Several scholars have suggested that in ancient Greek there was a low boundary tone at the end of a relatively small prosodic constituent such as a clitic group or maximal prosodic word. The boundary tone may phonologically motivate some puzzling pitch-accentual phenomena in the language. One is the diachronic pitch-peak retraction that led to the circumflex pitch accent (HL) on penultimate syllables (the “sōtêra rule”). Another is the intonational phrase-internal downstepping or deletion of a word-final acute accent (H); that conversion of an acute to a grave accent is known as “lulling” or “koímēsis”. If such a low boundary tone existed, its effects should still be audible in ancient Greek non-strophic vocal music, where there is a significant correlation between the pitch movement of the text and the movement of the melody to which it is set, i.e. between tone and tune. Specifically, proponents of such a low boundary tone would predict that the turning point between falling and rising melody, the “musical trough”, should center around the word-final mora or syllable. The present study provides the first full description of troughs in the Delphic Hymns and finds that they are indeed closely aligned with word-end. Furthermore, once other factors that could lead to word-final troughs are set aside, i.e. once potential confounds are controlled for, the association of the trough with word-end remains strong, suggesting that we should in fact reconstruct the low boundary tone.

Document Type


Publication Date


Publisher Statement

Copyright © 2023, Koninklijke Brill NV.


The definitive version is available at: