AIRS 3rd Annual Meeting: 2013

Title: Are infants’ and toddlers’ songs recognizable without the words?
Authors: Helga Rut Gudmundsdottir (University of Iceland /Brams, Montreal), Sandra E.
Trehub (University of Toronto Mississauga/Brams, Montreal)

Abstract

There is little systematic research on singing in early childhood. Nevertheless, there are claims
that recognizable songs emerge between children’s first and second birthday, a period
corresponding to children’s typical progression from one-word to multi-word utterances.
Prelinguistic babbling reveals melodic aspects of the ambient language. Some scholars propose
that individual differences in language acquisition reflect divergent inclinations towards words
or intonation, resulting in so-called “word-babies” or “intonation-babies.” Case studies of song
acquisition reveal “melody-first” and “words-first” babies. The primary goal of the present
study was to ascertain whether songs of infants and toddlers (16 months - 3 years) are
recognizable on the basis of melody (pitch and rhythm patterns) alone. The singing, which was
recorded by parents at home, was uploaded to YouTube or sent directly to the principal
investigator (P.I.). Songs were chosen that were (a) familiar or unfamiliar to North American
listeners and (b) produced by several children with foreign lyrics. Two songs had familiar
melodies (Twinkle, Happy Birthday) and two had unfamiliar melodies, with each song
produced by six different children. English- and French-speaking adults will listen to the song
samples, naming the songs when possible and declaring them unfamiliar if not. Subsequently,
the key words of the songs will be isolated for identification by native speakers of those
languages. The results will shed light on adults’ ability to recognize children’s early songs in the
absence of linguistic cues and their ability to recognize the words of those songs in the absence
of contextual cues. [relevant also to 1.2]