Laryngeal theory

node

The proposal that Proto-Indo-European contained a small set of consonants — conventionally written h1, h2 and h3 — which had disappeared from every language known when the theory was framed, leaving behind only their effects on neighbouring vowels. Ferdinand de Saussure argued in his Mémoire of 1879 that the vowel alternations of the reconstructed language became regular if one posited abstract elements, which he called coefficients sonantiques, in specific positions. It was a purely internal argument: the entities were required by the system, and no language attested any of them. That is precisely what made the theory falsifiable in a way linguistic reconstructions rarely are, and it is why the decipherment of Hittite is treated as the discipline's decisive test. Hittite, written down in the second millennium BCE and read only from 1915, turned out to preserve a consonant conventionally transcribed h in a large share of the places Saussure's system demanded one. Jerzy Kuryłowicz identified the correspondence in 1927. The number of laryngeals, their phonetic realisation and their behaviour in individual environments are all still argued; that they existed is not.

A node is not a place. Drawing it on a map would assert something about the world that no stored fact supports.