Search

When a free-listing task is used to elicit verbal concepts from a given semantic domain, it provides two indicators of the salience of each word for that linguistic community. These are the proportion of the subjects who include a word in their lists, and its average ranking priority position across the lists. The data also contain cues about the cognitive representation of the semantic domain, and in particular about the conceptual closeness among words. Closely associated words tend to prime each other and to appear in the lists in close succession. Clusters of mutually associated terms can be recognised, listed in one another's company, although with different priority for different subjects. We applied this approach to the domain of colour terms, converting lists for fourteen European languages into matrices of inter-term similarity, for analysis with multidimensional scaling (MDS) and hierarchical clustering. Two-dimensional MDS solutions or 'maps' were typically required to reflect two competing criteria by which terms were sequenced. Speakers of each language tended to follow a salience gradient, but also made separate clusters of fully-chromatic concepts - colour terms in strictu sensu - and unsaturated or desaturated concepts defined primarily by lightness rather than by hue. This and other features recurred across the languages despite their geographical and phylogenetic diversity, as cross-cultural universals in colour language, in addition to the well-known regularities governing basic colour terms and the stages of colour-lexicon development. [ABSTRACT FROM AUTHOR]