The study examines the relationship between hallucinations/delusions a
nd violent behaviour in a sample of long-stay inpatients with chronic
schizophrenia. Thirty-one subjects defined as violent and meeting DSM-
111-R criteria for schizophrenia were compared with 31 matched non-vio
lent schizophrenia patients with respect to detailed phenomenologies o
f auditory hallucinations using the Mental Health Research Institute U
nusual Perceptions Schedule (Carter and Copolov, 1993; Carter et al.,
1995) and delusions using the Maudsley Assessment of Delusions Schedul
e (Taylor et al., 1994). Patients in the violent group were significan
tly more likely to experience negative emotions, tone and content rela
ted to their voices than those in the non-violent group, whilst patien
ts in the non-violent group were more likely to experience positive em
otions, tone and content related to their voices, Patients in the non-
violent group were significantly more likely to report success in copi
ng with their voices. There was no association between command halluci
nations and violent behaviour. Patients in the violent group were more
likely to hold persecutory delusional beliefs than those in the non-v
iolent group, while patients in the non-violent group were more likely
to hold grandiose delusions than those in the violent group. Patients
in the violent group were also more likely to report that the delusio
n made them feel angry, while those in the non-violent group were more
likely to report that the delusion made them feel elated. The results
suggest specific aspects of the phenomenologies of hallucinations and
delusions that should be clinically assessed to determine the likelih
ood of violence as a result of such psychotic symptoms. (C) 1997 Elsev
ier Science B.V.