You are viewing the SNL 2017 Archive Website. For the latest information, see the Current Website.

Poster D19, Thursday, November 9, 6:15 – 7:30 pm, Harborview and Loch Raven Ballrooms

Prior knowledge influences in learning and consolidating new words

Emma James1, M. Gareth Gaskell1, Lisa Henderson1;1University of York

The complementary learning systems account of memory proposes that a newly encountered word is initially bound and stored as a distinct representation in the hippocampus, but that offline reactivation enables a strengthening and integration of this information with neocortical-based vocabulary (McClelland et al., 1995; Davis & Gaskell, 2009). Extant evidence suggests that this memory consolidation process is facilitated by related prior knowledge and/or by sleep rich in slow oscillations. James et al. (2017) have proposed that these mechanisms might differentially support word learning across development: children benefit from larger amounts of slow-wave sleep that supports ongoing neural maturation, whereas adults experience a greater benefit from their richer knowledge base. We present two behavioural experiments that examine the role of existing knowledge in consolidating new linguistic information in children and adults. The availability of prior knowledge was manipulated experimentally by training novel words with/without close orthographic neighbours in the English lexicon. For example, a single letter swap from ballow can form ballot, wallow, bellow etc., whereas no existing words can be formed from marpan. Participants were children aged 7-9 years (n = 232) in Experiment 1, and adults (n = 79) in Experiment 2. Prior knowledge contributions to word learning was further examined on an individual level by including a standardised measure of existing vocabulary knowledge. In order to test hypotheses regarding prior knowledge contributions at learning and over a period of consolidation, we tested memory for the new words using cued form recall and a recognition task at three time points: the same day, next day, and one week later. Both children and adults showed better cued recall performance for words with neighbours, suggesting that they could access prior knowledge to support learning in this paradigm. However, whilst adult memory performance remained influenced by prior knowledge over the course of the week, the effect of word neighbours was reduced at subsequent test points for children. These contrasting patterns are consistent with a model in which children have superior sleep-associated consolidation processes compared to adults, which can better support novel learning. An additional condition in Experiment 2 showed that a single word neighbour was sufficient to boost learning, and that learning in this condition only was predicted by adults’ existing vocabulary knowledge. Our results are consistent with the proposal that children’s superior sleep architecture may facilitate learning of more novel stimuli during this period of development, whereas adults are more reliant on prior knowledge during consolidation. A third experiment is underway to re-examine the contributions of vocabulary knowledge in children, by including the single neighbour condition and by removing potentially confounding contributions of orthographic knowledge. We discuss implications for the complementary learning systems account of word learning in light of these developmental differences.

Topic Area: Language Development

Back to Poster Schedule