How do high frequency words assist language acquisition in infant and adult learners?

Frost, R. L. A. & Monaghan, P. (2015, October). Talk presented at the LuCiD Annual Mini Conference.

That learners can extract transitional information from speech and use it to help infer word boundaries and linguistic rules has been well documented in both infant and adult literature. Interestingly, recent findings have indicated that the presence of high frequency words in speech may be benefit language learning in infants. Bortfield, Morgan, Golinkoff and Rathburn (2005) found that 6-month-olds were able to identify new words more easily when they appeared in speech next to words that they were already familiar with (e.g. their own name). Monaghan and Christiansen (2010) propose that such high frequency words may act as anchors for speech segmentation to occur around, and they suggest that these words may also assist with grammatical categorisation. The present set of studies addresses these claims, and explores how the presence of high frequency words affects learning in infants versus adults.
In Experiment 1 we familiarised adults with a language containing four target words, and four marker words. Words were split into two categories (A and B), and these categories could be identified only by co-occurrence statistics between words and the marker words that preceded them in the speech stream. To examine the extent to which word frequency influences learning,usage of marker words was manipulated over three conditions:Marker 0 (contained target words only), Marker 1 (contained target words and two marker words, with words in each category preceded by one marker word), and Marker 2 (target words in each category were preceded by one of two marker words). After training, participants completed three tests of segmentation and categorisation. Findings from preliminary analysis indicate that all participants were able to segment the stream. Performance was slightly higher for participants whose training stream contained marker words as well as target words, however there were no significant main effects, indicating that the benefit of high frequency items for segmentation may be subtle in experienced learners. Performance on the categorisation tasks will also be discussed.
In Experiment 2, we will expose 8- and 12-month-old infants to the artificial language, and will use eyetracking to assess segmentation and categorisation. Anticipated findings of this study will be discussed.
Bortfeld, H., Morgan, J.L., Golinkoff, R.M., & Rathbun, K. (2005). Mommy and me: familiar names help launch babies into speech-stream segmentation. Psychological Science, 16, 298-304.
Monaghan, P. & Christiansen, M. H. (2010). Words in puddles of sound: modelling psycholinguistic effects in speech segmentation.