From Yellow to Blue — or Not?

Abstract

In Indo-European languages, the reflexes of PIE root *ĝhel- are typically used as colour terms for ‘yellow’ or to denote yellow objects like gold. In Slavic languages there are no less than three different reflexes (e.g. Russian желтый, зеленый and голубой). While the original root is traditionally thought to have had the primary meaning ‘yellow’, there is nothing unusual in the fact that the root often acquires the meaning ‘green’, as PIE was almost certainly a language where green and yellow were not distinguished on the level of basic color terms. The fact that some reflexes expanded into the blue part of the spectrum, although it has a parallel in another PIE root (Lat. flauus ‘yellow’ vs PGmc *blēwa- ‘blue’), seems rather interesting. A similar semantic transition of *ĝhel- can be seen in Celtic languages (e.g. OIr gel and glas). But while in Celtic there could have been two reflexes of the same root, one of which stayed in place and the other drifted away as Proto-Celtic evolved into a Stage IV language, the Balto-Slavic word cannot be so easily explained away: both голубой and its Baltic cognates (Lith. gelumbe ‘blue cloth’, OPruss. golimban ‘blue’), unlike the words for ‘green’, ‘yellow’ and ‘gold’ in the same languages, have retained the unpalatalized *gh-. While by no means a borrowing from Lat. columba, the Balto-Slavic lexeme does share the word-formation with columba and Grk. κóλυμβος ‘little grebe’ – the words that are traditionally connected with the cluster of Lat. calidus ‘with spots’, Grk. κηλίς, OIr caile ‘a spot’, OInd. kāla- ‘(blue-)black’ etc. Here a different version is proposed: neither are columba and κóλυμβος connected with the aforementioned ‘black spots’ cluster, nor is голубой connected with PIE *ĝhel-; they represent a separate and possibly non-Indo-European group of cognates.