Reply: Neural detection of complex sound sequences or of statistical regularities in the absence of consciousness?

Sir, We read with interest the letter by Naccache and colleagues (2015) on our recent paper showing unconscious processing of global auditory regularities. We thank the Editor for the opportunity to provide further details about our results especially in the context of a growing interest around the interpretation of the global effect in disorders of consciousness (Piarulli et al. , 2015) and in sleep research (Strauss et al. , 2015). Naccache and colleagues questioned the reliability of the conclusion of our paper arguing that previous studies have never shown a global effect unless healthy controls/patients were aware of the global regularity violation. In this context, they propose a different interpretation of our study along three lines: (i) the observed effect is not ‘global’ because it does not have the characteristic features previously reported in the literature in terms of latency (>300 ms) and associated EEG component (i.e. P300) both in controls and in patients; (ii) the reported effect might be driven by violation detection between two consecutive sounds and not at the level of groups of sounds as required by a true global discrimination; and (iii) comatose patients showing the global effect were in fact conscious. In response to the first point, we should first clarify that the latencies of the observed global effect in control subjects are indeed in accordance with what was reported previously in the literature (i.e. Fig. 4 in Bekinschtein et al. , 2009) as all subjects exhibited discriminative periods after 300 ms post-stimulus onset (Fig. 1). We have never reported that the global discrimination ‘… occurred during the early time period (0–250 ms) following the onset of the irregular sound’ as claimed in the letter. In our study …