AIRS 4th Annual Meeting: 2012

Title: Early Song Singing Infants Discover a Rule-Based System
Authors: Stefanie Stadler Elmer (University of Zurich), Daniel Muzzulini (Kantonsschule
Alpenquai Luzern)

Abstract

The phenomenon of early song singing allows studying the infants’ emerging productive
musicality. Song singing consists in combining linguistic and musical elements. Both types of
elements are vocal sounds differing in fundamental frequency, duration, intensity, and timbre.
Here, we focus on children growing up with German as their native language. What does early
song singing reveal about the young singer’s ability to adopt rule-based behaviour? The
singing-before-speaking hypothesis expects musical features to prevail, whereas some music
educators expect the linguistic elements to appear first. To address this question, the following
steps are taken: We carry out microgenetic analyses of young children’s song singing using
acoustical methods. We also analyze traditional German children’s songs in order to derive and
formalize rules. Finally, a meta-analysis is carried out to match the theoretical rules with the
children’s productions. Results show that musically stimulated children at the beginning of
their second year are able to produce pitch patterns in structured time, to create songs that
match well with culture-specific musico-linguistic rules. Melodic contour, phrase segmentation,
metric and rhythmic patterns, their repetition and variation, as well as onomatopoeic syllable
formation are major elements children use to create a stream of ordered sounds. Linguistic
features such as word formation, however, are less prominent or even absent. Typical for
emotional states of playfulness and wellbeing, musically stimulated young children’s
vocalizations show more differentiations with respect to musical features than to linguistic
ones. These children discover very early how to control pitch and intensity, which prepares
them for pre-musical and singing-like vocalizations. Less musically stimulated children tend to
focus primarily linguistic features. Altogether, results suggest that song timing rules provide a
temporal framework for regular motor movements and for filling in linguistic elements that still
may lack semantic meaning.