How do people learn numerals? Developing a cross-situational number word learning task.

<p dir="ltr">Numerals are relatively infrequent in children’s input, and when they are used, the referenced quantity is often absent or unclear. Although it has been argued that productive linguistic rules are central to the development of our understanding of cardinals (e.g., Schneider et al., 2020; Spelke, 2017; Yang, 2016a) and ordinals (Meyer, 2019), such linguistic cues do not sufficiently explain how people successfully learn numerals despite ambiguous input. We argue that cross-situational word learning (CSWL) — a form of statistical learning (see: Smith & Yu, 2008; Yu & Smith, 2007) — could offer an explanation, and investigated this by developing a novel CSWL tasks. The first version of this task (Experiment 1) was administered to 25 Dutch-speaking adults, the second version (Experiment 2) was completed by 30 typically developing Dutch-speaking children (4;01-7;09). Although the adult participants were able to learn cardinals and ordinals in the semi-artificial language, there was no evidence for learning in the child study. The adult findings thus suggest that CSWL could have explanatory value for the acquisition of numerals, but further research is necessary to better understand the potential role of CSWL in early numerical development. We therefore discuss methodological choices that could have influenced our results and make concrete suggestions for future research. As such, our study provides crucial first steps to understanding how people infer number word meanings from ambiguous input.</p><p dir="ltr">This collection contains the experiment files and analysis files (data files, R scripts, and html files) that belong to the study titled "How do people learn numerals? Developing a cross-situational number word learning task." These data were collected and analyzed as part of the research for a dissertation written Heleen de Vries, under supervision of Judith Rispens, Caitlin Meyer, and Alla Peeters-Podgaevskaja.</p>