Brain Cogn. 2016 Oct;108:1-10. doi: 10.1016/j.bandc.2016.06.008.

Pitch-class distribution modulates the statistical learning of atonal chord sequences.

Tatsuya Daikoku, Yutaka Yatomi, Masato Yumoto.

Department of Clinical Laboratory, Graduate School of Medicine, The University of Tokyo, Tokyo, Japan. E-mail address: yumoto-tky@umin.ac.jp (M. Yumoto), tdaikoku-tky@umin.org (T. Daikoku)

Pubmed

 

Supplment

  1. Intriduction

According to previous studies, pitch and chord perception partially depends on domain-specific modality called pitch class, which is a set of pitches that are separated by octaves [1]. In contrast, domain-general statistical learning has been considered to be an essential strategy for learning chord sequences [2]. Thus, general hypothesis is that music are learned via domain-general and domain-specific systems [3-5]. In the present study, we examined how pitch-class perception affects statistical learning of atonal chord sequences.

 

  1. Methods

Neuromagnetic responses to two chord sequences were recorded from right-handed participants. Using pure tones with frequencies in nine-tone equal temperament (F0 = 250 x 2(n-1)/9 Hz, n = 1-15: Fig. 1), we generated nine augmented triads consisting of two major thirds (250, 315, 397; 315, 397, 500; 397, 500, 630; 270, 340, 429; 340, 429, 540; 429, 540, 680; 292, 367, 463; 367, 463, 583; 463, 583, 735 Hz). The nine chords were divided into three groups that each contained three chords in advance. The three chords within each group were pseudo-randomly presented as a chord triplet and order of triplets was constrained by a first-order Markov stochastic model so that a forthcoming triplet was statistically defined (80, 14, and 6%) by most recent triplet (Fig. 2). In one sequence, each triplet consisted of three chords in same pitch class (pitch-class compact sequence). In the other sequence, each triplet consisted of three chords in different pitch classes (pitch-class incompact sequence). A repeated-measures ANOVA with the peak amplitude and latency of the neural responses were performed.

 

  1. Results

In the pitch-class compact sequence, neural responses to chords transitioned with lower probabilities were significantly increased compared with those with higher probability (P = .042: Fig. 3). In contrast, no significant result was detected in the pitch-class incompact sequence.

 

  1. Discussion

The neural response can be a probe for statistical learning of atonal chord sequences and pitch-class perception. Pitch-class perception that has been interpreted as one of domain-specific mechanisms in music may facilitate domain-general statistical learning of chord sequences.

 

 

Fig. 1. The pitch class-pitch height space

 

 

Fig. 2. Transition diagrams (a) and excerpts from the pitch-class triplet sequence (b) and pitch-class randomized sequence (c) Thick and dashed arrows were high and low transitional probabilities, respectively. Each colour represents three distinct pitch classes.

 

 

Fig. 3. Grand-averaged source-strength waveforms for P1m to chords in pitch-class triplet sequence (a) and pitch-class randomized sequence (b). The solid lines represent responses to triplet transitioned in clockwise direction with higher probability. The thin lines represent responses to triplet transitioned in counterclockwise direction with lower probability. The dashed lines represent difference waveforms.

 

  1. Reference

[1] Jackendoff R, et al, 2006. The capacity for music: what is it, and what’s special about it? Cognition 100, 33-72.

[2] Saffran JR et al, 1996. Statistical learning by 8-month-old infants. Science 274, 1926-28.

[3] Daikoku T et al., 2014. Implicit and explicit statistical learning of tone sequences across spectral shifts. Neuropsychologia 63: 194-204.

[4] Daikoku T et al., 2015. Statistical learning of music- and language-like sequences and tolerance for spectral shifts. Neurobiology of Learning and Memory 118: 8-19.

[5] Daikoku, T., Yatomi, Y., & Yumoto, M., 2017. Statistical learning of an auditory sequence and reorganization of acquired knowledge: A time course of word segmentation and ordering. Neuropsychologia 95: 1-10.