...'The Vocabulary of Color with Particular Reference to Ancient Greek and Classical Latin,' by John Lyons", by Paul Kay, in The Language of Color in the Mediterranean pp. 76-90.
Kay denies L's claim that the BK system includes only "second-order" predicates, and includes mainly adjectives and verbs for color words from many languages:
[76 note 1]Kay asserts that individual senses of words may qualify as BCTs
[80 n. 10] "For example, in Somali the (first order) words for black, white, and red are intransitive verbs, the words for yellow and green are denominal adjectives and the word for blue (buluug), also a first order basic color term, is a noun. To say that something is blue, one has to say that it 'has' blue (Maffi 1990b)."
[82-3] Kay also says that the monosemy of words like khlo^ros (ancient Greek) and latuy (Hanuno/o), including simultaneously meanings of color along with meanings of freshness, is not proven. [In fact, LSJ lists sand as a predicate of khlo^ros, indicating that it does in fact have an independent color sense. In addition, this kind of association of senses is present in Latin viridis, but Pliny has plenty of examples where the term is purely chromatic.]
Kay cites evidence that the perception of the Hering primaries as cognitive focal colors is not restricted to humans, but is also to the great apes and old world primates, quoting Sandell et al. 1979, "Color categories in macaques," Journal of Comparative and Physiological Psychology 93:626-635, pp. 628 and 634.
Finally, he argues that the thesis of L. that luminosity rather than hue was more important to Ancient Greek speakers is not necessarily incongruent with the BK thesis; it may be that this could be considered a pre-BK stage of color development.