Statistical learning of music- and language-like sequences and tolerance for spectral shifts

Abstract

a b s t r a c tIn our previous study (Daikoku, Yatomi, & Yumoto, 2014), we demonstrated that the N1m response could be a marker for the statistical learning process of pitch sequence, in which each tone was ordered by a Markov stochastic model. The aim of the present study was to investigate how the statistical learning of music-and language-like auditory sequences is reflected in the N1m responses based on the assumption that both language and music share domain generality. By using vowel sounds generated by a formant synthesizer, we devised music-and language-like auditory sequences in which higher-ordered transitional rules were embedded according to a Markov stochastic model by controlling fundamental (F0) and/or formant frequencies (F1-F2). In each sequence, F0 and/or F1-F2 were spectrally shifted in the last one-third of the tone sequence. Neuromagnetic responses to the tone sequences were recorded from 14 right-handed normal volunteers. In the music-and language-like sequences with pitch change, the N1m responses to the tones that appeared with higher transitional probability were significantly decreased compared with the responses to the tones that appeared with lower transitional probability within the first two-thirds of each sequence. Moreover, the amplitude difference was even retained within the last one-third of the sequence after the spectral shifts. However, in the language-like sequence without pitch change, no significant difference could be detected. The pitch change may facilitate the statistical learning in language and music. Statistically acquired knowledge may be appropriated to process altered auditory sequences with spectral shifts. The relative processing of spectral sequences may be a domain-general auditory mechanism that is innate to humans.

Statistical learning of an auditory sequence and reorganization of acquired knowledge: A time course of word segmentation and ordering

“…In addition, these statistical learning effects reflected in neural responses could be observed in both the ERPs and the magnetic counterparts of ERPs. This neural effect could be observed in the responses at approximately 50 ms (i.e., P1/P1m) (Paraskevopoulos et al, 2012;Daikoku et al, 2016), 100 ms (i.e., N1/N1m) (Abla et al, 2008;Furl et al, 2011;Daikoku et al, 2014Daikoku et al, , 2015Koelsch et al, 2016), and 200 ms (i.e., P2/P2m) (Furl et al, 2011) after stimulus onset.…”

“…This suggests that the P2 response resists adaptation more than the N1 response. Due to this long time constant of P2 response modification, in our previous study, when participants listened to auditory sequences for 5 min, statistical learning effects were reflected in the N1 but not the P2 responses (Daikoku et al, 2014(Daikoku et al, , 2015. By training for voice-onset-time perception, the N1-P2 peak-topeak amplitude increased over a long time scale of 10 days, whereas the P1-N1 peak-to-peak amplitude had no significant difference at any time scale (Tremblay et al, 2001).…”

“…This notion may, on the other hand, raise the question of whether adults retain the ability for statistical learning even after they have shifted to domain-specific native-language perception such as syntax and words. Regarding this question, the previous studies demonstrated that adults who had acquired native language also performed statistical learning when they tried to learn an unknown language (Saffran et al, 1999;Gómez, 2002;Peña et al, 2002;Daikoku et al, 2015): adults can extract statistical regularities such as transitional probabilities included in speech sequences and can recognize both word boundaries and word order in speech sequences (i.e., word segmentation and syntax, respectively) (Frost and Monaghan, 2016). However, previous studies have suggested that learners had difficulty in extracting grammatical-like regularity based on non-adjacent dependency included in speech sequences, and the discovery of the grammatical system appears to require a different type of computation as well as transitional probability (Gomez, 2002;Peña et al, 2002).…”

“…Previous studies on auditory statistical learning, using a secondorder Markov chain (Markov, 1971(Markov, , reprinted in 1971Furl et al, 2011;Daikoku et al, 2014Daikoku et al, , 2015 or a word segmentation task (Abla et al, 2008;Paraskevopoulos et al, 2012), have shown that auditory statistical learning can be reflected in neural responses. Previous studies using two different experimental paradigms have shown similar results: neural responses to tones that appear frequently in reference to the most recent tones are more rapidly reduced than responses to tones that appear infrequently.…”

“…Previous neurophysiological studies have investigated only a single aspect of language learning at a time, such as the acquisition of words (Abla et al, 2008;Paraskevopoulos et al, 2012) or grammatical structure (Furl et al, 2011;Daikoku et al, 2014Daikoku et al, , 2015. In our real-world learning environments, when humans listen to language sequences, they are exposed to the linguistic traits of word boundary and word order patterns.…”