The Role of Parental Speech in Infant Language Development: Insights from 8-9-month-old Norwegian Infants 

• In Event: The Ins and Outs of Language Development: Integrating Multiple Measures from Inside and Outside the Laboratory

Abstract

Intuitively, many parents use Infant-Directed Speech (IDS) when talking to their infants: they lengthen sounds, exaggerate and vary their pitch of voice, and hyper-articulate the corner vowels /i/, /a/ and /u/. IDS has been claimed to be universal and to facilitate early language learning. For instance, exaggerated variability in pitch can help infants segment and recognize words in fluent speech, whereas hyper-articulation has been claimed to enhance the clarity of speech. However, a number of limitations in previous research (e.g., little or no manual control over the acoustic processing, lack of variability in sound context), mostly with English-learning infants, and the results of recent advanced acoustic analyses, showing poor sound distinctiveness in IDS, challenge these interpretations and raise doubt about the facilitating role of IDS in language development.
The current project addresses limitations of previous research and examines the role of acoustic properties of IDS in early language development of Norwegian 8-9-month-old infants, as revealed by direct (word comprehension, in an eye-tracking looking while listening paradigm) and indirect (reported vocabulary size via parental reports, CDIs) measures. Norwegian, unlike English, uses prosody (changes in pitch tone) and sound lengthening to convey meaning, suggesting that remarkable variation is already present in adult-directed speech (ADS).
In a first study, fifteen Norwegian mothers were recorded while reading to their child a made-up child-friendly book that, among other, featured 9 Norwegian long vowels in 5 different contexts, repeated twice. This design allowed us to control for the quality and quantity of linguistic material across parents. We analysed, semi-manually, mothers’ productions and computed a number of acoustic measures typically examined in infant research: e.g., average pitch, pitch variability, vowel length and vowel space extension, that we complemented with a number of novel measures that index the preciseness/clarity of vowel categories in the acoustic space and their distinctiveness via clustering analysis. The results revealed a significant positive relationship between the two measures of language development, i.e., looking time and the vocabulary size; yet, their relationship with pitch, duration and vowel space extension was negative, suggesting that exaggeration in the use of lexically-meaningful cues might disrupt learning. Importantly, the results revealed that larger vowel extension is associated with less precise vowel categories, which, in their turn, are related to poorer distinctiveness of vowel categories in the acoustic space; these data suggest that vowel hyper-articulation does not necessarily lead to clearer speech.
In a second study, with 50 infant-parent dyads, we performed analogous acoustic analyses, this time, for both IDS and ADS, collected when reading the same book. The latter enabled us to assess the 'articulatory' effort that a parent does while talking to her infant and how it relates to infant's word comprehension as measured by an eye-tracking task and her vocabulary size. Data analysis is ongoing.
The results of this study will provide unprecedented knowledge on the acoustic properties of IDS (as compared to ADS), their relationship to each other, and will identify IDS properties that may predict infants’ early language development.

Authors

• Natalia Kartushina, University of Oslo, Norway
  Presenting Author

• Silje Robberstad, University of Oslo
  Non-Presenting Author

• Julien Mayor, University of Oslo
  Non-Presenting Author