On the 28th May 1960, at precisely 7:40pm, AD Cornell valiantly attempted to 'haunt' a cinema audience who were sitting down to enjoy an X-rated film. Before emerging from the shadows, Cornell draped himself in a white muslin sheet, the fabric covering him from head to toe. He then emerged before the unsuspecting audience and was bathed in the light of projector. He moved in front of the screen, from the left edge to the right edge and back again. For Science!

Cornell, a Cambridge-based parapsychologist, was conducting experiments in "apparitional experiences". In his reports, he tacitly accepts the contentious premise that the spirits of the dead may literally walk among us as physical apparitions. His experiment did not directly address this question. Rather, he tried to artificially induce the experience of seeing 'real' ghosts, so that he could observe how people reacted to this unexpected and apparently paranormal experience.

Cornell's reports have previously been aired in Mary Roach's Spook and Richard Wiseman's Paranormality, and as an experimental psychologist studying human attention, I found them intriguing. My own research into how people perceive illusions has led me into rather unusual historical rabbit holes related to magicians, psychologists, witches, and ghost hunters. But Cornell's simulated haunting initially seemed to be unlike anything I'd previously encountered. His methods were… unorthodox and his conclusions were dubious. However, by the time I finished, I had come to the delightfully absurd realisation that his designs had (inadvertently) anticipated developments in cognitive psychology by nearly half a century.

Before getting back to that cinema, let's consider Cornell's two preceding experiments. Initially, he attempted to haunt a cow pasture near King's College. A four-and-a-half minute 'Ghost Walk' was carefully choreographed. He would crawl behind a small mound, wrap himself in a white sheet, and abruptly stand up – as if he were appearing from nowhere. He would then stroll to the next mound before vanishing by diving to the ground. Dress-rehearsals seemed to confirm that the appearances and disappearances were sudden and startling. He stationed research assistants along the path to observe the unsuspecting by-passers.

Cornell noted that the local cows appeared to regard him with rapt attention

The stage was set, and the team conducted six walks in total, but there was an unexpected complication. To Cornell's dismay, none of the approximately 80 by-passers gave any indication of having noticed anything at all! Cornell did note that the local cows appeared to regard him with rapt attention, perhaps supporting his contention that his bed sheeted-self was clearly visible and in plain sight.

Reasoning that a more ghost-appropriate venue might lead to superior results, Cornell staged his next walks in the graveyard of St Peter's Church. Because the graveyard abutted a public road, the team took extra safety precautions. This time, Cornell's assistants were not merely standing by to observe; they were ready to step in to "avert any accident", resulting from hysteria at the sight of the 'Experimental Apparition' (which is how Cornell consistently refers to his bed sheet-shrouded self).

One passerby assumed the bed-sheeted Cornell was 'an art student walking about in a blanket'

Such precautions proved unnecessary. Between pedestrians, cyclists, and motorists, Cornell estimated that approximately 142 people would have been able to witness the apparition, but only four gave clear indications of having seen it. Under questioning, it emerged that none of them believed they had witnessed anything remotely paranormal. The first person described the apparition as "a man dressed as a woman, who surely must be mad" another assumed that it was "an art student walking about in a blanket". Two witnesses, when questioned together, did realise that the Experimental Apparition was probably intended to simulate a paranormal event, but went on to note that the effect was spoiled because "we could see his legs and feet and knew it was a man dressed in some white garment".

My favorite part of this write-up comes when Cornell reports an impromptu hypothesis. He wondered if the lack of awareness might be attributed to the "purely visual nature" of the apparition. In his own words: "During the last 6 minutes, the figure uttered low moans in order to attract attention." I feel that it's important to re-emphasise here that Cornell is referring to himself: a man draped in a sheet, standing in a graveyard at night, moaning. None of the 10 people he came across gave any indication of noticing him, although, in context, it's also plausible that they were deliberately avoiding eye-contact.

Which brings us back to the cinema, which Cornell argued was clearly the logical next step after the cow pasture and the graveyard. Still determined to induce a paranormal experience, he wrote that his next experiment needed to take into account two important considerations. His first consideration was a practical one: He wanted to present the Experimental Apparition in a place where the spectators would be focusing their attention. His second consideration was an ethical one: Cornell wrote that he had grown concerned about the potential for the Experimental Apparition to traumatise children. He specifically selected an X-rated film to 'safeguard' against this risk.

This one made less sense to me, especially considering preceding ghost walks. That being said, I do appreciate the sublime ridiculousness of the idea. He made arrangements to conduct his test during the trailers preceding the feature. The Experimental Apparition was clearly visible for approximately 50 seconds as it passed in front of the screen. Six research assistants, distributed about the audience, monitored the proceedings. Before the feature film started, Cornell without bed sheet, re-emerged with a microphone to survey the crowd.

One respondent thought they'd seen a polar bear

For Cornell, the experiment was another failure. None of the audience reported anything remotely paranormal. Many saw nothing unusual at all: 46% of the respondents had failed to notice the Experimental Apparition when Cornell first passed in front of the screen, and 32% remained completely unaware of it. Even the projectionist, whose job was to watch for anything unusual, reported that he had completely failed to notice the apparition. Those that did see 'something' were not particularly accurate in their descriptions. One person reported seeing a woman in a coat, another thought they had seen a polar bear, and another believed that they had observed a fault in the projector. Only one person accurately described a man dressed in a sheet pretending to be a ghost. Cornell ultimately concluded that such failures to see should be attributed to an absence of a "subtle psi-factor" or "telepathic stimulation" that would normally correspond with a 'genuine' haunting. He goes so far as to suggest that the number of 'true' hauntings may actually be grossly under-reported.

For me, these failures to see are by far the most exciting part of the experimental series. The pleasure of reading Cornell's original reports, which were published in 1959 and 1960 in the Journal of the Society for Psychical Research, is that he writes in matter-of-fact academic prose. He dutifully reports numbers and exact quotes from participants, and walks the reader through the details of his experimental designs without a glimmer of apparent irony. To him, the cinema audience and the X-rated film simply represented an elegant solution to a methodological problem.

While Cornell's psi-based conclusions would be largely out of place in a contemporary scientific journal, his results nevertheless pre-empt a modern understanding of the mind that has nothing to do with the paranormal.

To understand why, consider the following experiment. Participants shown a video of two groups of people, one group wears white shirts and the other wears black shirts. Each team has a basketball that they pass amongst themselves. The viewer is instructed to count the number of times the players in white pass the ball.  If you like, you can try it for yourself here.

The lurking gorilla

Did you spot anything strange? Surprisingly, many people fail to notice the gorilla – a phenomenon known as "inattentional blindness", which was a term first coined in 1998, by psychologists Irving Rock and Arian Mack, who showed that when viewers are concentrating on a task, they could be functionally blind to unexpected objects and events.

Mack and Rock's experiments largely involved presenting simple geometric shapes on computer monitors but Daniel Simons and Christopher Chabris extended the concept with the gorilla experiment. As you can see in the image below, the gorilla was in plain view – yet it escaped many people's notice.

Since then, attention research has exploded with reports of 'sights unseen,' ranging from motorcycles, simulated assaults, pirates, and unicycling clowns; doctors even fail to notice gorillas lurking in radiological scans.

Cornell's paradigm has several parallels with the basketball game: watching the movie trailer is somewhat like the counting task and the Experimental Apparition serves as an effective gorilla – meaning his results are arguably one of the first demonstrations of inattentional blindness.

Even though inattentional blindness is now an established phenomenon with the scientific community, in general, everyday people are not necessarily aware of it. Contemporary surveys have shown that most people firmly believe that they would notice unexpected objects and events, even if they were paying attention to something else.

Continued research into inattentional blindness can potentially have important implications for real-world domains like security, advertising, traffic safety, and eye-witness testimony. But, looking back, it's interesting that Cornell did (unintentionally) manage to design and conduct an experiment so excellently suited for demonstrating inattentional blindness. Appropriately, the most valuable element of the experiment was effectively hidden in plain sight.   

Matthew Tompkins is an experimental psychologist at the University of Oxford. He tweets as @MattLTompkins.

Join 700,000+ Future fans by liking us on Facebook, or follow us on Twitter, Google+, LinkedIn and Instagram.

If you liked this story, sign up for the weekly bbc.com features newsletter, called "If You Only Read 6 Things This Week". A handpicked selection of stories from BBC Future, Earth, Culture, Capital, Travel and Autos, delivered to your inbox every Friday. 


Cornell, A. D. (1959). An experiment in apparitional observation and findings Journal of the Society for Psychical Research, 40 (701), 120-124.

Cornell, A. D. (1960). Further experiments in apparitional observation Journal of the Society for Psychical Research, 40 (706), 409-418.