If the truth is out there, we've not found it yet

Is the idea of a telepathic tabby completely barking? Rupert Sheldrake says his experiments prove that pets have paranormal powers, but Susan Blackmore, right, is one of several academics disputing his methods.

If Rupert Sheldrake is right about pet telepathy, then vast areas of psychology, neuroscience and medicine must be wrong.

Standard double-blind trials in medicine rely on the assumption that telepathy does not exist - that the doctor cannot know which is the real drug and which the placebo. Participants in psychology experiments are routinely assumed ignorant of the experimenter's hypothesis. Neuroscientists assume that information cannot get inside brains by telepathy.

Conceivably we should overthrow all these assumptions, but if so it is deadly serious for science. So either Sheldrake is right, and a scientific revolution of enormous proportions is upon us, or he is wrong and his results are unreliable. So which is it?

Sheldrake's most important experiments in his latest book are those with the dog Jaytee, who seems to predict when his owner, Pam, is coming home. The results are statistically impressive, but designing an experiment to test this claim is not trivial. Jaytee apparently rushes to the window when Pam is coming home, or starting the journey home, or thinking about coming home. He also runs to the window at other times. So what do you measure?

Sheldrake videotaped the dog while Pam was away, and recorded the percentage of time he spent at the window. He then compared this percentage during the period before Pam set off for home, during the first ten minutes of her journey, and during the last ten minutes before she arrived. He found that Jaytee was more often at the window when she was coming home.

Fine so far - but in many experiments Pam herself decided when to come home. Although she may have thought her behaviour unpredictable, it is possible that Jaytee detected patterns in her journeys and reacted accordingly.

Realising this problem, Sheldrake did 12 experiments in which he bleeped Pam at random times to tell her to return. Now surely Jaytee could not be using normal powers could he? No. But there is another simple problem. When Pam first leaves, Jaytee settles down and does not bother to go to the window. The longer she is away, the more often he goes to look. But in these experiments Pam was never away for less than an hour, and yet the comparison is made with the early period when the dog rarely gets up. Sheldrake provides extra analyses to try to get round this artefact, but it remains a problem for his experimental design.

Richard Wiseman, sceptical parapsychologist at the University of Hertfordshire, carried out four experiments of his own with Jaytee, and concluded that the dog had no psychic powers. Although Sheldrake writes of these experiments in his book, they are mentioned only in a footnote.

Finally, Sheldrake claims that most psychic researchers confine their studies to humans, and there is a "taboo" against paranormal research. In reality there are many well known animal studies, including the famous rat precognition studies that J. B. Rhine, the founder of parapsychology, exposed as a fraud in 1974. As for other scientists, there are many who, like myself, know that if telepathy exists it could be the most important research we could ever do. So we have tried and failed. There are better ways to spend precious research time than chasing after something that lots of people want to be true, but almost certainly is not.

Susan Blackmore is reader in psychology at the University of the West of England, Bristol.