You are using an old browser with security vulnerabilities and can not use the features of this website.

Here you will see how you can easily upgrade your browser.

Collective music making in an ancient form

Svan funeral singers in the village of Latali.

A particularly interesting subset in the new research corpus are eleven recordings (both headset-  and larynx microphone recordings) of three-voiced male funeral songs called Zar in Svan and Zari in Georgian. These are believed to represent one of the oldest forms of collective music making in Georgia  (Bolle-Zemp, 1997; Kalandadze-Makharadze, 2004). An interesting  personal account of the context of this ritual has been given by Brae (2011).

Musically, these songs are very special in that they essentially lack clear mnemonic anchor points (e. g. a pronounced rhythmical structure) which facilitates learning and memorisation of the music. This notwithstanding, these 3–5 minute long pieces are not improvised. Two different recordings of the same 5 min long Zar in Zargash performed by  the same singers for example,  turned out to be more or less identical. 

Svan funeral singers in the village of Latali.
Pitch tracks of Zari showing pitch shift.

What adds to the special character of this music is the fact, that in the majority of the recordings the singers increase the pitch more or less linearly as a function of time.  The steepest observed increase reaches 100 cents/min.   As a consequence, the used pitch inventory continuously changes over time, making classical transcriptions practically  impossible as it would require to continuously add accidentals, which would wrongly suggest (musically meaningless) key changes, just to accomodate the pitch shift. 

Additional challenges are provided by the „lyrics“ of Zar, which primarily consist of semantically apparently meaningless vowels. Since the Svan language has a much larger vowel inventory (18) than Georgian (5), the proper recognition of the lyrics and its documentation in a phonetic system simply from recordings is an enormous challenge. 

Furthermore, since these songs are essentially melody-free, their origin seems difficult to explain in the context of models which assume that polyphonic singing has developed from melodies which are subsequently harmonized. 

Pitch tracks of Zari showing pitch shift.

For analysis, we intend to use previously developed tools for the simultaneous visualisation of melodic and harmonic content of three-voiced songs (Scherbaum, 2016). Furthermore, we will search for patterns and structural regularities in and between the individual voices in funeral songs using signal processing tools to identify mutual interaction between voices. By doing so, we hope to gain some insights into the fundamental question regarding the historic development of Georgian polyphony.


Bolle-Zemp, S. (1997).  Mehrstimmige Wehrufe, Georgica, pp. 134–148.

Bray, M. (2011). Echoes of the ancestors Life , death and transition. Caduceus, (81), 6–9. (PDF)

Kalandadze-Makharadze, N. (2004). The funeral Zari in traditional male polyphony, in Proceedings of the 2nd Inter- national Symposium on Traditional Polyphony, Tbilisi, Georgia,  pp. 166–178. (PDF)

Scherbaum, F. (2016). On the benefit of larynx-microphone field recordings for the documentation and analysis of polyphonic vocal music, in Proceedings of the International Workshop on Folk Music Analysis, Dublin, Ireland,  pp. 80–87. (PDF)