2026/07/01 by Yiqun Zhang, Subin Kim, Marisa Casillas
paper · doi:10.1111/cogs.70241
Abstract Children's early vocabularies have been observed to be predominated by nouns, particularly concrete and animate nouns. This phenomenon, known as the “noun bias,” is argued to reflect a universal conceptual advantage for concrete objects in early word learning, because such referents are perceived as bounded, individualized concepts that are ready to be mapped onto linguistic labels. The noun bias has been observed across a wide range of languages using multiple methodological approaches, establishing it as a robust feature of early lexical development. However, the magnitude and consistency of the noun bias have also been observed to vary across linguistic and cultural contexts, motivating further investigation of the factors that shape this phenomenon. The present study systematically reviews and quantitatively analyzes evidence from 68 published and unpublished papers spanning 28 languages over five decades (1973–2025) to examine how the noun bias is shaped by linguistic and other properties. The review suggests general evidence supporting a noun bias. However, research methodology and noun morphology emerge as significant predictors of variation in the observed noun bias magnitude. In addition, the presence of specific linguistic attributes—noun plural marking and verb past tense marking—explains variation in the observed noun bias outcomes from the 68 papers reviewed. This comprehensive view of the literature sheds light on limitations and inconsistencies in the literature that should be remedied in future work to further advance theoretical accounts of the noun bias.