Walo Paul, 'Saying "green" in yakoma ? On the discourse determination of colour terminology'.
One difficulty with an etnolinguistics enquiry about colour naming is setting up the semantics status of certain labellings. Thus the compound ngu-ngunza (water-manioc leaf) could be taken as a colour-term ("green") once a closed terminology -of which it would be one element- be established. Contrarywise, one can argue such a syntagm is produced by the very conditions of the ethnological survey (identifying a colour sample). A more detailed analyzis of the during-the-enquiry discourse shows both the instability of the motivated semantic unity, or contrarywise, aiming at a virtually closed terminology, can be manipulated during the enquiry and depend upon categorization cognitive operations.