About the tonal organization of traditional Georgian vocal music

Lecture by Prof. Frank Scherbaum

The specificity of the sound scales of traditional Georgian vocal music has been debated for decades, however, progress in research is limited due to their small scope and insufficient documentation of the datasets used in earlier analyses.

During the speech, the results of two recent computational studies will be presented, which address this issue through the analysis of two complementary corpus: (1) Erkomaishvili's collection — recordings from 1966 made at the Tbilisi State Conservatory, documenting the singing master Artem Erkomaishvili, who independently performed all three vocal parts of more than one hundred liturgical songs. covering more tracks; and (2) the 2016 GVM collection — a newly created corpus of high-quality multi-channel field recordings (video, head and larynx microphones) of traditional Georgian singing, collected during a research expedition in 2016 (together with Nana Mzhavanadze), focused on the Upper Svaneti region and supplemented with recordings from Guria, Racha and Tbilisi. The juxtaposition of historical solo recordings with the contemporary corpus of ensemble recordings raises the question of whether the conclusions about the tuning of Georgian music are universal and whether they remain valid regardless of time, region and performance context.

Both studies show that the untempered, uneven tuning based on pure quartets and quintes and neutral thirds and sexts is not a characteristic of only one historical singer, but is a lasting and describable property of traditional Georgian vocal music. This phenomenon can be observed in singers, in various bands and regions, over a period of about half a century. The formation of melodies is based on flexible pitches of scale sounds, dictated by the requirements of harmony, and not on rigidly defined intervals. The way melodies are performed is also influenced by deliberate pitch modulations and — in Swanian mourning songs (Zari) — the gradual, coordinated shifting of pitch of sounds.

Both studies are descriptive, not normative, and together demonstrate the possibility of applying a general, repeatable process of computational analysis — from interactive scale and interval modeling to pitch drift correction based on interval filtering. This method can also be used in the study of other untempered vocal traditions. At the same time, research points to limitations associated with writing this type of music using Western five-line notation.