Children who read more tend to be better readers than children who read less. Reading exposure captures not onlythe number of times words are experienced but also the breadth of the contexts words appear in. Using a large children’s corpusof written language, we quantified the former as word frequency and the latter as Semantic Diversity (SemD) (Hoffman etal., 2013). SemD was indexed using Latent Semantic Analysis to calculate the degree of semantic dissimilarity between thecontexts in which each appeared. We selected 300 words that varied in SemD for a visual lexical decision and naming task with9-year-old children (N=114). Results showed that both frequency and SemD were associated with performance, independentlyaccounting for variation in speed and accuracy. Those words high in frequency and high in SemD were read more efficiently.These findings show that factors beyond frequency are important in determining children’s word reading.

We investigated whether semantic diversity (SemD) influences immediate serial recall for words. SemD was calculatedusing LSA to quantify semantic similarity across contexts in large corpus. We examined the effects of SemD and im-ageability, a classic semantic variable. Participants saw and recalled the 6-word list by typing out the words in correctserial order. Experiment 1 was conducted in the laboratory (N=40). There was no main effect of SemD or imageabilitybut exploratory analyses showed that SemD was modulated by list position and imageability. Among high-imageabilitywords, low-SemD words were better recalled in latter positions (4 & 5) of the list. Experiment 2 conducted online (N=44)replicated the results, showing better recall of low-SemD words in the high-imageability condition at Position 5. Thesefindings suggest that the availability of more semantic connections induces more competition between items, which im-pacts on performance later on in serial recall.

People learn new words in narrative contexts. Little is known about the influence of the emotional valence of the text on word learning. In a pre-registered experiment, we investigated whether emotional narrative context shapes word learning. English adults (N = 76) read 30 novel adjectives embedded in 60 short narratives (20 positive, 20 negative, and 20 neutral valence). Post-tests assessed learning (immediate and 24 hours later) and examined whether the valence of the novel words can be inferred from contextual valence. Compared to the neutral context, emotional contexts (both positive and negative) facilitated word form learning in the immediate post-tests, but only negative emotion words were recognized better 24 hours later. Furthermore, the valence of the context was reflected in the word meanings participants generated for each novel word. These findings are discussed with reference to theories of affective embodiment and its implications for supporting the learning of abstract concepts.

Emotion is closely associated with language, but we know little about how children express emotion in their own writing. We used a large-scale data-driven approach to investigate whether emotional expression via writing changes through development, and whether it varies for boys and girls. We first used a lexicon-based bag-of-words approach to identify emotional content in a large corpus of stories written by 7- to 13-year-old children (N>100,000). Generalized Additive Models were then used to model changes in sentiment across age and gender. Two other approaches (BERT and TextBlob) validated and extended these analyses, revealing converging findings that positive sentiments in children’s writing decrease with age. These findings echo previous studies showing lower mood and increased acquisition of negative emotion words across development. We also found stories by girls contained more positive sentiments than boys. Future experimental work should further investigate the complex relationships between written language and emotion across development.