2026/06/20 by Loan Vuong, Patrick C.M. Wong, Patrick C. M. Wong
Psychology · Social Sciences · #Child and Animal Learning Development #Language Development and Disorders #Language and cultural evolution
paper · doi:10.1016/j.dr.2026.101280
Much of what we know about word learning comes from experimental paradigms that treat the extraction of word forms (segmentation) and the assignment of meaning (mapping) as separable processes. Although this division has yielded important insights, it may underestimate how their coordination supports learning in natural environments. Evidence from statistical segmentation—a widely studied mechanism for word discovery—suggests that representations derived from auditory statistics are often fragile and short-lived. These observations raise a critical question: how do children build stable lexical knowledge from initially fragile representations? Drawing on prior interactive accounts of segmentation and meaning mapping, we argue that semantic integration plays a distinct role in memory by helping to stabilize nascent word representations. This integration may be especially important under the sparse and acoustically variable conditions of natural speech, where learners must resolve uncertainty about linguistic relevance while balancing stability with flexibility. We outline testable hypotheses and propose specific indices for assessing robustness in word learning.