Journal of the Society for Psychical Research 63, October 1999
by Rupert Sheldrake

In August 1998 the British Journal of Psychology published a paper entitled 'Can animals detect when their owners are returning home? An experimental test of the "psychic pet" phenomenon' by Richard Wiseman, Matthew Smith and Julie Milton. This paper was widely publicized thanks to a media release issued by the Press Office of the British Psychological Society. The sceptical tone of this announcement, entitled 'Mystic dog fails to give scientists a lead', was reflected in the ensuing newspaper reports: 'Pets have no sixth sense, say scientists' (The Independent , Aug 21); '"Psychic" dog is no more than a chancer' (The Times , Aug 21); 'Psychic pets are exposed as a myth' (The Daily Telegraph , Aug 22). The wire services reported the story world-wide.

Together with Pamela Smart, I have carried out over 200 experiments with the dog in question, called Jaytee. The four experiments that form the basis of the paper by Wiseman and his colleagues were carried out at my invitation (and with the loan of my video equipment). I would like to take this opportunity of putting into context their paper and the publicity it excited.

Many dog owners claim that their animals seem to know when a member of the household is coming home, showing their anticipation by waiting at a door or window. Some dogs are said to react more than 10 minutes in advance (Sheldrake & Smart, 1997; Brown & Sheldrake, 1998; Sheldrake, Lawlor & Turney, 1998). In 1994 I proposed a simple experiment to investigate this claim (Sheldrake, 1994): the owner should come home at a non-routine time in an unfamiliar vehicle, such as a taxi, and the people at home should not know when the person is coming. Does the dog still show its usual signs of anticipation? Or is this phenomenon merely a matter of routine, selective memory, subtle cues from people at home, or hearing a familiar vehicle approaching?

In an initial series of 100 observations on Jaytee's behaviour during the absences of his owner, Pamela Smart (PS), on most occasions he began waiting around the time that PS set off to come home, at whatever time of day, or by whatever means of transport. He did so when PS returned at randomly chosen times unknown to the people at home with the dog (Sheldrake & Smart, 1998). However on 14 out of 100 occasions he did not show this anticipatory behaviour. On some of these occasions he appeared to be distracted by a bitch on heat in a neighbouring flat, on others he was sick, but sometimes he failed to wait at the window for no apparent reason. Nevertheless, taking all the data into account there was a highly significant relationship (p<0.0001) between the time that the dog began to wait at the window and the time his owner set off to come home (Sheldrake & Smart, 1998).

In November 1994, the Science Unit of Austrian State Television (ORF) filmed an experiment in which PS went out, leaving Jaytee in her parents' flat. PS was then asked to come home at a randomly-selected time. Her parents and the cameraman filming the dog in their flat did not know when she would be returning. She travelled by taxi. In this experiment, Jaytee got up and went to wait at the window 11 seconds after PS was told to go home, and he stayed there throughout her 15-minute homeward journey.

Richard Wiseman, well known in the media for his sceptical views, was asked by various newspapers and television programmes for his comments on Jaytee's anticipatory behaviour. He suggested a number of possible explanations, such as routine times of return and selective memory, which I had already tested and eliminated. I invited him to do his own tests, and PS and her family kindly agreed to help him.

Wiseman did four experiments with the help of his assistant Matthew Smith. In these experiments the behaviour of the dog was videotaped continuously by Wiseman throughout PS's absence, with the timecode recorded on the film. Meanwhile, PS travelled with Smith to locations 5.5 to 8 miles away, and later set off home with Smith at randomly-selected times known in advance to Smith but not to herself. They returned in cars with which Jaytee was unfamiliar, to avoid any possible sensory clues from familiar vehicles. Wiseman and PS's parents did not know when PS would be returning.

As in many of my own experiments, in the 3 tests conducted at the home of PS's parents, the dog sometimes went to the window when PS was not returning, for example to watch other dogs walking past or cars pulling up outside. But he waited there much more when she was on her way home (Figure 1). My own videotaped experiments show a very similar pattern (Sheldrake, 1999).

In these graphs the time over which the experiment took place is divided into 10-minute periods. The final 10-minute period on each graph represents the first 10 minutes of PS's return journey, while she was still too far away from home for Jaytee to detect the approach of the car. This car was in any case unknown to him.

Wiseman, Smith and Milton's criterion for Jaytee's success or failure
Instead of plotting their data on graphs and looking at the overall pattern, Wiseman, Smith and Milton used a criterion of their own invention to judge Jaytee's 'success' or 'failure'. They did not discuss this criterion with me, but based it on oversimplified remarks about Jaytee's behaviour made by commentators on two British television programmes which re-broadcast an extract from the ORF experiment (Wiseman, Smith & Milton, 1998). These television programmes stated that Jaytee went to the window every time that his owner was coming home. In fact, he did so on 86 per cent of the occasions (Sheldrake & Smart, 1998). And on one of these programmes it was said that Jaytee went to the window "when his owner Pam Smart starts her journey home." In fact Jaytee often went to the window a few minutes before PS started her journey, while she was preparing to set off (Sheldrake & Smart, 1998).

Wiseman et al. took Jaytee's 'signal' for his owner's return to be the dog's first visit to the window for no apparent external reason. In the first experiment this 'signal' lasted for 53 seconds. In their paper, they remark that when viewing this videotape,

PS correctly remarked that Jaytee only stayed at the porch for a fairly brief period of time.... and that a better indicator of his signal might be when he remained there for a longer period of time. For this reason the authors decided that any future study should not take the first time that he inexplicably went to the porch as his 'signal' but, instead, the first time that he inexplicably visited the porch for more than 2 minutes. [their italics]

The following experiments were judged according to this new criterion. Wiseman et al. found that Jaytee reacted according to their arbitrary criteria in all three experiments at PS's parents flat before PS set off at the randomly-selected time. These experiments were therefore classified as failures. They ignored the dog's behaviour after the 'signal' had been given.

In addition to these experiments at PS's parents' flat, they carried out a test at the house of PS's sister, where Jaytee had to balance on the back of a sofa to look out of the window. In the Wiseman-Smith experiment in this house, the first time he visited the window for no apparent reason coincided exactly with PS setting off, and her sister remarked at the time, on camera, that this was how Jaytee behaved when PS was coming home. But Jaytee did not stay there for long because he was sick; he left the window and vomited. Because he did not meet the two-minute criterion, this experiment was also classified as a failure.

The negative conclusions of Wiseman, Smith and Milton are unjustified for at least four reasons:

  1. When PS was on her way home, Jaytee's characteristic behaviour, as noted by the Smart family and documented in my own experiments, was his waiting by the window. This pattern can be seen by looking at all the data (as in Figure 1) rather than by confining attention to an arbitrarily brief period. Wiseman et al. correctly point out that pet owners could draw false conclusions about an animal's 'signal' if they considered only one of the possible signals and ignored others; they rightly emphasize the need for 'a complete and accurate recording of the pet's behaviour.' But having made such a record, they then disregard it in favour of a single 'signal' - a 'signal' defined by themselves on the basis of a remark on a television programme.

  2. Jaytee's 'signal' for PS's return was supposed to be a visit to the window for an inexplicable reason, i.e. for a reason that was not apparent on the videotape itself. Nevertheless, Jaytee's attention might have been attracted to incidents outdoors that were not visible in the narrow field of view of the camera, and hence an explicable visit could be wrongly classified as 'inexplicable'.

  3. In so far as the purpose of these experiments was test the possibility that Jaytee was capable of responding telepathically to PS's intentions to come home, the experimental design involved a serious flaw. Wiseman et al. assumed that PS, waiting with Smith in a pub or other location, would not think about returning until Smith told her to. In fact, she tells me that as time went on she could not help thinking about going home, with thoughts like "It won't be long now". In all three experiments conducted in PS's parents flat, the randomly-selected return times were in the second half of the experimental period, and PS knew that they would have to be leaving by the end of this period. Moreover, while they were waiting together, Smith knew when they would be setting off, and he could well have communicated his anticipation to PS unconsciously through subtle cues, such as glances at his watch. Hence the tendency of Jaytee to go to the window before PS set off (Figure 1) could have reflected her own anticipation that she would soon be leaving.

    Milton Pet Graph

    Figure 1. Results of the three experiments carried out by Wiseman and Smith with Jaytee at PS's parents' flat on June 12, 1995 (A); June 13, 1995 (B); and December 4, 1995 (C). The graphs show the total amount of time that the dog spent at the window in successive 10-minute periods. The 10-minute periods were defined in relation to the randomly-selected time at which PS was told to return home. The final point on each graph represents the first 10 minutes of PS's return journey, and is indicated by a filled circle ( -o- ). PS's return journeys took 20 minutes, 11 minutes and 13 minutes in experiments A, B and C respectively. Any visits to the window that Jaytee may have made in response to the car's approach and PS's arrival have therefore been excluded. The time in each experiment when the dog gave its 'signal', as defined by Wiseman, Smith and Milton, is indicated by an arrow.

  4. The data of Wiseman et al. show the same pattern of response as my own (Sheldrake, 1999) and hence confirm rather than refute my findings about Jaytee's anticipatory behaviour.

Publicizing a sceptical claim

In the summer of 1996, while I was continuing a long series of videotaped experiments with Jaytee, Richard Wiseman went to several international conferences, including the World Skeptics Congress, and gave lectures about the 4 experiments he and Smith had done with this dog, stating that the dog had failed their tests. For one of these conferences they wrote a paper (Wiseman & Smith, 1996) very similar to the paper published in the British Journal of Psychology (with the addition of Milton as an author), and sent me a copy.

In September 1996, Wiseman and I met to discuss these findings. He raised objections to the way I had plotted the data on graphs, and suggested an improved method, dividing up the experimental period into 10-minutes intervals. The graphs shown here (Figure 1) use this method suggested by Wiseman. I sent copies of these graphs to him before he and his co-authors submitted their paper to the British Journal of Psychology and suggested that they draw attention to the fact that the dog spent most time at the window while PS was actually on her way home. But they did not mention this striking effect either in their paper or when they publicized their sceptical conclusions.

Over the next two years, Wiseman announced repeatedly through the media that he had discredited this dog's ability to anticipate his owner's return. For example, on a British television programme called Strange But True (ITV; 1 November, 1996) he said of Jaytee: 'In one out of four experiments he responded at the correct time - not a very impressive hit rate and it could just be coincidence'. The three 'misses' are the experiments summarized in Figure 1.

On another British television programme called Secrets of the Psychics (Equinox, Channel 4; 24 August, 1997) he and several fellow sceptics debunked a series of bogus seances and fraudulent healers and in this context he said of Jaytee:

We filmed him continuously over a three hour period and at one point we had the owner randomly think about returning home from a remote location and yes, indeed, Jaytee was at the window at that point. What our videotape showed, though, was that Jaytee was visiting the window about once every 10 minutes and so under those conditions it is not surprising he was there when his owner was thinking of returning home.

In order to support this statement, a series of video clips showed Jaytee going to the window over and over again, eight times in all. The times of these visits to the window can be read from the timecode. They were taken from the experiment on shown in Figure 1A. Two of these eight visits were the same visit shown twice, and three took place while PS was on the way home, although they were misleadingly portrayed as random events unrelated to her return.

The British Psychological Society's Press Office helped Wiseman's sceptical claims to reach a yet wider audience, and a quote from Wiseman included in their media release was widely reported in the press: 'A lot of people think their pet might have psychic abilities but when we put it to the test, what's going on is normal not paranormal.' Smith was quoted as saying: 'We tried the best we could to capture the ability and we didn't find any evidence to support it' (Irwin, 1998).

Wiseman, Smith and Milton have succeeded in proving that the media can be misleading. They have also shown that the claims of sceptics need to be treated with scepticism. But in spite of their polemical intentions, their data support rather than refute the idea that some dogs anticipate their owner's returns, whatever the explanation for this ability may turn out to be.

Brown, D.J. and Sheldrake, R. (1998) Perceptive pets: a survey in north-west California. JSPR 62, 396-406..
Irwin, A. (1998) Psychic pets are exposed as a myth. Daily Telegraph, 22 August.
Sheldrake, R. (1994) Seven Experiments that Could Change the World. London: Fourth Estate.
Sheldrake, R., Lawlor, C. & Turney, J. (1998) Perceptive pets: a survey in London. Biology Forum 91, 57-74.
Sheldrake, R. and Smart, P. (1997) Psychic pets: a survey in north-west England. JSPR 61, 353-364.
Sheldrake, R. & Smart, P. (1998) A dog that seems to know when his owner is returning: preliminary investigations. JSPR 62, 220-232.
Sheldrake, R. (1999) Dogs that Know When Their Owners are Coming Home. London: Hutchinson (in the press).
Wiseman, R. & Smith, M. (1996) Can pets detect when their owners are returning home? An experimental test of the 'psychic pet' phenomenon. Proceedings of the 39th Parapsychological Association Convention , pp. 35-44.
Wiseman, R., Smith, M. & Milton, J. (1998) Can animals detect when their owners are returning home? An experimental test of the 'psychic pet' phenomenon. British Journal of Psychology 89,