The effects of contextual diversity on lexical processing: A scoping review

Research into the effects of contextual diversity on lexical processing has flourished in the past 20 years, encompassing diverse tasks, populations, and languages, and underpinning influential theories of word learning (Nation, 2017). This review provides a comprehensive synthesis of the field. Eighty-six articles (145 experiments) met our preregistered inclusion criteria, comprising three distinct experiment types; behavioural studies (N=111), computational modelling studies (N=20), and corpus validations (N=14). Across these experiments, the terminology used to refer to different diversity metrics has been applied inconsistently and somewhat arbitrarily. Behavioural studies have largely focused on word form processing (i.e., lexical decision tasks), showing consistent processing benefits for high-diversity words, regardless of the metric used. Effects of diversity on word-meaning processing were more mixed. We also review findings from a wide range of other tasks that are often overlooked (e.g., recognition memory). Computational modelling studies suggest that metrics that quantify the distinctiveness of contexts in which words are used better predict behaviour than metrics that simply count those contexts. Performance of these metrics is further improved by accounting for social patterns of language use across individuals or topics. Corpus validations show that diversity effects are consistent across languages. This review confirms that diversity in linguistic experience is a key organizational principle of the lexicon, but highlights that diversity effects vary across types of linguistic knowledge and task demands. We highlight key areas where more data is required, and make specific recommendations for improving future research based on a structured research cycle.