Influence of phonological expectations during a phoneme deletion task: Evidence from event‐related brain potentials

Abstract

Several studies have identified a negativity [the phonological mismatch negativity (PMN)] preceding the N400 during auditory sentence comprehension. The present study investigated whether the PMN reflects a prelexical or lexical stage of spoken word recognition. Event-related brain potentials (ERPs) were recorded to investigate phonological processing independently from lexical/semantic influences during a task requiring metalinguistic analysis of speech stimuli. Participants were instructed to omit the initial phoneme from a word ("clap" without the/k/) after which they heard a correct (lap) or incorrect (cap, ap, nose) answer. The PMN (peaking at 270 ms) was largest to incorrect items and did not differentiate between items that shared the same rime and items that were phonologically unrelated to the correct choice. Further, the PMN did not differ between word (cap) and nonword (ap) choices. The P300 was largest to correct items but was also seen to choices that rhymed with the correct answer. It is concluded that the PMN serves as a neural marker for the analysis of acoustic input merging with prelexical phonemic expectations.

Activation in the anterior left auditory cortex associated with phonological analysis of speech input: localization of the phonological mismatch negativity response with MEG

“…When the sentence-ending words were both phonologically and semantically unexpected, both PMN and N400 responses were elicited, and, in the opposite case, neither the PMN nor N400 responses were obtained. In agreement with this study, a PMN response, independent of the N400, has been obtained whenever a violation against the expected auditory phonemic features has been introduced [5,6,8,22,24,31].…”

Phonological oddballs in the focus of attention elicit a normal P3b in dyslexic adults

“…Phonemes /b/ and /g/ are therefore less perceptually salient than /p/ and /r/ and may require more processing. This effect possibly relates to the P2 modulation reported by Newman et al [34] who engaged participants in a phoneme deletion task. The task was to decide whether the second word of a pair (e.g., dlapT) was the first word (e.g., dclapT) devoid of its first phoneme.…”

“…The P2 has been shown to be modulated by short-term memory demands [9] and is also suggested to vary with acoustic differences between phonemes [34]. In this study, the difference in P2 is likely to represent the greater processing demands of /b/ and /g/ phonemes which have minimal perceptual cues for identification [30].…”

“…The task was to decide whether the second word of a pair (e.g., dlapT) was the first word (e.g., dclapT) devoid of its first phoneme. Newman et al [34] observed that the P2 was significantly larger for targets (correct phoneme deletion) than foils (irrelevant words). They suggested that this modulation reflected changes in the acoustic features of the different initial phonemes presented across conditions.…”

“…Studies of classical and dnoveltyT P300s have led to the differentiation of two subcomponents within the P3 wave: (a) the P3a more visible over fronto-central electrodes and thought to index automatic shifts of attention [17]; and (b) the P3b more visible over centroparietal electrodes and thought to index target detection and working memory updating [40]. In the auditory modality, P3a/P3b complexes have been studied using pure tone oddballs (e.g., [13]), phonological oddballs (e.g., [16,34]), and lexical oddballs (e.g., [39]). Because the P300 indexes awareness of stimulus change, it has been studied in dyslexic individuals in an attempt to characterize potential attentional deficits.…”