Infants’ Neural Tracking of Melodic Expectations in Children’s Songs

Authors

  • Larissa Reitinger University of Vienna
  • Trinh Nguyen Italian Institute of Technology
  • Roberta Bianco Italian Institute of Technology

Abstract

Musical engagement appears to be universal across human cultures. The ability to process music is already present in early stages of brain development, suggesting a predisposed mechanism that enables neural encoding of rhythm and pitch. Findings show that this early sensitivity to auditory input is facilitated by a cognitive mechanism called statistical learning, where probabilistic relations are extracted from incoming sensory information [1]. This allows for the formation of a continuously updated, feature-specific, internal model, which enables anticipation of subsequent auditory stimuli. Interestingly, already neonates show a clear differentiation in neural activity when the structural regularity of auditory cues is violated, indicating a context-specific model adaptation [2]. Since most studies on infant music cognition use non-naturalistic stimulation paradigms, it remains unclear whether the formation of melodic expectations based on statistical relations of prior sensory input is present during naturalistic music listening in infants. The present study aims to answer the critical question of whether distinct cortical responses in infants encode pitch and note-onset expectations during music listening. Using electroencephalography (EEG), we record cortical signals of healthy infants at the age of three, six, and twelve months (n=30 each) during passive listening to polyphonic MIDI (musical instrument digital interface) versions of two children’s songs. Subsequently, we investigate pitch and note-onset expectations by quantifying melodic surprise of the melody lines using IDyOM (Information Dynamics of Music), which is an empirically validated framework for computationally assessing musical structures based on variable-order Markov models [3]. Using temporal response function (TRF) analysis and ridge regressions, we investigate cortical encoding of melodic expectations. Here, we expect a higher predictive accuracy when predicting infants’ neural response from acoustic, timing, and pitch information than from each parameter alone. Additionally, we hypothesize this prediction accuracy to increase with age, reflecting an improvement in tracking structural regularities across development. These results contribute to understanding melodic expectancy during naturalistic music listening and neural tracking in ecologically valid settings in early infancy.

References

[1] J. R. Saffran and N. Z. Kirkham, ‘Infant Statistical Learning’, Annu. Rev. Psychol., vol. 69, pp. 181–203, Jan. 2018. doi:10.1146/annurev-psych-122216-011805

[2] J. Todd, G. P. Háden, and I. Winkler, ‘Relevance to the higher order structure may govern auditory statistical learning in neonates’, Sci. Rep., vol. 12, no. 1, Art. no. 1, Apr. 2022. doi:10.1038/s41598-022-09994-0

[3] M. T. Pearce, "The Construction and Evaluation of Statistical Models of Melodic Structure in Music Perception and Composition,” dissertation, City University London, UK, 2005.

Published

2023-06-05