ABSTRACT
Multilevel models are popular these days. Software engineering, semantics, organization studies, education, sociology, and psychology are among the branches of science with a keen interest in these models. Let us give an example. In the course of development children acquire common words for colors available in their language. At a young age children sharing the same language do not all know the same number of color words; there are individual differences. There are also cultural (linguistic) differences in the number of words known to young children. Each language has basic color terms, but the cross-language variability is considerable. Some languages have more terms than others; a given term may refer to a part of the visible spectrum for which there is more than one term in some other language. For example, many languages have a single word for the colors designated with blue and green in English (Berlin & Kay, 1969). It follows that the number of color words known by a child has an individual component and a language component. Differences between children within a language group will have to do with age and cognitive competence. Differences between children not speaking the same language are likely to be a function of the number of color words available in their respective languages. Thus, a cross-cultural study of children’s color vocabulary requires different explanations for individual and cultural differences. Individual differences may be related to age, intelligence, and socioeconomic status, while the explanation of cultural differences would require a model of the emergence of color terms in various languages. As a consequence, differences between individuals within cultures should not be interpreted in the same way as differences between children with different mother tongues: The analysis of individual and cultural variance in the number of used color terms requires a multilevel model.
