Children's early word productions are often inconsistent in form, showing apparent between-speaker and within-speaker variability in phonological and morphological shape. Recent results show more adult-like productions in certain contexts, suggesting that children's phonological and morphological representations are more robust than often assumed. Previous acoustic/phonetic studies have also shown that children sometimes make systematic acoustic or gestural feature contrasts not perceived as such by adults (covert contrasts). However, research on such covert contrast has been limited to a few case studies. Little is therefore known about the extent of these processes across young, normally developing learners, nor about how this relates to early variability in natural speech. This proposal addresses this problem by using acoustic and articulatory measures to test the hypothesis that many of children's early phonological and morphological representations are more adult-like than they appear. To investigate this issue we draw on recent theoretical and empirical work on adult speech, where variability in production has been analyzed in terms of individual acoustic/articulatory cues to grammatical feature contrasts. However, most of this research to date has focused on the beginnings of words (onsets): very little is known about how adults or children produce contrasts at the ends of words, where languages like English encode important grammatical morphemes. We therefore extend and compare previous findings to the acoustic/phonetics of child- directed speech (by adults) and to child speech, focusing on the question of when and how children acquire word-final (coda) consonants. To do this we first carry out detailed acoustic analyses of existing adult and child speech corpora, focusing on the production of simple non-inflected and inflected words (e.g., dig, digs). We then compare these results with those of 2-year-old's speech productions under more controlled experimental conditions. The goal is to better understand the principles that underlie phonological and morphological variation in child speech. We hypothesize that between-speaker variability will occur most often for feature contrasts that are signaled by multiple articulatory gestures and acoustic cues (e.g., the voicing of word-final consonants), where developing speakers may initially implement different feature cues. In contrast, we hypothesize that within- speaker variability will be highly predictable and context dependent (e.g., sensitive to word complexity and position within the utterance), reflecting constraints on phonological form. The results will provide a framework for better understanding children's early knowledge of phonological and morphological structure, and will lay the foundation for a developmental model of language planning. This will have implications for more accurate diagnosis and design of interventions for both child and adult populations with language impairments.

Public Health Relevance

The findings from this research will reveal much about the structure of children's early phonological and morphological representations, and how these develop over time. In so doing, they will provide insight into the mechanisms underlying variability in children's productions, and the possible impact of input, complexity and contextual factors in understanding between-speaker and within-speaker variability in language production. These findings will lay the foundation for constructing a developmental model of language planning, providing a framework for designing more effective tools for diagnosing and treating aspects of language impairment.

National Institute of Health (NIH)
Eunice Kennedy Shriver National Institute of Child Health & Human Development (NICHD)
Research Project (R01)
Project #
Application #
Study Section
Special Emphasis Panel (ZRG1-BBBP-M (02))
Program Officer
Freund, Lisa S
Project Start
Project End
Budget Start
Budget End
Support Year
Fiscal Year
Total Cost
Indirect Cost
Massachusetts Institute of Technology
Organized Research Units
United States
Zip Code
Yuen, Ivan; Cox, Felicity; Demuth, Katherine (2014) Three-year-olds' production of Australian English phonemic vowel length as a function of prosodic context. J Acoust Soc Am 135:1469-79
Mealings, Kiri T; Demuth, Katherine (2014) The role of utterance length and position in 3-year-olds' production of third person singular -s. J Speech Lang Hear Res 57:484-94
Song, Jae Yung; Demuth, Katherine; Evans, Karen et al. (2013) Durational cues to fricative codas in 2-year-olds' American English: voicing and morphemic factors. J Acoust Soc Am 133:2931-46
Mealings, Kiri T; Cox, Felicity; Demuth, Katherine (2013) Acoustic investigations into the later acquisition of syllabic -es plurals. J Speech Lang Hear Res 56:1260-71
Song, Jae Yung; Demuth, Katherine; Shattuck-Hufnagel, Stefanie (2012) The development of acoustic cues to coda contrasts in young children learning American English. J Acoust Soc Am 131:3036-50
Sundara, Megha; Demuth, Katherine; Kuhl, Patricia K (2011) Sentence-position effects on children's perception and production of English third person singular -s. J Speech Lang Hear Res 54:55-71
Theodore, Rachel M; Demuth, Katherine; Shattuck-Hufnagel, Stefanie (2011) Acoustic evidence for positional and complexity effects on children's production of plural -s. J Speech Lang Hear Res 54:539-48
Yung Song, Jae; Sundara, Megha; Demuth, Katherine (2009) Phonological constraints on children's production of English third person singular -s. J Speech Lang Hear Res 52:623-42