billion billion to one is “effectively the same” as a billion to one, except in the sense that they're both absurd.
When the quality of the studies was evaluated, there was no difference in outcome between the well designed and the poorly designed studies (though the quality of the studies improved over time). Radin thinks this is because the precognitive effect was “remarkably stable” (p. 115). He says Honorton & Ferrari identified eight elements of good experiments for precognition studies, but he only mentions four: 1) specifying how many samples would be collected; 2) planning the method of statistical analysis; 3) using proper randomization methods; and 4) using automated recording. In analyzing variables, they found that 42.6% of the studies that provided trial-by-trial feedback were successful, while none of the studies that didn’t provide such feedback were successful! (If you are not using true randomization, you might be measuring guesses based on pattern recognition rather than precognition.)
Radin thinks that because certain predictions were made that seem to have been validated, it shows that “precognition performance was not merely a statistical oddity” because it seems to vary in ways that made sense “psychologically.” He thinks this suggests some sort of lawful relationship going on that can guide future research. Skeptics might be concerned that a closer look at the randomization techniques and feedback methods might reveal the basis for this apparent lawful relationship. In any case, the whole idea of concluding anything important based on a meta-analysis of this kind seems preposterous. Finally, as psychologist Jim Alcock has pointed out many times: these guys don't give the null hypothesis a chance.
At times, Radin seems to lose touch with reality, as when he posits testing for “unconscious precognition” to investigate “the possibility that the mind is in contact with its own future state.” He suggests we test whether future perceptions interfere with present performance on reaction-time tasks, but he doesn’t tell us how he’d do this. He also suggests we test whether future emotional states are detectable in present nervous system activity (p. 116). Again, he gives no hints as to what he might be talking about. Even if we found, for example, that certain nervous system activity (such as the production of adrenalin) precedes the feeling of anxiety, that would have no bearing on the precognition issue. (Maybe Radin was having a presentiment of experiments on presentiment that he would do later! In Entangled Minds [pp.166-168] he reports on four small experiments he did on presentiment. He got mixed results, but when he did a meta-analysis, guess what? He found that his results showed odds of 125,000 to 1 against chance!)
Radin then launches into a description of a bizarre experiment by Holger Klintman. Well, the experiment isn’t bizarre, but the interpretation of the results sure is. The experiment has to do with showing a subject a color patch and then showing them a color word. The test is to determine the difference in reaction time to words that refer to the color patch (red patch followed by ‘red’) vs. the reaction time to words that refer to a different color than the patch (red patch followed by the word ‘green’). The latter requires more time and apparently is very difficult to do. Klintman’s twist on the experiment was to ask the subject to name the color patch after he showed the word. He “was astonished to discover that the initial reaction times were faster when the color patch and the color name matched, and slower when the following color name mismatched” (p. 117). He rejected all “conventional explanations” in favor of “time-reversed interference.” He believed the subject’s “precognitive sensing of the future stimulus somehow traveled back in time, causing cognitive interference, when the future stimulus was a mismatch.” (p. 118) Not knowing what conventional explanations he rejected, it is hard to comment on this. But I would have looked for an explanation in terms of how different parts of the brain are involved in processing visual images and words, that associations are established between color patches and color words (a red patch would be associated with ‘red’ and so ‘green’ would not be expected to follow a red patch in the neural circuitry. Likewise, the word ‘red’ would be associated with red patch and the pathway more direct than when the word ‘green’ appears after seeing a red patch. ‘Green’ would trigger the pathway to green patch memories and delay the response to the red patch).
Radin thinks he replicated his presentiment test with an experiment he describes on pages 119 ff. I have two comments. One, the confidence interval for the experiment is 65% (i.e., seven out of 20 chance of it being a fluke). The only other comment I have to make is that where he sees a vast difference between the graph lines measuring the before physiological state (orienting response) for the calm vs. the stimulating picture, I see very little difference. This seems to be a case of pathological science: Seeing something significant in something insignificant. [Click here to see a copy of his graph.]
He concludes the chapter with a reference to an article in Science by D’Amasio et al. on people with good and bad (damaged) brains to randomly pick decks of cards that produced more favorable outcomes than the ones not picked. They called it a “bias mechanism,” but Radin wonders if it isn’t psi.
* AmeriCares *