Analytical Approaches to World Musics
ISSN 2158-5296
Bernacki 2025
AAWM Journal 13/No. 1 (2025)
Embodied Metric Transformation in the Greek-Macedonian Dance-Song Daoular Tsalar
Nathan Bernacki
Non-isochronous Meter, Embodiment, Balkan, Dance, Entrainment
Choreomusical approaches to timing and coordination provide the opportunity to analyze the ways in which proprioceptive, haptic, and visual modalities interact and inform the production of sonic phenomena. The Greek-Macedonian vareis horoi (heavy dances) of the zourna/daouli tradition offer intriguing insight into the dynamics between movement and sound in musical performance, as choreographic events in this style are central to sonic onset synchronization and the long-term coordination of performance structure. To analyze how musical time is structured, coordinated, and embodied in this style, I first describe its general metric progression and how this form is realized in a specific dance-song Daoular Tsalar. I then demonstrate the choreography to a particular recording of this song and present a beat-level interonset interval (IOI) analysis of its sonic onsets. I contextualize these quantitative findings in the dynamics of live performance and conclude with implications for theories of entrainment through a critique of Clayton et al.’s (2020) Theory of Interpersonal Musical Entrainment.
Nathan Bernacki is a PhD Candidate in Ethnomusicology at the University of British Columbia.
Click for DOI, citation, PDF version.
Introduction
Choreomusical approaches to timing and coordination provide the opportunity to analyze the ways in which proprioceptive, haptic, and visual modalities interact and inform the production of sonic phenomena. The Greek-Macedonian vareis horoi (heavy dances) of the zourna/daouli tradition offer intriguing insight into the dynamics between movement and sound in musical performance, as choreographic events in this style are central to sonic onset synchronization and the long-term coordination of performance structure. To analyze how musical time is structured, coordinated, and embodied in this style, I first describe its general metric progression and how this form is realized in a specific dance-song Daoular Tsalar. I then demonstrate the choreography to a particular recording of this song and present a beat-level interonset interval (IOI) analysis of its sonic onsets. I contextualize these quantitative findings in the dynamics of live performance and conclude with implications for theories of entrainment through a critique of Clayton et al.’s (2020) Theory of Interpersonal Musical Entrainment.
General Characteristics
[2] The defining feature of the zourna/daouli tradition is its instrumentation, which involves a daouli (double-sided drum) hit with a bass beater on one side and switch stick on the other, and at least two zournades (double-reed oboes). Usually associated with Roma, this tradition is present in many life-cycle rituals, religious festivals, and celebrations in both Romani and non-Romani settings throughout the Balkan Peninsula. Most dances are performed in a line or open circle moving counterclockwise, with the right-most dancer as the lead-dancer.[1] The instrumentalists in this tradition frequently stand within the circle of dancers facing the lead dancer in order to execute specific dancer-musician interactions that characterize performances of vareis horoi. This performance interaction necessitates centering the relationship between dancer and musician in order to approach the fundamental components involved in the temporal coordination of this style. Therefore, I use the term beat to describe a frame of time in which a choreographic event can occur, or in other words, when dancers can move.
[3] Performances of vareis horoi often feature a telescoping metric form and suggest a hearing involving three sections. Section 1 can be characterized as a slow and loose treatment of the metric cycle.[2] In this section, sonic onsets are coordinated and determined by when the lead dancer moves, engaging in a visual coordination of onsets between the lead dancer and musicians. The supporting dancers also watch the lead dancer to time their steps, although the haptics involved in the dance line’s physical connection and proprioceptive considerations of each movement’s feeling also contribute towards this multimodal timing coordination. Section 2 functions as a transition period in which the previously wide range of acceptable onset timings gradually narrows as cycle durations decrease.[3] Concurrently, the mode of coordination gradually transfers from visual to purely aural, as the musicians become the locus for temporal coordination. In this process, the dance’s visual cues cease to coordinate beats and are replaced with a strict timing framework governed by the musicians’ sonic onsets. After this transition comes Section 3, which is defined by a fast and strict treatment of the metric cycle and a relatively short cycle length. These fixed onset timings and shortened cycle duration allow sonic and choreographic onsets to be coordinated aurally. In other words, all dancers move to the music, unlike in Section 1 where the music is timed according to the movements of the lead dancer. The performance then ends with a signal from the lead dancer.
[4] To analyze this metric transformation, I focus on the dance-song Daoular Tsalar (“The Daouli Strikes”) as performed in the city of Irakleia, in Serres, Greek-Macedonia. This particular song traditionally occurs in the beginning stages of a glendi (celebration) and therefore features a notably long transition period from slow to fast in order to carefully lift celebratory energy to an appropriate introductory level (Papakostas 2023). The recording analyzed here is from a CD compiled by Greek dance teacher and scholar Christos Papakostas, featuring Kostas and Ilias Matziris on zourna, and Ilias Diamanos on daouli (Audio Example 1).
[5] The choice of this audio recording rather than a video of a live performance was made for several reasons. First, one objective of this analysis is to quantitatively delineate the gradation between Section 1 and 3. This recording has a particularly long Section 2 and therefore is an ideal example of how each beat category adjusts throughout the course of performance. In publicly available videos of staged performances, Section 2 of many vareis horoi is virtually non-existent or only takes place over only a few cycles. This is because, according to Papakostas, it is the nature of many performance groups to shift from the slow and difficult Section 1 immediately to the more lively Section 3, skipping over a drawn-out Section 2 (Papakostas 2023). Cutting Section 2 also gives the lead zourna player time to perform a taksim (non-metric modal improvisation) during Section 3 within the time constraints and aesthetic preferences of a staged performance setting. The only publicly available non-staged video performance of Daoular Tsalar was made by Papakostas himself, but the visuals are cut in the middle of the performance and the audio is often inundated with the sound of arapides (mummers with bells).[4]
[6] Second, Audio Example 1 features a trio of master musicians who know the general timing expectations of slow sections through decades of performance experience. This is a necessary part of professional musicianship, as musicians may have to lead the timing of movements even in Section 1 depending on the experience of the lead dancer (a point that I will return to later). Third, this recording’s studio-like quality facilitates the marking of IOIs, as the waveforms are much clearer than those of a situated performance such as Video Example 2. Last, through my personal connections with Papakostas I can trust that this recording was made consensually, for the purpose of education and scholarship, and with the musicians’ best interest in mind.[5]
Figure 1. General timbral cycle of Section 1.
[7] Section 1 has five beats with corresponding dance movements. The fifth is left “silent” in this recording, although an articulation of this fifth beat by the daouli player is possible. In Section 1 of this recording there is a sequence of five strokes per cycle, represented in Figure 1 onomatopoeically with the stroke dum referring to a strike with the bass stick and tak referring to a strike with the switch. The timbral cyclicity of this basic stroke pattern matches exactly with the series of weighted and non-weighted choreographic events. Here, “weighted” refers to a transfer of weight between feet, or when weight is emphasized through a bounce on a single leg. Similarly, a “non-weighted” movement would be a gesture that does not involve weight transfer or emphasis, as with a mere lift of the leg. In Section 1, each dum corresponds with a weighted movement, and the two switch hits match one non-weighted lift of the leg. There is a final anacrustic movement at the end of the cycle which again is not articulated sonically by the daouli in this case.
[8] With these considerations in mind, the basic stroke template of Section 1 can be described with a bass hit marking Beat 1, the first of two switch hits marking Beat 2, single bass hits on Beats 3 and 4, and no sonic articulation on Beat 5. Notably, the second of these two taks seem to articulate some sort of subdivision layer. However, while this beat’s second attack technically subdivides the beat, it functions more as a gestural cluster reflecting the corresponding dance movement. In Section 2, progressive timing adjustments towards a 2:3 beat ratio are made to each of the five beat categories while cycle length gradually decreases.
[9] This process gradually forms the beat proportions of Section 3, which vary between recordings and regions. Those of this recording can be described as containing five non-isochronous beats (short-long-long-short-short) with an approximate 2:3 ratio between short and long beats (see Figure 2).
[10] These beat proportions are common in performances of Daoular Tsalar from the area of Irakleia where these musicians are from, but can differ in other regions of Greek-Macedonia. For example, while some timing patterns resemble what I describe as 12/8 (2+3+3+2+2), others feature a 7/8 (1+2+2+2, Video Example 3) or (3+2+2, Video Example 4) subdivision grouping.[6] Nevertheless, as long as beats are marked consistently between cycles, the metric characterization of Section 3 should not adversely affect the overall demonstration of how the beat cycle transforms over the course of a performance. Furthermore, the nearly identical series of dance steps found in most performances of Daoular Tsalar underscores the unity between the slight regional differences in timing patterns (Graziosi 2023).
Figure 2. Visualization of Section 3’s meter.
Section Dance Part
Introductory Taksim N/A
Section 1 Part 1
Section 2 Part 1 (gradually faster)
Section 3 Part 2
Figure 3. Correspondence between sections and dance parts.
[11] The dance is three cycles in duration, and I refer to each of these individual cycles as a dance cycle.[7] The first two out of three dance cycles travel to the right, and the final one travels to the left. This dance sequence progresses through two closely related parts throughout the performance (Figure 3). Part 1 of the dance begins after an introductory non-metric modal improvisation (taksim) and continues throughout Sections 1 and 2. Part 2 begins at the start of Section 3 and is a slight variation of Part 1, gesturally adjusted for a faster tempo. To give a sense of what the choreography to this particular recording would be, Video Example 5 is a demonstration of the dance with beat, cycle, dance cycle, and section counters in the lower left-hand corner.[8] This video example is meant to depict only the dance movements and not the intricate processes of coordination that occur in live performance. These choreographic and metric observations can also be used to inform a quantitative approach such as IOI analysis to visualize this metric transformation.
Procedure, Compartmentalization, and Data
[12] IOIs in each section were marked according to the daouli strokes that correspond with choreographic onsets. The 58 contiguous cycles (243 IOIs) measured include all 23 cycles of Section 1, all 14 cycles of Section 2, and the first 21 cycles of Section 3. These were marked first by tapping, followed by two rounds of manual revision adjusting each IOI to its corresponding stroke. Figure 4 shows the durations of beat IOIs grouped into their respective cycles with dashed lines marking the boundaries of the three sections. It should be noted that the hard boundaries drawn between these sections are blurred rather than distinct, and are used here purely for the convenience of analysis. I discuss the data of Sections 1 and 3 before that of the transitional Section 2.
Figure 4. IOIs of beat durations.
[13] As mentioned previously, only four out of the five beats in the cycle are marked sonically by the daouli in Section 1, and therefore only four beat IOIs comprise this section’s cycles. This section features an average beat proportion of 12:26:21:41 at an average cycle duration of 5.88 seconds.[9] While the standard deviations of each beat category are expectantly large, this data still illustrates a general proportional structure between beat durations. The first beat is the shortest, the second beat is longer than the first, the third beat is slightly shorter than the second but longer than the first, and the fourth beat is the longest. These trends are consistent throughout every cycle of Section 1. Even though these beats are timed visually, this finding supports the notion that there still exist expectational ranges for each beat category’s onsets.
[14] The concept of a timing range as an organizing metrical factor has been proposed in other work based in adjacent Balkan musical traditions. In Daniel Goldberg’s dissertation on timing in Bulgarian tŭpan performance, he proposes that “meter does not dictate a single point in time for a given musical event, but rather sets a timing range during which a musician may play that event,” drawing conceptual similarities with Stover’s “beat spans” (2009), Danielsen’s “beat bins” (2010), and Johansson’s “rhythmic tolerance” (2012, 2017) (Goldberg 2017, 196–199).[10] This redefinition finds support within the present analysis, except in this case, a visual mode of coordination allows the temporal range for each beat category in Section 1 to be much larger in comparison with these previous studies. This visual mode of coordination also allows for non-isochrony to occur on the cycle, beat, and subdivision level. This finding is similar to that of Johansson’s (2017) timing analysis of Swedish Polska in which these non-isochronous levels are analyzed according to their choreographic and melodic contexts. Despite featuring differing explanations for their non-isochronous nature, the timing analysis of both Swedish Polska and vareis horoi call into question the necessity of isochrony on any metric level to coordinate onset timings.
[15] Jumping ahead, Section 3 begins with cycle 38, where cycle durations begin to stabilize after their gradual contraction in Section 2. Section 3 features an average beat proportion of 17:25:25:16:17 at 1.97 seconds per cycle.[11] As mentioned earlier, the meter is a five-beat short-long-long-short-short cycle with a 2:3 ratio between beat types. Distinct to this section is the articulation of an isochronous subdivision layer by the daouli player. However, the function of this isochronous stream is ambiguous, as the zourna melody does not always lock in with this isochronous subdivision level. Depending on how the lead zourna player ornaments the melody, a range of melodic subdivision timings are possible that do not always adhere to the isochronous timing of the daouli player. Similarly, the choreography of Parts 1 and 2 do not articulate the subdivision level, with the gestures of both moving to the non-isochronous series of beats. What the various subdivision timings of the daouli and zourna have in common is that they all reference or align at the choreographic beat level throughout the entire performance. This correspondence may suggest that even in Section 3, participants use the beat level as the reference point for metric coordination, even when a layer of seemingly isochronous subdivisions is referenced by the daouli.
[16] The gradual transition defining Section 2 allows for the loose timing of Section 1 to transform into the strict timing of Section 3 without a break in performance. Section 2 begins with cycle 24, where cycle length noticeably decreases after the relatively long and loose cycling of Section 1. In this transition, each cycle is more contracted than the previous (with one exception) and this gradual process creates a clear sense of acceleration. As mentioned previously, this section features incremental adjustments of beat proportion and cycle duration towards the strict proportions of Section 3’s five-beat meter.
Figure 5. Beat proportions underlain with 12 subdivisions.
[17] To visualize this transitional process in terms of beat proportion, Figure 5 shows how each beat category adjusts from Section 1’s loose and gestural timing towards the strict beat proportions reached in Section 3. This depiction is aided by laying equally spaced subdivisions under the proportion data of all cycles to indicate the timing expectations of Section 3’s 2:3 beat ratio. In taking the average beat proportions for Sections 1 and 3 as reference, the proportion of Beat 1 increases by 5%, Beat 2 decreases by 1%, Beat 3 increases by 3% throughout this transition. Although the sonic proportions for Beats 4 and 5 are not known in Section 1, if these two beats are combined in both sections, the beat proportion of this macro-category decreases by 4%.
[18] However, the takeaway from these exact adjustment percentages is unclear. Firstly, the sense of proportion between Section 1 and Section 3 is heavily influenced by cycle duration, secondly, these adjustments will differ drastically from performance to performance, and lastly, beat proportions may be heavily influenced by the interdependency created through proportioning, especially in the case of Section 1.[12] Instead, what is perhaps more intriguing are the macro-level similarities between proportions that are maintained throughout the entire performance despite these adjustments. For example, the duration of Beat 1 in each cycle of the performance is much shorter than that of Beat 2, with Beat 2 and Beat 3 remaining relatively equal and still longer than Beat 1. It is difficult to determine how the fourth and fifth beat adjust throughout the performance in this recording, but speculatively, if grouped together they would make the longest beat duration in the cycle. These trends suggest that in this style the loose timing expectations of Sections 1 and 2 may in some way correspond to the more strict proportional structure of Section 3, a relationship that would require further confirmation by timing analyses of many other vareis horoi. Nevertheless, through these visualizations, incremental changes in beat proportion and cycle duration can be seen throughout Section 2, and through this transformation, it is clear how Section 1 and 3 are connected yet distinct with regard to timing.
[19] Before proceeding with a review of descriptions of varies horoi in ethnographic literature, it is worth reviewing the various revisions I have made throughout the course of this analysis to conceptions of meter as a hierarchical structure of points in time with at least two levels (London 2012; Clayton et al. 2020). These amendments include reorienting the beat level rather than the subdivision level as the primary organizing referent, reconceptualizing beats through ranges rather than points in time, and (for Section 1) accepting that non-isochronous timing can occur simultaneously on every metric level.[13] Connecting these points is an emphasis on gesture as an integral component in timing coordination, suggesting that expanding theories of meter into different modalities may bring differing priorities regarding temporal organization. To contextualize these findings further, I will now discuss elements of performance dynamics that may affect timing in this metric transformation.
Ethnographic Descriptions and “Expanded Form”
[20] Descriptions of the musician-dancer interaction featured in vareis horoi have appeared in various literature on Balkan music and dance.[14] Before reviewing these cases, it is essential to reiterate the transnationality of this style, as the research cited below draws from the nation-states of Bulgaria, North Macedonia, and Greece. Not only are “heavy dances” characteristic of the transnational Macedonian region that straddles these countries, the Romani musicians who perform them work within a similarly transnational network that involves knowledge of both broad aesthetic similarities and localized performance particularities. Not only do musicians have to know the general slow to fast metric progression of heavy dances, they also must be familiar with the preferred repertoire and correct sequencing of songs for a variety of occasions across various communities.[15] Given the presence of the heavy dance style across national boundaries and the translocal professionalism of Romani musicians, the ethnographic descriptions below reference a variety of regions in order to reflect broader musical and choreographic similarities.
[21] In an English translation of Dimov and Peycheva’s work The Zurna Tradition in Southwestern Bulgaria: Romani Musicians in Practice (2002) the interaction between musician and dancer is described as such:
After guessing the song wished by the audience, the zurnacies play it with regard to the horo dancers. The leader, or basha, being the best dancer, governs the step, rhythm, and music. This is why the maestro zurnaci walks close to him, watches, and follows his movements, changes the tempo and tact if necessary, and improvises, reacting immediately to his dancing.[16] (340)
[22] In commenting on the unity between the slow and fast sections of heavy dances, they go on to say that:
[T]here is always a slow part in the beginning but it is gradually into a fast one…They are always united in a whole of content…The Rusalia Plays, typical for the zurnaci repertoire and formed by several different ritual melodies, always start with a slow part and end up with a fast one. (349).[17]
[23] There is also evidence that zourna players verbally mark slow and fast sections, and this dancer-musician interaction can only occur during the slower part of the performance:
The most esteemed performers have the ability to improvise and predict the dancer’s steps, and this moment is cherished by the performers. In this quote the informant mentioned this occurs “‘cause the horos are slow…”.[18] (375)
[24] As this tradition is performed throughout the transnational cultural region of Macedonia and its surrounding areas, much of Dimov and Peycheva’s Bulgaria-centered work can be applied towards understanding the zourna/daouli traditions in Greek Macedonia of which “Daoular Tsalar” is a part.[19]
An account of heavy dance performance in North Macedonia is found in Robert Liebman’s dissertation on Balkan folk dancing:
…in the men’s heavy dances in Lazaropole (western Macedonia), after the dancing has progressed from the slow, heavy part at the beginning to the faster, more spirited part in which the leader may perform turns and squats…Depending upon their mood, these other leaders can do the entire dance again with all of the gradual increase from the slower part to the faster or they may just continue at the higher speed and energy. Likewise, depending on their mood, they may end the dance at its most feverish pitch or they may gradually reduce it and return to the calm which they had at the beginning. (1994, 190–191)
[25] Balkan dance instructor Richard Crum’s entry on heavy dances in the Oxford Encyclopedia of Dance, similarly describes them as:
A typical dance of the tesko type is performed exclusively by men. It begins slowly but then speeds up; the musicians follow the cues of the lead dancer, who sets the tempo and dynamics while the drummer matches his beats to the steps and pauses he sees performed. The music defies conventional notation.[20] (1998)
[26] In commenting generally about Macedonian folk music, Alice Singer (1974) observes that “Villagers dance ‘to’ the music, but musicians also play to the dancers, adjusting the tempo and sometimes the rhythm to the actions of the leader,” (383) noting that this progression also occurs outside of the zourna/daouli tradition.
[27] These descriptions not only point towards the difference between slow and fast sections in a performance, one noted by music scholars, dancers, and musicians, but also towards the connection and graduality linking this duality. This seamless change is underpinned by a multimodal connection between the choreographic and the sonic, an interdependence that arises as another major theme from these accounts. Furthermore, an interview with Greek dance instructor, daouli player, and scholar Christos Papakostas suggested that situating these ethnographic and analytical facets within the social context of vareis horoi can lead to insights about how to conceive of musical time in this style.
[28] Musicians accompanying a lead dancer consider a range of factors that inform Section 1’s onset-to-onset interaction. Papakostas stressed the importance of playing for a specific lead dancer, considering their place in the community, age, skill, and gestural style, as to properly reflect the timing and spirit of their movements in the musicians’ own playing (Papakostas 2023). These attributes have an effect on performance dynamics. The lead dancer does not always determine each beat onset of Section 1, as in some cases this section can be musician-led or performed as a negotiation (Papakostas 2023). For example, a younger, inexperienced dancer may be led at times by the musicians, whereas an older dancer familiar with the style may be completely in charge of the pace. In any case, a sonic aesthetic of gestural looseness is always present. For musicians, this relationship with various lead dancers involves drawing upon years of performance experience as well as an intimate cultural knowledge of various communities.
[29] Papakostas described the collective quality of varies horoi through the concept of methexis, a term originating in ancient Greek theater referring to the extensive use of audience improvisation and participation in the execution of a scene (Papakostas 2023). Methexis in this context emphasizes how each musical participant—the musicians, the lead dancer, and the supporting dancers—contributes in different ways and at different moments to a successful performance, one that ultimately invokes kefi, a particular sense of joy or ecstasy. In vareis horoi, this process involves a communal balancing of slow to fast, loose to strict, and visual to aural progressions, with unifying choreographic, metric, and melodic features.
[30] Through approaching this style as community-based and unified across timing changes, perhaps Section 1 could be conceived as a “gesturally expanded” form of Section 3. This suggestion is supported by other performance factors. First, Section 3 is the most participatory part of the performance, in the sense that everyone in the dance line can entrain to the music on their own, without depending on the lead dancer. This communality is also reflected in the onset regularity afforded by an aural mode of coordination. Second, due to the strict treatment of Section 3’s metric cycle, the feeling of this section’s beat proportions are already known to participants throughout the preceding sections. This stands in contrast with the beat timings of Section 1, which are ultimately determined by the lead dancer, their improvisational abilities, and overall gestural style. Finally, in considering a broader Balkan dance context, most traditional dances are performed in the type of aurally coordinated meter found in Section 3. Even within the zourna/daouli tradition there are many dances that do not include the characteristic loose to strict progression found in vareis horoi. The idea of “gestural expansion” accounts for these communal and holistic considerations even across significant timing changes, and has intriguing implications for the role of movement in current models of entrainment.
Implications for Theories of Entrainment
[30] A primary concern in studies of entrainment is assessing how participants coordinate and structure the temporality of their musical phenomena. In the present case study, this interaction involves a metric framework, a choreographic cycle, and shifting modes of coordination, all of which feature gesture as a common and integral link. This emphasis on embodiment in the performance and coordination of vareis horoi provides an opportunity to assess the role of movement and visual information in current models of entrainment. Though the role of gesture (Davidson 1993; Williamon and Davidson 2002; Clayton 2007a) and visual cues (Clayton 2007b; Bishop and Goebl 2017) has a history within musical entrainment studies, the designation of “performer” is often restricted to musicians, and therefore the agency of dancers’ and other musical participants’ movements in relation to the coordination of musical time has received little attention.[21] Similarly, entrainment studies that do include dance often study dancers as recipients that respond to sonic signal rather than musical actors with agency in co-creating musical temporality.[22] As Clayton et al.’s (2020) theory of interpersonal entrainment stands as an overview and synthesis of the literature at large, it is an ideal case to investigate exactly how movement is defined within entrainment research more broadly.
[31] In this model, “Musicians make use of two broad categories of movement when performing. Sound-producing movements (e.g., key presses, bow strokes) occur over the same timescales as auditory onsets. Sound-facilitating, or ancillary, movements (e.g., head nods, body sway) are movements that do not play a direct role in the production of sound and typically occur over longer timescales than sound-producing movements.” (Clayton et. al. 2020, 145).
[32] In short-term coordination, the fine motor actions responsible for the production of zourna and daouli onsets fit the description of sound-producing movements that “occur over the same timescales as auditory onsets,” (Clayton et. al. 2020, 145). However, the continuous visual mode of coordination in Section 1 is an essential component to the timing of these sound-producing movements. This visual and gestural dynamic therefore suggests a type of movement that is non-sound producing, yet integral in onset-to-onset synchronization. Such a category also blurs the two second threshold that distinguishes synchronization and coordination, as according to the above IOI data, these gestures consistently occur above and below this mark (Clayton et. al. 2020, 140).
[33] Examples of ancillary movements involved in long-term coordination of vareis horoi include cueing by the lead dancer of transitions between sections, dance parts, and performance endings. Indeed, these singular movements are important for transitioning through formal and choreographic boundaries, but this category cannot include the ongoing gestural cycling of dance events that assist in the gradual transfer of coordination between the visual and aural found in Section 2. This suggests a possible category of movement involving a series of repeated gestures that, through their repetition, facilitate periods of intermodal coordination.
[34] These two types of visual information, sound-producing movements and ancillary movements, are situated in a conscious exclusion of dancers and other musical participants, a sentiment expressed through statements such as, “For the purposes of this paper, IME refers to the temporal coordination between co-performers…bodily movement of musicians is considered, although full consideration of dance is deferred,” (Clayton et. al. 2020, 140), and, “We concentrate on interactions between musicians, rather than taking into account audiences or dancers…,” (155). But as theories of entrainment begin to include the influence of dancers and other musical participants on IME, numerous other types of movement will be necessary to approach styles such as vareis horoi. This expanded consideration of movement will also have to address not only visual, but also haptic and proprioceptive modes inherent in embodiment that, in the example of Daoular Tsalar, directly inform the timing of sonic events and the expectations of metric coordination. Amendments such as these are very much in the spirit of current models of entrainment, as culturally specific knowledge of the expectations involved in performance timing are recognized as crucial to a broader understanding of IME (Clayton et. al. 2020, 154). The explicit use of multi-modal and shifting forms short-term and long-term coordination in vareis horoi provides an opportunity to recognize how this cultural specificity may manifest with regard to modality.
Conclusion
[35] The above IOI analysis and review of ethnographic literature demonstrates how movement and gesture connect musical phenomena across significant timing changes. Though compartmentalization was used to highlight timing differences between sections, factors uniting these sections such as choreography, beat proportion, and beat cardinality, hold equal analytical value. Furthermore, analyzing a musical style such as vareis horoi requires connecting sound with visual, gestural, haptic, and proprioceptive sensory dimensions in order to account for the integral coordinative factors that lie outside the purely sonic. The study of such coordination lies in the domain of entrainment, and a multi-modal approach could build upon and benefit the current role of movement and gesture in current models such as IME. More broadly, advocating for embodied analytical approaches stems from the reality that all music and dance is at first an embodied experience, and various musical phenomena creatively use this unity in a myriad of ways. Analyses of these styles require a methodological turn inward to include somatic feeling, and towards the sensory interactions that constitute musical experience.
REFERENCES
Bishop, Laura, and Werner Goebl. 2017. “Communication for Coordination: Gesture Kinematics and Conventionality Affect Synchronization Success in Piano Duos.” Psychological Research 82: 1177–1194.
Clayton, Martin. 2007a. “Time, Gesture and Attention in a Khyal Performance. Asian Music 38(2): 71–96.
———. 2007b. “Observing Entrainment in Music Performance: Video-based Observational Analysis of Indian Musicians’ Tanpura Playing and Beat Marking.” Musicae Scientiae 11: 27–59.
Clayton, Martin, Kelly Jakubowski, Tuomas Eerola, Peter E. Keller, Antonio Camurri, Gualtiero Volpe, Paolo Alborno. 2020. “Interpersonal Entrainment in Music Performance: Theory, Method, and Model.” Music Perception 38(2): 136–194.
Crum, Richard. 1998. “Yugoslavia: Vardar Zone.” In The International Encyclopedia of Dance. Oxford: Oxford University Press.
Danielsen, Anne. 2010. “Here, There and Everywhere: Three Accounts of Pulse in D’Angelo’s ‘Left and Right’.” Musical Rhythm in the Age of Digital Reproduction, edited by Anne Danielsen, 19–35. Farnham, UK: Ashgate.
Davidson, Jane. 1993. “Visual Perception of Performance Manner in the Movements of Solo Musicians.” Psychology of Music 21(2): 103–13.
Dimov, Ventsislav and Lozhanka Peycheva. 2002. The Zurna Tradition in Southwestern Bulgaria: Romani Musicians in Practice, translated by Ivanka Ivanova and Angel Igov. Sofia: Bŭlgarsko Muzikoznanie Izsledvaniya [Bulgarian Musicology Studies].
Engel, Annerose, Sebastian Hoefle, Marina Carneiro Monteiro, Jorge Moll, Peter Keller. 2022. “Neural Correlates of Listening to Varying Synchrony Between Beats in Samba Percussion and Relations to Feeling the Groove.” Frontiers in Neuroscience 16: 1–19.
Graziosi, Joseph. 2023, May 8. Personal Communication.
Goldberg, Daniel. 2015. “Timing Variations in Two Balkan Percussion Performances.” Empirical Musicology Review 10(4): 305–328.
———. 2017. “Bulgarian Meter in Performance.” Ph.D. diss., Yale University.
———. 2019. “What’s the Meter of Elenino Horo? Rhythm and Timing in Drumming for a Bulgarian Folk Dance.” Analytical Approaches to World Music 7(2): 69–107.
Jankowsky, Richard. 2013. “Rhythmic Elasticity and Metric Transformation in Tunisian Sṭambēlī.” Analytical Approaches to World Music 3(1): 34–61.
Johansson, Mats. 2010. “The Concept of Rhythmic Tolerance: Examining Flexible Grooves in Scandinavian Folk Fiddling.” In Musical Rhythm in the Age of Digital Reproduction, edited by Anne Danielsen, 69–83. Farnham, UK: Ashgate.
———. 2017. “Non-Isochronous Musical Meters: Towards a Mulitdimensional Model.” Ethnomusicology 61(1): 31–51. https://doi.org/10.5406/ethnomusicology.61.1.fm.
Kvifte, Tellef. 2007. “Categories and Timing: On the Perception of Meter.” Ethnomusicology 51(1): 64–84. https://www.jstor.org/stable/20174502.
Liebman, Robert. 1993. “Dancing Bears and Purple Transformations: The Structure of Dance in the Balkans.” Ph.D. diss., University of California: Los Angeles.
London, J. 2012. Hearing in Time: Psychological Aspects of Musical Meter. 2nd edition. Oxford: Oxford University Press.
lovingvirtue93. 2013. ΒΑΡΟΥΝ ΝΤΑΟΥΛΑΡΙ. June 26, 2013. https://www.youtube.com/watch?v=XBxOvKin_-4.
Moulopoulou, Katerina. 2011. “Ρουμλούκι – Λυπιρίδα.avi.” May 16, 2011. 3:51. https://www.youtube.com/watch?v=Fmjn2wVkHiE.
Pantelis Voukantsis. 2023. “ΓΚΑΪΝΤΑ ΠΕΤΡΙΤΣΙΟΥ στον 10ο χορό των ντόπιων Ν. Σερρών, “Να δω τα Σέρρας τα χωριά”.” March 14, 2023. 8:55. https://www.youtube.com/watch?v=5JtQXFo2F10.
Papakostas, Christos. 2012. “Flambouro Babougera1998.Φλάμπουρο Μπαμπούγερα. Papakostas mpg.” November 15, 2012. 10:12. https://www.youtube.com/watch?v=HfE0MHculwU&t=176s.
———. 2012. CD for North American Teaching Tour 2012. Self-released.
———. 2023, March 27. Personal Communication.
Polak, Rainer. 2010. “Rhythmic Feel as Meter: Non-Isochronous Beat Subdivision in Jembe Music from Mali.” Music Theory Online 16(4). http://www.mtosmt.org/issues/mto.10.16.4/mto.10.16.4.polak.html.
Silverman, Carol. 2012. Romani Routes: Cultural Politics & Balkan Music in Diaspora. New York: Oxford University Press.
Stover, Christopher. 2009. “A Theory of Flexible Rhythmic Spaces for Diasporic African Music.” Ph.D. diss., University of Washington.
Williamon, Aaron, and Jane Davidson. 2002. “Exploring Co-Performer Communication.” Musicae Scientiae 6(1): 53–72.
Μακεδονική Παράδοση. 2019. “ΝΤΑΟΥΛΑΡ ΤΣΑΛΑΡ – Ηράκλεια Σερρών [Μακεδονία].” August 26, 2019. 4:40. https://www.youtube.com/watch?v=ZEOElZut1iQ.
Χορευτικός Όμιλος «Ειρήνη». 2012. “Βουρούν νταουλαρί, Φλάμπουρο Σερρών – Χορευτικός ‘Ομ. Ειρήνη.” May 13, 2012. https://www.youtube.com/watch?app=desktop&v=GfscjbYEE2U.
[1]. For an example of this style in a live performance setting see Video Example 1, a video of the dance Lypirida from the region of Roumouki. This is a typical performance of a varis horos (sing.), and clearly shows its associated instrumentation, slow to fast formal progression, and general musician-dancer dynamics.
[2]. This compartmentalization is akin to Rice’s (2000) comments on the form of the Macedonian dance Pushteno Oro (or Beranče), in which he refers to slow, medium, and fast tunes (Rice 2000, 206).
[3]. This transformation is similar to the type found in Tunisian Sṭambēlī (Jankowsky 2013).
[4]. The video described here is Video Example 2 in which footage of Daoular Tsalar can be seen from 2:57-6:02.
[5]. In my interview with Papakostas, it is interesting to note that, when commenting on what he would change in making these audio recordings again, he mentioned that he would like to include more soundscape elements, as can be heard in celebratory performances seen in Video Example 2.
[6]. For more on singleton beats in a Balkan musical context, such as the first beat in a four beat 7/8 (1+2+2+2), see “What’s the Meter of Elenino Horo? Rhythm and Timing in Drumming for a Bulgarian Folk Dance” (Goldberg 2019).
[7]. Virtually all performances of this dance are three cycles in length, but there is one documented case of a four-cycle long version. Thank you to Greek dance scholar Joseph Gaziosi for guiding me through his personal collection of fieldwork videos, allowing me to make these larger-scale assessments.
[8]. As the dancer in this video, I am able to time my movements to this recording because I am accustomed to its general cyclic pace through my research engagement with it (and practice). This is supported by my familiarity with the general timing expectations of varies horoi through my lifetime experience in Balkan folk dance groups.
[9]. The standard deviations in milliseconds for these averages are 41 (Beat 1), 118 (Beat 2), 54 (Beat 3), 374 (Beat 4), and 411 (cycle duration).
[10]. Tŭpan is the Bulgarian word for daouli.
[11]. The standard deviations in milliseconds for these averages are 16 (Beat 1), 24 (Beat 2), 18 (Beat 3), 18 (Beat 4), 17 (Beat 5), and 34 (cycle duration).
[12]. An example of this interdependency is in Cycle 2 in Figure 5, where the long final beat causes the other beats to appear shorter in comparison with the surrounding cycles. A potential cause of this appearance could be that Beats 1, 2, and 3 were played shorter than usual in this cycle. However, the durational measurements of Figure 4 clarify that Beats 1, 2, and 3 of Cycle 2 are roughly the same duration compared with their respective beat categories, and therefore confirms that this proportioning was caused by a relatively long Beat 4. For more on this interdependency, see (Goldberg 2015, 310).
[13]. All three points have precedent in recent literature concerning meter and timing. For support of emphasizing the beat level in metric analysis see (Kvifte 2007; Polak 2010), for redefining beats as acceptable ranges of onset timings see (Stover 2009; Danielsen 2010; Johansson 2012; Goldberg 2017), and for a case of non-isochrony at all metric levels see (Johansson 2017).
[14]. The terms for these heavy dances are teshki hora in Bulgarian, teshki ora in Macedonian, and valle të randë in Albanian. The singular horo in Bulgarian, Oro in Macedonian, and valle in Albanian refers generally to traditional styles of dance in which heavy dances are a part. Thank you to Jane Sugarman for bringing the Albanian version of heavy dances to my attention during a presentation of this research.
[15]. For more on zourna and daouli groups as a transnational practice, see (Silverman 2012, 23–25).
[16]. The term for ‘zourna players’ in Bulgarian is zurnaci (pluralized in the quotation with the English -es), with maestro zurnaci referring to the lead zourna player. The word “tact” here refers to the Bulgarian word for meter takt.
[17]. Any grammatical mistakes in these quotes are from the original translation.
[18]. This translation pluralizes the Bulgarian horo with the English ‘-s’.
[19]. By this designation of the “transnational cultural region of Macedonia” I am referring to the country of North Macedonia; the Greek administrative regions of Eastern Macedonia and Thrace, Central Macedonia, and Western Macedonia; the Bulgarian Pirin-Macedonia region; and small portions of Serbia, Albania, and Kosovo.
[20]. While most often danced by men, this statement is not entirely true. The Roumluki region of Greece is famous for its many women’s heavy dances of which the above Video Example 1 is an example. Thanks again to Joseph Graziosi (2023) for bringing these dances to my attention.
[21]. For more on these studies, and a more extensive list of citations regarding visual cues and entrainment, see Clayton (2007b, 29) and Clayton et. al (2020, 140, 145–146).
[22]. For an example, see (Engel et al. 2022).