How Ghosts Work

­Richard Wiseman of the University of Hertfordshire has researched the phenomenon of haunting in Great Britain. He has studied locations considered to be haunted, like the Haunted Gallery at Hampton Court Palace, the Edinburgh Vaults and Mary King's Close. First, he has consulted written records and interviewed employees to determine exactly where in each location people have reported ghostly activity. Then, he has asked visitors to document their experiences and report anything out of the ordinary.

His results have been pretty consistent -- people report more strange experiences in the areas where others have experienced unusual phenomena in the past. In other words, people have more ghostly experiences in the places that seem to be the most haunted. This is true regardless of whether people have any prior knowledge of the area or its ghostly history. However, people who say they believe in ghosts or who already know about supernatural activity in a particular area report strange events more often.

These findings can seem to support the idea that a building can be haunted. But Weisman's projects have also involved looking for the source of the apparently paranormal phenomena. In addition to gathering reports of strange occurrences, he has evaluated physical conditions in each haunted area. He and his research team have used instruments to measure light, humidity, sound and magnetic fields. His measurements suggest that the signs that a building is haunted often have a rational, physical cause. The Ghost Experiment site includes synopses of several of Weisman's experiments.

Other researchers have used similar methods to try to determine the causes of ghostly activity. While no one has conclusively proven that ghosts do not exist, researchers have proposed a number of alternate explanations about physical or psychological causes for strange experiences. Some are simple - people can hallucinate or mistake reflections, shadows and unidentifiable noises for ghosts. Other theories are more complex. We'll look at some examples in the next section.