An acoustic investigation of traces to children’s early omitted English articles: a case study

Young children are more likely to omit unstressed function words when they do not form a disyllabic foot with the previous word e.g., Tom [pushes]Foot the dog than when they form a disyllabic foot with the previous word e.g., Tom [pushed the]Foot dog (e.g., Gerken, 1991, 1994; Demuth & McCullough, 2009). However, previous research indicates that when children appear to omit weak syllables in content words, they leave imperceptible prosodic traces to the omitted syllables (Carter & Gerken, 2004) suggesting that the omission of unstressed syllables in content words is incomplete. However, it is not clear whether children also leave traces to the other type of unstressed syllables i.e, function words. This thesis, therefore, explored whether there is an acoustic trace to children’s omitted function words, namely English articles. To this end, a longitudinal speech corpus of one English-speaking child between 1;5 to 1;9 (years; months) (Demuth et al., 2006) was screened for three types of utterances in which articles were 1) produced e.g., find the dog; 2) omitted from obligatory context e.g., find dog; or 3) not required (control) e.g., find dogs. From these utterances, three durations were measured that could be indicative of a trace and compared across the utterance types using regression analyses. Results show no evidence of an acoustic trace to omitted articles which may imply that children delete the entire syllable instead of leaving a prosodic trace. A discussion of the implications of this finding was provided for phonological/ prosodic accounts of unstressed syllable omission and for speech production processes in general.