You are here: home / telephone experiment
Experiments on Telephone Telepathy
by Rupert Sheldrake
Many people have found that they think about someone they haven't thought of for a while, and then that person calls. Also, many people say that sometimes when the telephone starts ringing they sometimes know who is calling them, even when they had no reason to expect this person to call.
Skeptics usually dismiss these experiences by suggesting that they are a matter of chance coincidence, combined with selective memory: people only remember the times the were right and forget when they were wrong. But skeptics have no data to back up these suggestions. Surprisingly, there as been almost no research on this subject, so the question is entirely open.
I outline below a simple way of doing experiments on this phenomenon.
In the simplest version of this experiment, the subject nominates four people to whom a telepathic response seems likely. The subject gives the experimenter the names and telephone numbers of these people, and also identifies times at which all concerned will be free to take part in tests. Obviously this experiment needs to be conducted with a telephone that does not have a caller-identification display. For this reason the experiment cannot be done with a mobile phone, since all available models have a caller-ID display.
Fifteen minutes before the time chosen for a trial, the subject sits quietly reading, or engaged in some other relaxing activity (but not watching TV or a video, which can be too distracting). In videotaped trials, the subject should remain on camera for the whole of this period. For example, if the test telephone call is scheduled for 11.15 a.m., the subject sits down, within the field of view of the camera, at 11.00 a.m. Shortly afterwards, say at 11.02, the experimenter throws a die to pick one of the four potential callers, each of whom has been assigned a number from 1 to 4. (If the die shows 5 or 6, it is thrown again until 1, 2, 3 or 4 comes up.) The experimenter calls this person, and asks him or her to call the subject at 11.15. If the other potential callers have not heard from the experimenter by 11.05, they know they are not taking part in this particular trial.
At the prespecified time, the chosen caller telephones the subject, having thought about the subject for a couple of minutes beforehand. When the phone rings at 11.15 the subject knows it is one of the four potential callers, and has to guess which one it is before picking up the phone. The subject says this guess to the camera, and then picks up the phone, saying "X," or "Hi, X," or "Hello, X," before the other person says anything.
There is a 25 percent chance of being right by chance. If the subject's guesses are more than 25 percent correct, the data need to be analyzed statistically to assess their significance. The most appropriate statistical method is the binomial test. According to this test, in a series of 10 trials with four potential callers, if 6 or more are correct, the result is statistically significant at the conventional threshold level of p = 0.05, or in other words with odds against chance of 20 to one. In a series of 15 trials, scores of 9 or more correct guesses are significant. In a series of 20 trials, again 9 or more are significant. In 25 trials, 10 or more correct guesses are significant.
Another way in which this experiment can be done is to have all four callers in the same place, together with the experimenter. Then all of them can be videotaped continuously by a single camera, while the subject, in a different place, is filmed continuously on another camera with a synchronized time code. As before, the experimenter picks one of the four callers at random for each trial, and the person calls at the prespecified time. Having all the callers in a single place makes it possible to control the experiment more tightly, and to record every aspect of it on videotape.
Experiments could also be carried out with call-anticipating animals. One simple experimental design would be to videotape the animal during a series of test periods in which four different people made phone calls, in a random sequence and at randomized times. One of these people would be someone to whose calls the animal usually responded. The others would be strangers. If the animal showed its usual reactions when the person it knew rang, but ignored calls from the other three people, then this would provide strong evidence for telephone telepathy in the animal.One possible complication could be telephone calls from other people during the experimental period. That is why it is important to choose a time, or use a line, where such calls are unlikely.
A simple form for recording the results is given below
Date and time of experiment:
Right or wrong
Please send your findings to me at:
20 Willow Road, London NW3 1TJ, England
or email a summary to Pam Smart.