Another Baltic country, Latvia, could be considered the most polyphonic among the Baltic States. Although Latvian polyphonic traditions are not as internationally known and as dazzlingly unique as the Lithuanian canonic polytonal sutartines style, and although there is no such variety of polyphonic styles as in Lithuania, the Latvian tradition of polyphonic singing covers most of the ethnic territory of the Latvian state. The only region where no polyphonic recordings have been made is the Northeast part of the country. In the western part of Latvia the tradition of polyphonic singing is still well alive (as in the regions of Nica, Barta, Alsunga).
Interestingly, virtually the only type of polyphony recorded in the territory of Latvia (according to the works of Latvian ethnomusicologists) is drone polyphony. Written sources mention the tradition of drone polyphony in Latvia from the 16th and 17th centuries. The drone is mostly pedal, but there are instances of the rhythmic drone as well. There are different terms for the drone performer in Latvia: vilceja (“the one who drags”), duceja (“the one who gives a low, continuous droning sound”) and ruceja (“a grumbler, the one who murmurs”). The drone changes its pitch and moves always a major second up. The main melody is always sung by a solo performer, and the range of the main melody is very small (usually a third). The drone is always performed by a group of singers. Rhythmically Latvian drone polyphony is based on a simple duple metre (2/4). Two-part singing dominates:

A. Yurian documented a fascinating tradition of drone three-part singing in Latvia in the 1890s. This tradition of three-part singing was documented at two places – in central-Southeastern and Southwestern regions of Latvia. (I am very grateful to Latvian ethnomusicologist Villis Bendorf for making archive transcriptions of A. Yurian available to me while I was in Latvia in November 1988.) Some of these recordings were published in 1907. A fascinating tradition of drone three-part singing is represented on these recordings (see the last example from Latvia). The pedal drone is in the middle, the main melody (solo performer) sings a small-range (within a third) melody, and the third part mostly sings under the drone. The third part is also performed by a solo singer. It is very interesting that the third part is also based on a small-range melody and mostly sings (recites) on a major second below the pedal drone, thus creating plenty of sharp secondal dissonances with the drone. (I do not want to go into “comparative” discourses here, before the second part of the book, but the parallels between three-part Latvian drone polyphony and some three-part Balkan dissonant styles are quite obvious.) This tendency to create secondal dissonances is particularly evident at the very end of the musical phrases. Look at the second last note: there are the following notes together: “D” in the pedal drone, “C” in the lowest part, and “E” in the main melody. Therefore, at this moment there are two seconds sounding together (C-D and D-E). Most importantly, this sharp dissonance is not a result of a random coincidence of free melodic parts. The fact that this sharp dissonance has a special “fermata” sign on top [a semicircle with the dot inside] at this very moment means that singers were consciously trying to achieve this sharp dissonant harmony and held it longer [“fermata” means that these notes must be sung considerably longer, “drawn out”]:
As I have mentioned before, Latvian ethnomusicologist Martin Boiko researched and found some interesting elements of sutartines-style singing in Latvia as well, although no songs of the unique “secondal polytonal sutartines” have been found in Latvia.