Where do ghosts come from?

AS MEDIEVAL CASTLE bedrooms go, this one looks the part. Disturbing Flemish tapestries share the walls with stern portraits. On close inspection, the ornate fireplace's iron firedogs turn out to have devils' heads. This place is supposedly haunted by the ghost of Tom Skelton, a 16th-century jester said to have committed murder. The malevolent face of "Tom Fool" stares from a dimly lit oil painting just outside the bedroom.

My assignment is to stay overnight in the Tapestry Room at Muncaster Castle in Cumbria, UK. Having earlier reassured my editor that I laugh at ghost stories, my bravado is crumbling. I still don't believe in ghosts, but I'm scared the atmosphere will wind me up into a panic. Two previous guests have bolted in the night, one a premiership footballer, the other a diehard sceptic who came to scoff. Then I learn that I will not be able to leave the room without tripping the castle's burglar alarms. What have I let myself in for?

I am here because of a controversial theory that some reports of ghosts could be caused by unusual magnetic fields triggering strange reactions in the brain. There's a long tradition of hunting for such fields at supposedly haunted locations - and even of trying to produce them in the lab. So far, results have been mixed, so I have followed neuroscientists and psychologists to Muncaster Castle to see if, in this case at least, science can lay a ghostly mystery to rest.

Chief investigator Jason Braithwaite is a cognitive psychologist at the University of Birmingham, UK. Braithwaite is a sceptic with a long-standing interest in the psychology of paranormal experiences and beliefs. "These weird experiences appear to be part of the normal operation of the brain," says Braithwaite. "No model of brain function can be viewed as complete until it explains them."

It was in the 1970s that Michael Persinger, a neuroscientist from Laurentian University in Sudbury, Ontario, Canada, proposed that some hallucinations could be triggered by magnetic fields. It is well established that magnetic pulses of 1 or 2 teslas can stimulate neurons in the brain; it is sometimes used to treat depression. Persinger, however, was interested in much weaker fields, of about 1 to 10 microtesla, which can arise from electrical equipment such as a hairdryer, or simply exist in natural background fields.