Examining the Relationship Between Speed of Processing, Parental Speech Input and Vocabulary Development

• In Event: Lexical Development: environment and mechanisms

Abstract

Infants show wide variability in their early language acquisition. Past research has linked parental speech input, both in terms of quantity and quality, and the infant’s own abilities to this variability. Hurtado, Marchman and Fernald (2008) investigated the relationship between various measures of parental speech input with the concurrent vocabulary size of 18-month-olds and their vocabulary growth between 18 and 24 months as well as links with lexical speed of processing (SoP). They showed that input affects both infants’ vocabulary growth and their individual SoP. Weisleder and Fernald (2013) further looked into how the amount of child-directed speech impacted vocabulary growth through SoP, revealing that the effect of the input on vocabulary is mediated by the infants’ individual SoP.
However, it is still unclear which aspects of the input influence infants’ linguistic development and to which extent and how they interact with SoP.
Two of the aspects that Hurtado and colleagues studied were maternal mean length of utterance (MLU) as an indicator of input complexity and number of word types as an index of lexical diversity. While both of these are expected to play a major role in determining how children acquire language, we suggest their impact on language development might differ.
Parental MLU might provide benefits, but only if infants can process these more complex sentences. We predict that the influence of parental input complexity on vocabulary development will be mediated through infants’ individual SoP. We hypothesise that fast processors will benefit from higher parental MLU, whereas slow processors may not benefit or even be hindered by regularly hearing such complex input.
Regarding lexical diversity, we would not expect an influence on SoP. Rich variety in lexical input aids infants in building their lexicon, which in turn can affect their processing. Thus, we hypothesise that lexical diversity correlates with vocabulary and has no direct link to SoP, which might indirectly be affected through vocabulary growth.
In order to test these predictions, we assessed 60 18-month-old infants learning Dutch. We measured their lexical SoP in a looking-while-listening paradigm and used the Dutch CDI to determine their vocabulary size at 18 and 24 months. Parental speech input was sampled during a lab-based play session at 18 months. We used CLAN to compute parental MLU and lexical diversity with VOCD. We were only able to include 38 infants, who provided both valid SoP measures and CDI data at 18 and 24 months.
To quantify support for both the alternative and the null hypothesis, we used Bayesian mediation analyses to test our predictions. Results reveal anecdotal evidence against a relationship between parental MLU and vocabulary development (BF10 = 0.234). However, visual exploration of our data suggests that our prediction that faster processors benefit more when exposed to more complex language input (Figure 1) might hold true. Furthermore, we find preliminary evidence for a relationship between lexical diversity and vocabulary growth. Taken together, our results demonstrate how we can further elucidate the role of input and infants’ own processing skills in lexical development.

Authors

• Julia Egger, Max Planck Institute for Psycholinguistics
  Presenting Author

• Caroline Frances Rowland, Max Planck Institute for Psycholinguistics
  Non-Presenting Author

• Christina Bergmann, Max Planck Institute for Psycholinguistics
  Non-Presenting Author