Lexical Input to Young Children at Home. A Study with Three Social Groups in Argentina

ROSEMBERG, C. R.; STEIN, A.; ALAM, F.; PIACENTE, T.

Congreso; 2013 SCRD Biennial Meeting; 2013

The study analyzes the lexical properties of the linguistic context that children from 3 social groups from Argentina -marginalized-urban neighbourhoods(UM), middle income families with university education(MI) and Toba native suburban communities(TN)- are exposed to daily in play, meal and hygiene situations. Differences in the size of children?s vocabularies develop in infancy as a consequence of the opportunities offered by their environment rather than as a consequence of individual capabilities. At 4 years of age, the size of a child?s vocabulary is determined to a great extent by the number of word tokens and word types that their parents use, and by their parents? use of uncommon vocabulary (Weizman & Snow, 2001). Children?s acquisition of a particular type of sophisticated words, those referred to abstract phe¬nomena, general categories, properties not directly observable, cognitive verbs or verbs of will and emotion, is closely related to the way in which adults use them. Studies have shown that although the quality of the lexical surround differs between families from a same social group, differences between socio-economic groups are marked (Hart & Risley, 1995). This fact points to the importance of studying the lexical environment that young children are exposed to at home. The sample of the present study included: a) 78,664 interactional turns (children: 36,656; interlocutors: 42,008) recorded in the homes of 17 children from middle income households; b) 78,445 recorded in the homes of 19 children from poor marginalized neighbourhoods (children: 33,549; interlocutors: 44,896) and c) 26,468 interactional turns (children: 6,925; interlocutors: 19,543) recorded in the homes of 11 children from suburban indigenous population. Corpora consisted of spontaneous situations -play, hygiene, meal- that were transcribed according to the Code for the Human Analysis of Transcripts and analyzed using the Computerized Language Analysis. The quantity and diversity of the vocabulary that children were exposed to were measured. The vocabulary was also analyzed taking into account its level of abstraction. Findings show differences in the input children were exposed to: quantity: TN: 27,153, MI: 18,650, UM 13920; diversity: MI: 2,512, TN: 2453, UM: 1,899. The ANCOVA analysis showed significant differences between the groups (quantity: F (2, 43) = 6.95, MSE = 82427198.74, p < .01; diversity: F (2, 43) = 5.00, MSE = 515560.90, p < .05). It also showed a significant relationship between the quantity of children's interlocutors (mean MI: 7; UM: 21; TN: 28) and the amount of words they produced. (F (1, 43) = 6.90, MSE = 82427198.74, p < .05). The correlation analysis only detected a positive and significant relationship between the quantity of interlocutors and the quantity and diversity of words produced in the data from the Toba native community (quantity: r = .83, p < .01; diversity: r = .87, p < .001). Differences in the amount of non-observable nouns, adjectives and verbs, were found between the three groups. The analysis suggests an association between words referencing non?observable phenomena and the social group (Table 1). Findings are discussed taking previous research in other cultures and languages into account.