also suggest that the environmental support provided by American middle-class parents reflects the demands of a technological society in which a cultural priority is preparing children for academic achievement and managerial and professional occupations (Hart and Risley, 1995).

However, Fernald and Morikawa (1993) compared the interactions of 30 middle-class American parents to those of 30 monolingual wives of affiliates of Japanese companies visiting in America. All the parents, given identical sets of toys and video-taped in 10 to 15 minutes of toy play during home visits, were found to fine-tune their speech to the skill levels of their children. The American parents talked with their 6-, 12-, and 19-month-old children primarily about objects (naming them), and the Japanese parents talked primarily about social relations (polite verbal routines accompanying the exchange of objects, encouraging positive actions on toys: “pat it gently”). The major influence on the language children learn is the culture’s socialization practices, which aim to establish and maintain the “language learning games” of the culture (Tomasello, 1992).

Socioeconomic Status. A significant association between children’s performance on cognitive tasks and parent income and years of education is well documented (see Gottfried, 1984; Neisser et al., 1996; Stipek and Ryan, 1997), both within and across cultural groups. Parents with the advantages of education are reported to interact with their infants in ways relevant to mainstream schooling. They prompt their infants to respond to books and pictures, ask questions that promote organizing knowledge into names and categories (Schieffelin and Ochs, 1983), and arrange for children to have materials, uninterrupted time, and adult support for exploratory play that challenges them to initiate actions and combine and modify them in order to achieve a goal (Bruner, 1974). Duncan et al. (1994) demonstrated that the effect of poverty is partially mediated by the home environment. One-third of the variance in age 5 IQ scores that was associated with income was eliminated when measures of the home learning environment, family social support, maternal depression, and active behavioral coping were included in the model. The extent to which poverty is related to quality of the home environment depends on the de-