Predictive coding of phonological rules in auditory cortex: A mismatch negativity study

Abstract

The brain is constantly generating predictions of future sensory input to enable efficient adaptation. In the auditory domain, this applies also to the processing of speech. Here we aimed to determine whether the brain predicts the following segments of speech input on the basis of language-specific phonological rules that concern non-adjacent phonemes. Auditory event-related potentials (ERP) were recorded in a mismatch negativity (MMN) paradigm, where the Finnish vowel harmony, determined by the first syllables of pseudowords, either constrained or did not constrain the phonological composition of pseudoword endings. The phonological rule of vowel harmony was expected to create predictions about phonologically legal pseudoword endings. Results showed that MMN responses were larger for phonologically illegal than legal pseudowords, and P3a was elicited only for illegal pseudowords. This supports the hypothesis that speech input is evaluated against context-dependent phonological predictions that facilitate speech processing.

Predictive coding accelerates word recognition and learning in the early stages of language development

“…Note that the major source of co-articulation, the consonant /k:/, was the same for all stimuli. Acoustical measurements conducted in a previous study (Ylinen et al, 2016) suggested that in similar bisyllabic Finnish stimuli, first-syllable vowels had a minimal co-articulatory effect on following vowels, not exceeding just noticeable difference.…”

“…In both conditions, the standard syllable was expected to create sequence-level predictions of hearing this particular syllable ( Figure 1C and D, longer arrows). In the Context condition, however, the context syllable representing a familiar word beginning was expected to create predictions at another level, namely, to create word-level predictions of hearing the ending of some familiar word ( Figure 1D, shorter arrows; for discussion on knowledgebased predictions, see Poeppel & Monahan, 2011; for phonological predictions in pseudowords, see Ylinen, Huuskonen, Mikkola, Saure, Sinkkonen et al, 2016). To probe the predictions at sequence and word levels, we presented two deviant syllables ( Figure 1A and B).…”

MisMatch Negativity-study showing pre-lexical sensitivity to both primary Final Accent and secondary Initial Accent in French

“…Winkler et al, 2009; Garrido et al, 2009) and, as such, the mmn has had a substantial contribution in investigations of underspecification of phonemic representations (e.g. Eulitz & Lahiri, 2004; Näätänen et al, 2007; Winkler et al, 2009; Deguchi et al, 2010; Ylinen et al, 2016; Scharinger et al, 2016, 2017), as well as the phonological representation of stress patterns (e.g. Ylinen et al, 2009; Honbolygó et al, 2004; Honbolygó & Csépe, 2013; Aguilera et al, 2014; Honbolygó et al, 2017; Garami et al, 2017).…”

Sensory learning and inference is impaired in the non-clinical continuum of psychosis: a replication study

“…Stable environments allow for the development of a robust predictive model, which simultaneously enables an efficient encoding of sensory stimuli while minimizing the demand on cognitive resources 10 . In these environments, healthy individuals form strong predictions about forthcoming sensory stimuli, and their brains consequently produce large prediction error (PE) responses to events that violate such predictions [11][12][13] . The PE response is commonly gauged using electroencephalography (EEG) and an auditory oddball paradigm 14 , in which surprising sounds are embedded in a sequence of predictable sounds.…”