O modelo de exemplares e o desenvolvimento lexical de crianças gêmeas: o que nos informam os dados de produção e não-produção

Abstract

In this study, we investigate the lexical development of twin children from different regions of Brazil, with the aim of analyzing data marked as “non-production” (NP) and as “word production”. This research is based on the theoretical assumptions of Complex Adaptive Systems (Beckner et al., 2009; Baia, 2013) and the Exemplar Model (Bybee, 2016; Gomes, 2020). We conducted a picture-naming experiment in which the target words were monosyllabic, disyllabic, or trisyllabic, with different stress positions. The data analyzed come from five pairs of twins, aged between 1 year and 6 months and 3 years and 7 months, from the cities of Ferraz de Vasconcelos (SP), Itu (SP), Curitiba (PR), Jequié (BA), and Vitória da Conquista (BA). According to the results, some lexical items were not produced by the children because they did not recognize the images, which suggests that these items are not part of their exemplar clouds. Additionally, some words were produced differently from the expected target due to the children having different exemplars for those items. We also found variation in production between members of the same twin pair. Thus, the results demonstrate that usage frequency and linguistic experiences impact the formation of exemplar clouds, with more frequent items creating more robust representations, which, in turn, influence the trajectory of lexical development in early childhood.