Professor Oyebode draws attention to a number of interesting issues in response to our study of social cognition and face processing in schizophrenia. A key question raised by our study is why deficits in emotion recognition were state-dependent, being limited to individuals experiencing positive symptoms, while impairments in social cognition were stable. One possibility, as discussed by Professor Oyebode, is that those who are free of positive symptoms are able to use alternative cognitive strategies to identify basic facial emotions. This view is supported by a functional magnetic resonance imaging study in which individuals with schizophrenia, none of whom was experiencing positive symptoms, were able to identify facial emotions correctly but nevertheless showed deficits in amygdala activation when processing facial affect (Reference Gur, McGrath and ChanGur et al, 2002). These findings suggest that other brain regions compensate for the normal functions of the amygdala in facial affect processing when individuals with schizophrenia are free of positive symptoms. More difficult tests, such as our social cognition task, may prevent such compensation and thus reveal an underlying stable deficit.

Professor Oyebode also points out the apparent discrepancy between the finding that people with schizophrenia have impairments in facial recognition memory on formal testing, but are able to recognise familiar people in day-to-day life. In our study we found no deficit in the ability of those with schizophrenia to recognise the identity of novel faces presented concurrently, suggesting that the deficits seen in previous studies resulted from the mnemonic and attentional demands of the tasks used, which may be lower for familiar people.