NCBI ROFL spotted this 1994 scientific paper extract describing an Indian man in the UK who blamed his crimes on ghostly possession. When exorcisms failed, he was treated for paranoid schizophrenia. That apparently helped. From the abstract in the British Journal of Psychiatry:
Treatment commenced using trifluoperazine and clopenthixol. RESULTS. The patient underwent remission during neuroleptic treatment, despite previous evidence of genuine possession. CONCLUSIONS. Many cultures give rise to apparently genuine cases of ghost possession. Neuroleptics may relieve symptoms of exorcism-resistant possession.
"Exorcism-resistant ghost possession treated with clopenthixol." (Br J Psychiatry)
While searching for the whole paper, I found a great analysis of the case by Vaughan over at Mind Hacks.
So this might be an otherwise unremarkable psychiatric case if it were not for the fact that the prison chaplain, and several of the patient's cellmates, saw the spirit possess the patient as a ghostly mist. The chaplain was convinced this was a genuine case of possession, as had priests from several other faiths who had previously carried out exorcisms on the patient.
This begs the question, if the patient was treated for his belief in spirit possession and his apparent hallucinations as to the reality of the ghost, why were the chaplain and the others not considered to be ill ?
This article highlights the uncomfortable relationship between beliefs in the paranormal and the assumptions of psychiatry. The results of a recent Gallup poll suggested that over 40% of Americans believe in possession by the devil and 15% believe spirits can 'temporarily assume control of a human body'.
Although psychiatrists would argue that the content of a belief is not enough in itself to diagnose a delusion, the criteria for distinguishing between 'healthy' and 'pathological' beliefs are notoriously incoherent.