Third, the intact P1 source in hallucinators for unambiguous, high-contrast and meaningful visual stimuli (i.e., words and faces) used in the recognition memory and working memory paradigms, which differ in these regards from more ambiguous stimuli used for visuospatial discrimination tasks (e.g., Butler et al., 2007; Doniger et al., 2002; Foxe et al., 2001), suggests that the N1 sink reduction did not result from a deficit in earlier visual processing. Rather, it involves interference during a subsequent stage of cognitive processing beginning about 150 ms after word or face onset, likely coinciding with stimulus categorization (Rossion & Jacques, 2008) as opposed to low-level perceptual analysis (Butler et al., 2007). This implies that auditory hallucinations, or the impaired functional processes that cause auditory hallucinations, interfere with the integration of incoming perceptual information as reflected by N1, that is, perceptual driven or bottom-up processing involving the ventral occipito-temporal stream is impaired by top-down modulation or cognitive interference (cf. Twomey, Kawabata Duncan, Price, & Devlin, 2011).
