Is the Chinese Number-Naming System Transparent? Evidence from Chinese-English Bilingual Children

Canadian Journal of Experimental Psychology, Mar 2006 by Rasmussen, Carmen, Ho, Elaine, Nicoladis, Elena, Leung, Joyce, Bisanz, Jeffrey

Abstract

Chinese-speaking children have been shown to have an advantage over English-speaking children in a variety of mathematical areas, including counting. One possible explanation for the advantage in counting is that the Chinese number-naming system is relatively transparent, compared to English, in that number names typically are directly indicative of base-10 structure (e.g., 12 is named "ten-two" rather than "twelve"). To determine whether the transparency of the Chinese number-naming system influences counting in bilingual children, we tested 25 ChineseEnglish bilingual children between the ages of 3 and 5 years, both in English and in Chinese. Children were asked to count as high as they could (abstract counting) and also to count objects in small, medium, and large arrays (object counting). No evidence was found for transparency or for transfer from one language to the other. Instead, relative proficiency in the two languages influenced counting skill. These results are discussed in terms of linguistic and cultural variables that might account for cross-linguistic differences in counting.

R�sum� Nous savons que les enfants d'expression chinoise ont un avantage sur les enfants d'expression anglaise dans divers domaines des math�matiques, y compris l'aptitude � compter. L'avantage des enfants d'expression chinoise peut s'expliquer par le fait que le syst�me chinois de d�nomination des nombres est plus ou moins transparent, comparativement au syst�me anglais, en ce sens que, g�n�ralement, le nom des nombres se rapporte directement � la structure en base 10 (par ex., le chiffre 12 se nomme � dix-deux � et non pas � douze �). Afin d'�tablir si la transparence du syst�me chinois de d�nomination des nombres a un effet sur l'aptitude � compter chez les enfants bilingues, nous avons test�, en utilisant l'anglais et le chinois, 25 enfants bilingues (chinois-anglais) de 3 � 5 ans. Les enfants �taient invit�s � compter le plus loin possible (num�ration abstraite) et � compter les objets appartenant � des petits, des moyens et des grands ensembles (comptage d'objets). Nous n'avons observ� aucun �l�ment probant mettant en �vidence la transparence ou le transfert d'une langue � l'autre. Au contraire, nous avons constat� que la ma�trise relative des deux langues agissait sur l'aptitude � compter. Nous analysons les r�sultats � la lumi�re des variables linguistiques et culturelles qui peuvent prendre en compte les diff�rences interlinguistiques touchant l'aptitude � compter.

Chinese-speaking children have been shown to have an advantage over English-speaking children in a variety of mathematical areas (Geary, 1996; Miller, Major, Shu, & Zhang, 2000; Miller, Smith, Zhu, & Zhang, 1995; Miller & Stigler, 1987; Muira, Okamoto, Vlahovic-Stetic, Kim, & Hann, 1999). For example, Miller et al. (1995) found better performance in Chinese-speaking than English-speaking 3-, 4-, and 5-year-old children on two counting tasks, abstract counting and object counting. In abstract counting children counted aloud from 1, and in object counting they counted the number of items in small (3-6 items), medium (7-10 items), and large (14-17 items) arrays. Chinese children at all three age levels were more accurate on large sets in the object counting task, and were able to count higher in the abstract counting task than English-speaking American children.

Miller et al. (1995) suggested that this advantage is due at least in part to the Chinese number-naming system, which is more transparent than the English number-naming system. A comparison of Chinese and English number names illustrates the linguistic transparency of number names, particularly for the teen quantities in Chinese (Table 1). For example, in Chinese, whereas "ten-two" is clearly indicative of the quantitative composition of 12, in English, "twelve" is less so. Miller et al. argued that this difference in linguistic transparency may be a basis for higher performance in number-related skills in Chinese children. Consistent with this conclusion, differences in the patterns of errors made by Chinese and American children were most evident where the relative transparency of the Chinese number-naming system is greatest. That is, American children were less able than Chinese children to count accurately beyond 10. Similarly, in a longitudinal study on counting development, Miller, Smith, and Zhang (2004, cited in Miller, Kelly, & Zhou, 2005) found that both Chinese (living in China) and American 2-year-olds had difficulty counting up to 10, but between 3 and 4 years of age, the Chinese children learned to count in the teen quantities much more accurately than the American children. If transparency of the Chinese number-naming system were the cause, it would suggest that linguistic transparency could lead to or support changes in conceptual knowledge or representation (see Bloom, 2000, and Siegler, 1998, for further discussion).

Because the Chinese children in the Miller et al. (1995) study lived in China and the English-speaking children in the U.S., the effects of language could not be disentangled from effects of culture. To assess whether linguistic transparency and other aspects of culture may be responsible for the Chinese advantage in counting, we studied abstract and object counting in Chinese-English bilingual children. The bilingual children in this study were growing up in Canada, allowing us to look at possible language differences in a single country. These children were regularly exposed to both English and Chinese. If linguistic transparency of Chinese numbers affects counting, then these children should count higher in Chinese than in English. The Chinese advantage should be particularly obvious in numbers greater than 10 (as in Miller et al., 1995). If Chinese children benefit from the transparency of the number system and if their knowledge has become somewhat detached from the meaning of the words themselves, then knowledge of counting in one language might transfer to the other and performance on the two languages would be equivalently high. For example, French immersion students learning to read in French do not require any special instruction to learn to read in English; learning how to read transfers from one language to another (Genesee, 1987).