In their classic study of Basic Colour Terms (BCTs), Berlin and Kay (1969:6) identified four basic criteria of a BCT: i) it is monomorphemic, ii) its signification is not included in that of any other colour term, iii) its application must not be restricted to a narrow class of objects, iv) it must be psychologically salient to informants. Four subsidiary criteria were also suggested in those cases where the first four produced non-conclusive results. Although frequently debated at first, and slightly modified by Crawford (1982), it seems that the basic criteria have stood the test of time and are still considered helpful in the identification of BCTs.
The purpose of this paper is to describe a possible connection between basicness and the function of a colour term modifying a noun. Following Warren (1984) and several others, three functions of adjectives in attributive position in English can be distinguished: characterising, identifying and classifying. Identification and characterisation involve modification of an instance of a noun, thus indicating the colour of the object in the case of colour terms. In contrast to the other two, classification is more abstract as it involves modification of type, rather than instance. Its abstract quality allows colour terms to be used in extended senses (i.e. outside their normal area of designation). Examples of classifying use are phrases such as red onion and blue oak, where the colour terms do not refer to the colour of an individual onion or oak, but modify onion and oak as types, creating subtypes accordingly. Significantly, the colour of the leaves of a blue oak would not be called blue when judged in a colour array – i.e. out of context.
This type of extension can be accounted for if type modification is analysed as a kind of reference-point phenomenon (Langacker 1999), where it is the combined salience of the colour term and the colour category that is significant. In the process of extension, a salient reference point is used to access another less salient entity. Since the purpose of classifying use is to create a subtype, the emphasis is on distinction and contrast to the most common instances of the general type (in our examples onion and oak). In the case of blue oak, the weak bluish tinge of the leaves marks the distinction and motivates the name despite the green element being stronger. Thus, in naming the subtypes a salient colour term (= Basic Colour Term) is preferred even if the nuance could be considered a very poor example of the colour term.
This paper presents details of a study of English colour terms (Steinvall 2002), based on the Bank of English text corpus and the Oxford English Dictionary. Examples and figures are used to substantiate the theoretical models. The results show a correlation between the frequency of classifying use and the Berlin and Kay hierarchy, and it is therefore suggested that classifying use may serve as a further criterion for basicness.
Finally, the wider implications of the findings are discussed, and the results from previous studies, such as Conklin’s (1964) observations of Hanunóo colour categories, and Forbes’ of the use of brun and marron in French, are reviewed and reanalysed from the perspective of classifying use. In addition, a connection between classifying and figurative use is briefly explored.