Carlos Quiles, Fernando López-Menchero
(First draft published in October 2017)
The evolution of the velar system in the attested Indo-European dialects gave rise initially to the three-dorsal theory, which was immediately—and has been since then—rejected by an important part of Indo-Europeanists.
Nevertheless, this artificial reconstruction, based on the centum-satem distinction, remains a prevalent hallmark of the most common handbooks on Indo-European linguistics used in university courses around the world.
In this paper we examine the reasons in favour of a two-dorsal system and against the reconstruction of a series of palatalised velars, illustrating it with the history of the development of both theories, highlighting the weak finds that seem to be the strongest link to an original system of three velars.