Fourth, the fact that visual N1 to words and faces was fully preserved in those schizophrenia patients without auditory hallucinations does not support the bottom-up processing deficit hypothesis in schizophrenia (e.g., Javitt, 2009), at least not without additional qualification. Instead, this finding may indicate that reductions of early perceptual ERP components for any given sample of schizophrenia patients may be primarily due to top-down cognitive interference in patients prone to auditory hallucinations. In this sense, the present findings may offer a cogent explanation for some of the inconsistency in demonstrating a neurophysiological deficit of early visual processing in schizophrenia using electrophysiologic methods. It should also be noted that our experimental procedure and stimuli mainly probed the ventral (parvocellular) visual processing stream or the ‘what' pathway (Ungerleider & Haxby, 1994), whereas Javitt (2009) reviewed evidence that reflects visual processing deficits in schizophrenia involving the dorsal (magnocellular) processing stream or the ‘where' pathway. In any case, this hypothesis regarding early processing deficits in schizophrenia and its relation to specific symptom features is in need of further study.
