Jump to ContentJump to Main Navigation
Making It Up TogetherThe Art of Collective Improvisation in Balinese Music and Beyond$

Leslie A. Tilley

Print publication date: 2020

Print ISBN-13: 9780226661131

Published to Chicago Scholarship Online: September 2020

DOI: 10.7208/chicago/9780226667744.001.0001

Show Summary Details
Page of

PRINTED FROM CHICAGO SCHOLARSHIP ONLINE (www.chicago.universitypressscholarship.com). (c) Copyright University of Chicago Press, 2021. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in CHSO for personal use. Subscriber: null; date: 28 July 2021

Finding an Unspoken Model

Finding an Unspoken Model

The Boundaries of Reyong Norot

(p.51) Chapter Two Finding an Unspoken Model
Making It Up Together

Leslie A. Tilley

University of Chicago Press

Abstract and Keywords

The first of two chapters in a music-analytic case study on collective improvisation, this chapter introduces reyong norot, a collectively improvised practice for the Balinese gamelan gong kebyar’s four-person melodic gong chime reyong. The chapter’s goals are two-fold. On the smaller scale, it investigates the consciously-known but unspoken model guiding reyong norot performance: a model derived from the fixed norot playing of the ensemble’s gangsa metallophones then further informed by a knowledge base unique to reyong. Analyses reveal how the gangsa’s complementary polos and sangsih melodic strands, which are played using either specific interlocking figuration (kotekan) or parallel harmonizing notes (kempyung), are differently conceived on the reyong because of its construction and performance practice. More broadly, the chapter outlines a research methodology for uncovering unspoken models for improvisation and analyzing their manifestations in improvised practice. Proposing five basic steps – listening, learning to play, immersion, transcription, and one-on-one interactions – it delineates processes of discovery, ethnographic field methods, and musical analysis techniques from the author’s research on reyong norot as suggested research methods for analytical ethnomusicologists. Discussions throughout the chapter assume an outward bearing, applying concepts to diverse practices worldwide and suggesting rich potential for comparative and cross-cultural research.

Keywords:   reyong norot, reyong, norot, gangsa, gamelan gong kebyar, collective improvisation, research methodology, models for improvisation, model, musical analysis

Improvisation in Gamelan, After All

The familiar din of happy voices and the tinkling of wood on bronze greet me as I rush down the basement stairs of UBC’s Asian Center, characteristically two minutes late for class. I drop my backpack in the corner of the room and kick off my shoes. Not a regular occurrence in most university classes, but after a year of playing Balinese gamelan gong kebyar for credit, it seems as normal as brushing my teeth before bed. Our teacher Alit1 offers me a quick wave and a smile as I make my way over to the reyong, a long row of horizontal bossed kettle gongs played simultaneously by four people (see figure 2.1).I sit on the floor at the left-hand edge of the instrument, pick up my two panggul—mallets—and wait for Alit to come show us our parts.

We’ve been learning a new melody for the famous dance piece Oleg, and I’ve predictably forgotten much of what I learned last class, thanks to the traditional oral learning method for Balinese gamelan that I have affectionately dubbed “learn-forget-learn-forget-learn … remember!” I should be able to retain this new section by the end of today’s rehearsal, though. If only Alit would stop changing my part! (p.52)

Finding an Unspoken ModelThe Boundaries of Reyong Norot

Figure 2.1. Four musicians play the reyong. Left to right: I Dewa Putu Rai, I Dewa Madé Sakura, I Dewa Putu Berata, I Gusti Nyoman Darta.

(Photo by Nicole Walker, 2003. Used with permission.)

This is a new problem, and its cause is a bit of a mystery. The last year of playing has taught me that, though gamelan pieces are learned by rote, they are as precomposed and fixed as any notated composition. Most musicians in the ensemble, including we reyong-ers, play formulaically composed elaborations on slower-moving melodic lines, and artistic license is not within our purview. So I’ve been trusting Colin McPhee’s claim that “other than in solo parts there can be no place for spontaneous improvisation” in gamelan (1966, xvii).

All the reyong playing I’ve done thus far exemplifies McPhee’s assertion to a tee. The ensemble’s jack-of-all-trades, we sometimes join a small tray of cymbals in a rhythmic support role, tapping running 16th notes with the brassy clang of wooden mallets on gong rims. In other moments, our playing exudes the fiery unpredictability of kebyar, a musical aesthetic meaning to “flare up” or “burst open.” In perhaps our most iconic technique, percussive 16th notes are interspersed with syncopated shots exploding in a perfect synchrony of 8-note cluster chords.2 But my favorite thing of all is playing fast interlocking melodies: sharing a phrase that’s been carefully divided into two complementary strands, allowing us together to reach speeds that none of us could alone.3 In the last piece Alit taught us, the warrior dance Baris, these interlocking melodies were as composed and unchanging as our other techniques: two parts meticulously designed to fit together like jagged puzzle pieces. In first position at the reyong’s left-hand edge, I played the phrase in figure 2.2.My friend (p.53) Jocelyn in third position doubled my part up the octave. I later learned that this reflected a principle of octave equivalence fundamental to reyong performance, where notes in lower octaves are functionally the same as their scalar equivalents in higher octaves.

Our partners in second and fourth positions interweaved the rhythm in figure 2.3. Together we made a continuous, interlocking melody from four adjacent tones. Shown in figure 2.4, this technique is called ubit empat, or simply empat, meaning “four.”4

We eventually discovered that removing the highest pitch, which in empat figuration always plays in harmony with the lowest, revealed a carefully composed 3-note melody snaking its way around this small range.5 The top staff in figure 2.5 shows this melody divided into its two voices; in the bottom staff, all the notes of the composite melody are beamed together.

The nature of Balinese interlocking meant that each musician needed to stay true to his or her part in order to maintain the integrity of this melodic composite. Gamelan was not an improvising art. Or that’s what I had always believed.

So what in the world was happening now? When Alit teaches gamelan he does it with the assurance of someone who’s been playing since he was a small child. He can deftly

Finding an Unspoken ModelThe Boundaries of Reyong Norot

Figure 2.2. Baris interlocking pattern on reyong, 1st and 3rd positions

Finding an Unspoken ModelThe Boundaries of Reyong Norot

Figure 2.3. Baris interlocking pattern on reyong, 2nd and 4th positions

Finding an Unspoken ModelThe Boundaries of Reyong Norot

Figure 2.4. Baris interlocking empat pattern on reyong,all players

Finding an Unspoken ModelThe Boundaries of Reyong Norot

Figure 2.5. Baris interlocking empat pattern on reyong, composite melody

(p.54) move from instrument to instrument, switching between vastly divergent parts without the slightest hesitation or memory lapse. But last rehearsal it took him several attempts to settle on figuration for each of the four reyong players, as though he were weighing alternatives, and today he has forgotten what he taught us last class and seems to be choosing from a whole new set of slightly different options with an out-of-character indecision. This new figuration style—norot it’s called—is not the predictable empat we’ve grown accustomed to! When I finally summon the courage to say, “I don’t think that’s what you showed us last time,” his answer changes everything: “It’s ok,” he smiles. “You can just improvise on what I’ve shown you. Go ahead. Try it.” And he walks away confidently to help another student.

My three compatriots and I look back and forth with confusion and some concern, and much less confidence in our abilities than Alit appears to have. Then, seeing no other choice, we begin hesitantly to “improvise.” More quickly than ought to be possible, Alit is back. “Not like that,” he says, with a pained look. But when we ask what can and can’t be done while “improvising,” our questions are met with uncertainty; Alit simply knows what sounds right and what doesn’t. He can show us idiomatic patterns, but he can’t explain what makes them work. It comes from his rasa he says: his “feeling” developed though years of playing the music in Bali. “OK,” he finally concedes. “I’ll just show you one for now.”

The combination of Alit’s indecision, his pained looks, his verbal instructions, and his Balinese rasa means three things to me: (1) though Balinese interlocking generally necessitates fixed idioms, reyong players have freedom to improvise in norot; (2) this freedom is not without its boundaries; and (3) though Alit obviously knows what the boundaries are—he can perform many versions of the norot melody and is quick to recognize inappropriate note use in our playing—it’s difficult for him to express these things verbally. His conceptual space for norot improvisation, it appears, is largely unspoken.

Unraveling Conceptual Spaces for Improvisation

The Importance of the Model

To understand Alit’s conceptual space for norot, I first needed to be familiar with the model on which he was improvising. I could not truly grasp the mastery of his performance, appreciate the ways in which he individualized techniques of interpretation, embellishment, recombination, and expansion, or gain insight into his embodied experience without first knowing the foundation for his creative choices. I might still enjoy his performance, of course, but the craft behind his creativity was a more elusive animal. The same is true in the study (p.55) of any improvised practice. Research in cognition and perception has found that a listener’s level of familiarity with a piece and her experience within its musical system work in concert to determine the subtlety with which she can recognize alterations to the piece in different performances.6 Knowing the musical and aesthetic expectations of a specific raga lends depth to a listener’s appreciation of an improvised alap in a way that a casual interest in Hindustani music cannot. In much the same way, only fluency in the written and tacit rules of Italian partimenti reveals an improviser’s often understated choices in that practice. And knowing to listen for creative references to famous recordings in the performance of well-known Brazilian pagode songs, or recognizing the new spin that a skilled jazz soloist puts on an old favorite, can allow for fuller insight into each of these musicians’ artistries than hearing them improvise on something unfamiliar.7 Whether or not we are consciously aware of it, our indoctrination into these models as listeners helps us navigate performances to identify moments of particular ingenuity. Like an ornithologist’s field guide:

You can approach [this knowledge] pragmatically to ID particular flora and fauna. Or you can read it for more general tips on how to go a-looking. A field guide can help you figure out how to be ready to go into the woods, can give you a framework for observing once you’re out in the woods, maybe even help you get over an inhibition about venturing into the woods. A framework, a chassis, a scaffold, some tips, a dram of encouragement.

(Corbett 2016, 7)

Whether a specific template for formulaic variation or a broad schema for exploration, a model contextualizes a performance, allowing real-time access to an improviser’s musical processes. Yet this link between model and performance can also present one of the greatest initial hurdles to the analysis of improvised musics. When a model is unconscious, or conscious but unspoken, analysts and uninitiated listeners are often left in the dark. To find the models behind such practices, we cannot simply open a “fake book,” as we might to learn the basic tune and changes of a jazz standard, or read a treatise, as we might to ascertain the desired aesthetic of a given raga.8 Nor will attending many performances by the same ensemble, as Corbett suggests for discovering free jazz models, be sufficient.

I showed in chapter 1 how Simha Arom used a combination of fieldwork and analysis to find the consciously known but unspoken models of Central African polyphonic horn improvisers. An expert reyong player likewise wields a storehouse of fine-grained musical knowledge communicated to him through (p.56)

Finding an Unspoken ModelThe Boundaries of Reyong Norot

Figure 2.6. Four “equivalent” improvised norot melodies

practice alone: a collection of very specific musical and aesthetic requirements that nevertheless remain unspoken. In performance, this model and knowledge base are materialized through continuous, varied, rapid-fire choices made in an exquisitely limited musical field of just a handful of tones. Four short norot improvisations demonstrate this constrained flexibility. Alit played each of these phrases, marked for analysis in figure 2.6, in different iterations of the same piece, though one might also hear four such phrases played simultaneously by the full reyong complement.9 Thus, like multiple improvised performances of the same jazz tune, these phrases can be considered paradigmatically equivalent.

The concept of musical equivalence is culturally bound and system specific; variations on a model retain perceptual similarity “to the extent that the transformations do not violate the ‘grammaticality’ of the musical system within which the material has been conceived.” This “suggest[s] a close link between comprehension of the musical system and the categorization process that one may hypothesize to operate in the recognition of variations of an original theme” (McAdams and Matzkin 2003, 91). Understanding how Balinese musicians deem these four melodies equivalent involves deciphering unspoken conceptual spaces of cognition and musical expectation, and a preliminary analysis can help define the boundaries of this task. One might first notice that Improvisations 1 and 2 share certain similarities of range and contour, as do Improvisations 3 and 4: pitch use in their first beats is identical, while rests create some variation between them. Yet Improvisations 1 and 4 feature larger ranges: pitches marked with asterisks at A and B are not used in the other two improvisations. What determines these range choices, and when should extra notes be employed? Further, while (p.57) Improvisations 1 and 3 use almost continuous 16th-note motion, Improvisations 2 and 4 contain many 16th-note rests, often creating syncopated gestures, as in the third and fourth beats of Improvisation 2. What is the function of these rests, placed seemingly at random throughout Alit’s improvisations? Are they a part of the rhythmic quality of the model or do they represent a creative move away from it? Equally curious are instances where, in two similar improvisations, Alit chooses to play different pitches in the same subdivision, as at C and D. Does the model contain both these notes, and if not, is their concurrence considered dissonant? Are there even operative concepts of consonance and dissonance in this practice? Next, even in closely related improvisations, Alit appears to have a relatively wide range of freedom in note placement within his small ambitus. At E, Improvisations 3 and 4 have the same end goal—the 3-note ascending gesture—but the approach to it differs. At F, Improvisations 1 and 2 use the same limited pitch palette, but temporal placement of the tones is different in all but one subdivision. What determines Alit’s choices here? Does the model dictate specific rhythmic placement, as might be suggested by the extensive neighbor-note oscillations in beats 3–6 of Improvisation 3? And if so, what are the idiomatic ways of diverging from this rigidity? Finally, how do Improvisations 1 and 2 relate to Improvisations 3 and 4, and what makes them “equivalent”?

Both the unspoken model and the unconscious knowledge base of reyong norot are needed to answer these questions, and over the next two chapters I will explore this conceptual space, using fieldwork and analysis to gain insight into the embodied experience of collectively improvising musicians. In many ways, reyong norot presents an ideal first case study. Each norot improvisation is based on the same specific, unspoken but consciously known template for formulaic variation: a simple set of rules that determines melodic construction for every player in any piece. Unlike in jazz, where each standard has a different melody and chord changes, or in Hindustani music, where individual ragas each have their own musical and aesthetic guidelines, the norot model’s standardization across the repertoire allows for easy comparative analysis between pieces, eras, villages, and performers. What’s more, gamelan gong kebyar uses a limited five-tone scale, with each reyong player controlling just two to four of those tones. While reyong musicians have developed ingenious techniques engendering a virtually infinite number of possible variations on a melody, the model’s limited scope and specificity and the instrument’s narrow pitch range provide a small pool into which we may dive deeply. They enable comprehensive discussions examining how improvisers diverge from models in performance, and provide relatively straightforward examples of each of the four general improvisatory techniques discussed in chapter 1.

(p.58) This chapter serves two purposes. At the microlevel, it introduces the specifics of the reyong norot model and preliminary knowledge base, preparing the reader to fully appreciate the subtle wizardry of the improvisations in chapter 3. But it begins at the macrolevel with a discussion of research techniques that fuel good analysis—methods for finding unspoken models and formulating culturally relevant hypotheses for their knowledge bases.

Searching for a Model

Although the process for uncovering an unspoken or unconscious model and effectively analyzing its improvisations will be different in every music culture, in my forays into collectively improvised genres in Bali, I found five general steps to be essential. I present them here in what I consider to be the most useful chronological order, though it will quickly become apparent that each step leads back to the other four in a continual feedback loop of information and inspiration.10 Along the way, I introduce some of the basics of gamelan gong kebyar necessary to an analysis of reyong norot. While the research process outlined here may seem obvious to the seasoned ethnomusicologist, I hope that articulating it explicitly will underscore the challenges and benefits of each step.

Step 1: listening

When Mantle Hood proposed the then-radical concept of bi-musicality11—the idea of acquiring performing fluency as an ethnomusicological research method—he opined that the first obstacle to success would be developing “an ability to hear” (Hood 1960, 56). The same is true for music analysis. An English speaker wishing to learn Cantonese, Hmong, or Ewe needs first to become accustomed to hearing the various tones in those languages. In much the same way, to fully understand Arabic or Persian classical musics, for instance, we must learn to hear the quarter tones that an ear trained in Western traditions might be inclined to “equalize.” Gamelan gong kebyar, too, uses a non-Western tuning system, all the more complex because it is not standardized. Generally referred to using the Balinese solfège syllables ding-dong-deng-dung-dang (which I abbreviate i-o-e-u-a), its five-tone scale translates roughly to C#-D-E-G#-A, as in figure 2.7.12 Perceiving this collection of pitch classes as “normal,” and not expecting them to behave as notes in Western scales might, takes time, patience, and exposure. The clangorous timbre and complex overtones of the ensemble’s (p.59)

Finding an Unspoken ModelThe Boundaries of Reyong Norot

Figure 2.7. The five tones of gong kebyar’s scale, with solfège and abbreviations

bronze keys and gongs, as well as the pulsating shimmer of its characteristic paired tuning, equally require acclimation.13

Repeated and attentive listening goes much deeper than simple pitch discernment, however. It builds familiarity with a full musical system, from the general swing feel in many jazz styles to a specific musician’s favored unfolding of the pitch palette in a Hindustani alap. Typical harmonic progressions or melodic trajectories, common structures, appropriate ornaments, characteristic rhythms or riffs—these and many other aspects of a musical language are revealed and made to feel familiar through thoughtful listening.14 One might initially notice, for instance, that gamelan gong kebyar pieces are often cyclic, and that while some instruments play very quickly, others consistently play sparsely. With repeated listening yet more details will come into focus and begin to seem both obvious and expected. This familiarity is even more vital when analyzing improvised traditions. Where each performance of a piece may be quite different from any other, learning to recognize when a performer has reached the boundaries of his or her idiom is that much more challenging. Collectively improvised genres, moreover, demand that we tackle these questions for multiple musicians simultaneously. But “if the spirit, melodies and rhythms of the music are already a part of you,” if you have embodied the tradition, “a theoretical analysis or a transcription is more effective” (Blake 2010, 13).

Step 2: Learning to Play

Most ethnomusicologists agree that understanding and embodiment of any form of music can be enhanced through performance.15 Recent studies in embodied cognition support this notion, proposing that “the human motor system, gestures and body movements play an important role in music perception” (Leman and Maes 2014, 236).16 This means that “the concepts with which we consciously and verbally make sense of musical experience are likely to be shaped by the structure of our bodies and by our experience of interaction with our environment” during performance (Clayton and Leante 2013, 197). (p.60) Finding the unspoken model behind an improvised tradition like reyong norot requires fluency both on the instrument and within its larger musical system, and playing the music allows us to understand it “from the ‘inside,’ so to speak. The structure of the music comes to be apprehended operationally, in terms of what you do, and, by implication, of what you have to know” (Baily 2001, 94).17

One need not become a virtuoso in the genre, but basic musical proficiency on the instrument being studied and performance experience in its ensemble context create a baseline of embodied idiomatic knowledge that fosters more culturally faithful analyses. In his call for bi-musicality, Hood argued that this “basic musicianship” would ensure our “observations and analysis as musicologist[s] do not prove to be embarrassing” (Hood 1960, 58). Yet at a more profound level, as Michael Bakan observes, “learning to play” eventually becomes “playing to learn”; we reach a point when we can “productively engage what are often [our] most direct and subjective experiences of the ‘alien musics’ [we] study, the experiences of music-learning and performing, and in so doing move toward new and deeper musical and scholarly understandings” (1999, 316). In his research into Balinese gamelan beleganjur, Bakan’s playing experience gave him access to local concepts of trust and partnership in music performance, taking him far beyond the notes on the proverbial page. In my own research in Bali, performance taught me some of the music’s fundamentals while also facilitating connections with potential teachers and collaborators and providing insight into traditional pedagogies. It helped me, as Bakan puts it, “inadvertently become more ‘Balinese’ through the course of my development” (318) as a learner, listener, and player.18

Performing and the Fundamentals

When Western-trained musicologists like me analyze Western improvised music genres, we can take so many of the fundamentals for granted. Performing within these and other Western practices has helped us embody basic assumptions about how the music should feel and sound: the make-up of the major and minor scales, the strong pull of the leading tone toward the tonic, the feeling of an accidental or syncopation, a suspension or cross-rhythm, the culturally agreed-on concepts of consonance and dissonance, the location of C on our instrument, or on a treble clef, and so on. The more we perform, the more refined our foundation becomes. We begin to instinctively know which note combinations are idiomatic for a particular style, which contours or gestures most appropriate for a given moment in an improvisation. We begin to (p.61) internalize idiom, to develop rasa, “feeling.” But different music cultures have different foundations with different assumptions, idioms, and rasa.

At the heart of any gong kebyar ensemble, for instance, is a collection of hanging gongs. Many gamelan pieces are built in contrasting sections of cyclic melodies, and strokes on these gongs emphasize important structural points within them.19 I once played a recording of Bach’s chaconne for violin to one of my Balinese drum teachers, a longtime friend and very skilled and curious musician. After a minute or two of intense listening, he looked at me quizzically and said, “It’s interesting … but where’s the gong?” In asking this, he was not actually implying that what Bach had been missing all these years was a good bronze gong; he simply felt the music lacked recognizable structural markers. My indoctrination into Western music allowed me to feel the chaconne’s structure through the implied harmonic motion of the violin’s melodies; his training in gamelan had prepared him for a very different kind of musical understanding. In Balinese gamelan, the end of a melodic cycle is generally marked by the largest hanging gong: the gong ageng or great gong (often abbreviated “G”). Other gongs highlight other important beats. Underlying the 8-beat Baris melody from figures 2.22.4 is a pattern using the gong ageng and medium-sized kempur (P) in the following configuration: (G) ___ G P _ P G.20 Learning to play gamelan helped me understand the importance of such cyclic structures. I quickly learned that if I got lost in the unfamiliar rhythms of my interlocking melody, I could listen to the gongs to find my way again. But I would later discover that their function as structural markers influenced reyong norot improvisation too (discussed in chapter 3).

Performing and Traditional Pedagogies

If one is fortunate enough to work with musicians from living music cultures, performance also allows access to traditional pedagogical practices, and these often reveal essential musical concepts.21 Central to gamelan music, for instance, is an end-weighted metrical system where the stress is felt not at the beginning of each cycle but at its end, and melodies are composed to lead toward strong beats. “This is the exact inverse of tertian harmony where bass provides the harmonic foundation for the material to come (for example, a C chord is played on the downbeat of a bar, followed by a measure of melodic material based on C)” (Steele 2013, 230). As a Western-trained musician, I naturally felt Balinese gong cycles as starting on beat 1, and would count them gong-2-3-4-5-6-7-8-gong. My Balinese teachers, though, would always count gong-1-2-3-4-5-6-7-gong. (p.62)

Finding an Unspoken ModelThe Boundaries of Reyong Norot

Figure 2.8. End-weighted cycles and beats: the reyong’s composite Baris melody

To them the gong ageng did not begin the cycle but ended it; the strong beat was not 1 but 8. In the Baris gong pattern (G) ___GP_PG, the first gong ageng is notated in parentheses because it actually “belongs” to the previous cycle, and each gong stroke after it is considered to be leading toward the beat-8 gong ageng at pattern’s end.

Learning gamelan from Balinese musicians also taught me that, when considering the subdivisions of a single beat, the accent likewise falls at its end. Two beats, each divided into four subdivisions, are not thought of as 1-2-3-4-1-2-3-4-1, but rather 4-1-2-3-4-1-2-3-4; accordingly, Balinese musicians almost always begin singing or playing a melody after the beat. While beaming conventions of Western staff notation make representing this perception difficult, in solfège notation, I follow each strong tone with a backslash (/) per figure 2.8; each marks not the beginning of a collection of four notes, but its end.

This end-weighted conception of meter and beat influences how melodies, both composed and improvised, are constructed: motives are designed to lead to strong beats and subdivisions. Yet only through traditional pedagogies and performance can the proper feeling for this metrical organization be embodied and the appropriate analysis pursued.

Performing and Social Connections

The benefits of performing also reach beyond these technical, aesthetic, and kinesthetic dimensions. When beginning fieldwork, playing experience can “provide one with an understandable role and status in the community, and it can be very useful in early orientation. It explains why you are there and what you are doing” (Baily 2001, 95). Being able to play can help an ethnomusicologist gain the trust and acceptance of potential teachers and collaborators, and may in time unlock access to more advanced musical, pedagogical, social, or philosophical concepts. As Helen Myers observes:

(p.63) Ethnomusicologists are more fortunate than anthropologists and sociologists because the private feelings we study are publicly expressed in musical performance. Cultural barriers evaporate when musicologist meets musician. There is no substitute in ethnomusicological fieldwork for intimacy born of shared musical experience. (1992, 31)

Ingrid Monson (1996) credits her familiarity with jazz and “considerable professional experience as a trumpet player” (16) as well as her connection to well-known radio producers for opening doors to interviews and lessons with jazz musicians. Yet she also recalls: “in some cases, it took several trips to performances to make arrangements [for an interview], as musicians checked me out” (16). Recounting his experience in Afghanistan, John Baily observes that playing the music “gave me an immediate and large area of common experience with people to whom I was a complete stranger. We were all heirs to a common musical tradition [… and] musical relationships form[ed] the basis for social relationships” (2001, 96). And Paul Berliner tells of the many years of relationship building and practical study required to truly gain the trust of his Shona mbira teacher and community, out of which stemmed access to privileged information about the instrument. He notes: “the elders who are the guardians of an oral tradition do not treat their knowledge lightly. Rather, they ‘give what they like.’ Moreover, they give only the amount of information they believe to be appropriate to the situation and to the persons involved” (1993, 7).

In my research into improvised arja drumming, performing gamelan gong kebyar in Bali connected me with musicians my own age who could teach me the basics. Once they deemed me sufficiently proficient, they would introduce me to older master drummers they admired. These men, like Monson’s jazz musicians, would always “check me out” before agreeing to work with me. When my friend Sudi brought me to meet the Peliatan-based arja master Cok Alit, I was offered a cup of coffee while the two of them spoke at length in Balinese (I knew only Indonesian). Cok Alit initially seemed reluctant to teach me, in part because I was a woman looking to learn a man’s instrument, and in part because he could not be sure that I was both capable and serious. After what seemed an eternity, he picked up his drum, played a short pattern, and handed the drum to me. Sending silent thanks to my parents for their musicians’ genes and my undergraduate professors for their aural perception classes, I played the pattern back to him on the first try. A smile crept across his lips, growing wider and wider until he broke into uproarious laughter. “OK,” he said at last. “I’ll see you tomorrow at 10:00.” Over time it became clear that Cok Alit enjoyed teaching me because I wanted to learn the older classical styles many young Balinese (p.64) musicians no longer deemed interesting. Our shared interest gave us common ground and brought us closer, both as musicians and as people. As our friendship developed, I gained access to Cok Alit’s intimate thoughts and ideas about music, and he seemed to feel excited and proud to play in recording sessions I organized. In our lessons, though, I would only be taught more difficult drumming patterns once I had mastered proper technique on easier ones. To Cok Alit, commitment and potential got me through the door, but playing proficiency, not simply mental understanding, was the key to my advancement. The same was true of my norot lessons with Alit. And because each pattern I learned taught me more about models as well as collective improvisation, my playing abilities were essential to my research.

The insights gained through performance, and the questions they raise, lead back to Step 1 for a more nuanced listening experience. But they also inspire a move into Step 3: a fuller immersion into the existing knowledge base of the genre.

Step 3: Immersion

Research in second language acquisition has shown that the most effective approach toward fluency is intensive immersion. In a comparative study of college students learning French in different educational contexts, for example, not only the number of class hours but also, significantly, the consistency with which the target language was used in varied out-of-class activities, affected the acquisition of oral fluency (see Freed, Segalowitz, and Dewey 2004).22 Learning a musical language likewise benefits from full immersion. Whether this takes the form of the ethnomusicologist’s once-standard “year in the field,” a collection of shorter, more intense research trips complemented by conversations and lessons over Skype, or many evenings and weekends of research done locally, immersing ourselves into the music cultures we study—improvised or otherwise—is essential to analysis. Even before taking lessons or doing one-on-one interviews, there are dozens of ways to dive headfirst into a music culture. For me, in Bali, immersion primarily meant hanging out with musicians. Lots of them. It meant being a fly on the wall, listening to and participating in as many conversations as I could, even if I didn’t always understand what was being discussed or have the background knowledge to make intelligent contributions. So it also meant sometimes feeling awkward or stepping out of my comfort zone. It meant making friends and accepting invitations, letting conversations flow naturally without trying to steer them too much; these were musicians and inevitably music was what they loved talking about most. (p.65) It meant attending every performance I could find and watching every rehearsal my musician friends allowed me to (which was all of them) to start getting a sense of different improvisers’ styles. It meant actively participating in these events whenever I was invited, whether or not I felt equal to the task. It meant getting permission to record rehearsals, performances, lessons, and performance postmortems. All of these activities helped me better understand the music cultures of gamelan gong kebyar and arja, confirming, contradicting, or adding nuance to the things I had read and the hypotheses I had begun to formulate.

My first gamelan immersion experience, however, was not in Bali, but rather in the town of El Cerrito, California, the then stomping grounds of the long-standing American gamelan group Gamelan Sekar Jaya. After spending a year

playing gamelan gong kebyar twice a week at the University of British Columbia (UBC), I’d decided it was time to up the ante. So for two months in the summer of 2001, I rented a room in the home where Sekar Jaya housed their instruments, living downstairs from their artists in residence, radical composer I Madé Subandi and revered dancer Ni Ketut Arini. Each morning, I rolled out of bed and played reyong before breakfast, slowly mastering the difficult technique of damping the instrument’s bright ringing bronze with one mallet while playing my next note with the other. I took lessons every day, sat in on rehearsals and performances, hung out with Subandi and Arini in the evenings, learned a little Indonesian, and soaked up as much music and culture as I could. It was at a Sekar Jaya rehearsal that I gained insight into one of the most foundational aspects of Balinese compositional process, central to the reyong norot model: the “core melody,” or pokok.

Aside from the reyong, I had only tried my hand at one other gamelan instrument: one of a group of eight two-octave metallophones called gangsa. Like the reyong, these instruments were responsible for the ensemble’s fastest-moving melodies. In the last year of playing with my group at UBC, I had noticed that slower-moving metallophones seemed to loosely track these faster melodies, sparsely striking our notes every few beats. Like the early scientists who believed that the sun and stars revolved around the earth, because that’s how it looked from their perspective, I assumed that these instruments’ melodies were derived from ours. I was wrong.

Tonight I am sitting at a gangsa in Sekar Jaya’s gamelan room learning a melody from the piece Oleg Tumulilingan (Oleg), the very same melody that, six months later, will spark my curiosity in reyong norot. “OK, here’s the pokok, the core melody” Subandi begins, singing the tune in figure 2.9 while tapping the beat on his thigh.23

The lead metallophone ugal begins to trace this core melody, adding grace notes and (p.66)

Finding an Unspoken ModelThe Boundaries of Reyong Norot

Figure 2.9. Oleg Tumulilingan’s core melody (pokok)

Finding an Unspoken ModelThe Boundaries of Reyong Norot

Figure 2.10. Learning a new melody on gangsa, I Dewa Ketut Alit and Leslie Tilley.

(Photo credit unknown, 2001.)

passing tones, and the small horizontal gong, kempli, joins him, tapping a steady beat with loud, dry strokes. “Just like in Jaya Semara,” Subandi calls out, smiling at the woman sitting among the hanging gongs. She nods and begins to play a 16-beat cycle of strokes on the large gong ageng (G), medium-sized kempur (P), and small klentong (t): (G) ___P___t___P___ G. I notice that the three pairs of one-octave metallophones considered “core melody instruments” have also begun to figure out their parts, the penyacah playing every beat, the calung every two beats, the jegogan every four. Subandi watches them for a moment, leans forward to help one guy who seems to be struggling more than the rest, then decides that the others will figure it out on their own. He turns to the reyong—peopled with some of the ensemble’s senior members, I know—and says “OK, norot.” Then, “I’ll be back.”

Finally, Subandi turns to us gangsa players and begins to play face-to-face with the person nearest him, a common pedagogical approach shown in figure 2.10.

(p.67) I crane my neck to see what he’s playing, piecing my melody together one beat at a time. But I notice with a blend of surprise and awe that, with just the core melody and the word “norot” to work from, many of the more advanced gangsa players have already figured out their parts. And suddenly, like Copernicus, it dawns on me: perhaps the core melody is not following us. Perhaps the earth revolves around the sun after all; perhaps we derive our melodies from theirs.

I was to call this rehearsal to mind months later, back in my own gamelan ensemble at UBC, when Alit told us reyong players to just “improvise” our norot. Although they would eventually need some help from Subandi, the reyong players in Sekar Jaya could clearly formulate norot improvisations from the core melody alone. The gangsas, too, seemed to know how to proceed as soon as Subandi said “norot.” And the questions began swirling. How did the core melody determine the reyong and gangsa parts? Did the gangsas, who were never allowed to improvise, also play a kind of norot? If they did, how was it connected to reyong norot? Did the two instruments share a single model, and if so, why did their melodies seem so different?

Each of these questions was born out of immersion. Immersion let me experience a Balinese style of teaching, watch and record more experienced players working through problems, and gain new Balinese vocabulary to contextualize the things I was learning. It would eventually help me to ask new questions of my teachers in their own musical vernacular and to formulate new, more relevant hypotheses. Although I did not yet fully understand how they had done it, the Sekar Jaya musicians helped me think about the music in more culturally accurate ways. I was now ready to start transcribing the things I had learned.

Step 4: Transcription

Like analysis more generally, the use of transcription in ethnomusicological research has been the topic of an ongoing negotiation since the mid-twentieth century.24 The choices we make about what to emphasize, what to omit, and which notation systems to use will unavoidably affect both our analyses and the perceptions of our readers; like the “theoretical cuts” of Seeger’s banana, each approach “will yield different visions of what music is” (2002, 189). This is why Steps 1 through 3 are so vital early in the research process; they help us make culturally sensitive decisions about visual representation.

When using staff notation for gamelan, I have chosen to approximate pitch to the Western scale. In a tradition where almost all melodic instruments are (p.68)

Finding an Unspoken ModelThe Boundaries of Reyong Norot

Figure 2.11. Oleg Tumulilingan’s core melody instruments: stratified polyphony with end-weighted meter

idiophones with no moment-to-moment pitch variability, and individual sets may be tuned quite differently from one another, the clutter of additional pitch detail seemed unnecessary. Instead, I include Balinese solfège in all transcriptions, which allows the material to be played on any set of gong kebyar instruments and, importantly, to be accessible to Balinese musicians not versed in staff notation. The audio and video files on the companion website further clarify how the music should look and sound. I have also opted to transcribe both reyong norot and kendang arja using strict 16th notes, ignoring any nuances of microtiming that may occur in performance. While interesting for a study of player precision and flexibility, or for examining potential subconscious relationships of leaders and followers,25 this level of detail was not pertinent to the questions I was asking.

What I did consider important was a visual representation of end-weightedness. Because of the limitations of Western staff notation, strong beats in all transcriptions are notated at the beginnings of measures, strong subdivisions at the beginnings of beats. But the first tone aligning with gong (the “gong tone”) is placed in parentheses, implying that it belongs to the previous cycle, and each transcription ends with the last gong tone—a single note standing by itself. As in Oleg’s various core melody parts in figure 2.11 (to be discussed), there are no bar lines and no repeat signs; the first and last note transcribed are both the gong tone.26

Despite the inherent limitations of any notation system, transcription facilitated my research on norot models and improvisations in three important ways: (p.69) it solidified my understanding of the larger musical context, confirmed and reinforced things learned in lessons and rehearsals, and allowed start-and-stop replay of recorded improvisations from Balinese musicians, thus enabling more detailed analysis.

Transcription for Context

My earliest transcriptions for norot research were not yet concerned with the minutiae of improvised patterns; they served instead to better understand the larger musical context of an unnotated ensemble genre in which I played only a single instrument. We learned in the previous section that gamelan pieces are built around a pokok, a core melody from which musicians on fast-moving instruments like reyong and gangsa somehow formulaically derive their parts.27 Thus, discovering the norot model first required an examination of core melody, and recording rehearsals allowed me to transcribe what other people were playing, helping to contextualize my own part within the piece as a whole.

The melodic building blocks of gong kebyar comprise multiple variants of a single melody running simultaneously at different note densities. The Oleg melody sung by Subandi and played by the lead metallophone ugal dictates notes for three pairs of core melody instruments, which result in a stratified polyphony characterized by regular vertical pitch class convergences.28 Each instrument pair interprets the ugal’s melody at its own density, coming into pitch class unison every two beats. Shown in figure 2.11, the penyacah play at the quarter-note density, compressing the melody into their one-octave range and interpreting off-beat notes in their own idiom.29 The calung play every other note in the penyacah’s melody, while the jegogan play every four notes, each beginning and ending with the gong tone and maintaining vertical pitch class convergence (circled with dotted lines).30

As we will see, the concept of a stratified polyphony based on a single melody also extends to the fast elaborating instruments, and it is only through this core melody that the norot model becomes clear.

Transcription for Confirmation and Detail

While my initial transcriptions were mostly useful in providing context, they also aided in retention and clarification. This was especially true of melodies for which I only played half the notes. When interlocking, the eight gangsas (p.70)

Finding an Unspoken ModelThe Boundaries of Reyong Norot

Figure 2.12. Polos part in interlocking gangsa melody, Oleg Tumulilingan

Finding an Unspoken ModelThe Boundaries of Reyong Norot

Figure 2.13. Composite gangsa melody, Oleg Tumulilingan

are divided into two roles: four so-called polos musicians generally align more closely with the beat, while four sangsih players sound the notes in between. I thought of these respectively as the “easy” and “hard” parts, though their names actually mean “basic” and “complementary.” For Oleg, I was learning the polos, shown in figure 2.12.

I knew that the sangsih players filled in the spaces between my notes in a reasonably strict alternation, but I was not at the point where I could focus on this interlocking while playing. Still a relatively new gamelan player, shutting out the sangsih when things were fast or difficult was a survival tactic for me. So it was through recording and transcription that I could both confirm my own part and understand the interlocking melody I made with my partner. In figure 2.13, the two parts are shown on the top staff with polos stems down and sangsih stems up; the bottom staff shows the composite melody formed through their interlocking.

A little research told me that this melody was indeed a kind of norot, a figuration style characterized by a “wavering or neighbor-note motion” (Vitale 1990, 5).31 But although they had the same name, this unimprovised gangsa figuration was not the same as reyong norot. How were these two elaboration styles connected? Conversation, lessons, and transcription of improvised norot would all point to the fixed gangsa melody as a model for reyong improvisation.

Step 5: One-On-One Interactions

The first four steps in the research process prepare ethnomusicologists to get the most out of our one-on-one interactions with teachers and collaborators. (p.71) Performance illuminates the aesthetic and technical challenges of the practice; reading, listening, and immersion foster familiarity and generate useful vocabulary; transcription clarifies musical context and points out mysteries to be solved. These skills and understandings, when combined with relationships formed through immersion, generate the social capital to ask for private lessons, interviews, and recording sessions with master musicians as well as the intellectual and musical frameworks to more fully benefit from them. The full research package teaches us how to think about the music in more culturally relevant ways and thus to better pose questions of musicians whose knowledge of their tradition may be tacit.32 It introduces us to the scene, exposing us to improvising musicians we might want to take lessons with, interview, or record. And it teaches us something about the most effective ways, socially and musically, to achieve these goals.

Recording Sessions

When analyzing music without readily available recordings, or where the details of improvisation in recorded tracks, as with reyong norot, are buried beneath other louder instruments, organizing recording sessions is a must. These will provide the raw data from which our analyses emerge. Particularly when looking at ensemble or collectively improvised music, a high quality recording is essential. The first time I recorded a pair of arja drummers, I had only a single handheld recorder placed on the floor between them and no musician keeping a steady beat for reference. I loved listening to this recording, and it offered some sense of the practice’s characteristic interlocking. But the speed and rhythmic complexity of the drumming, as well as the abundance of quieter subordinate strokes, made it impossible to transcribe from this recording with certainty. On my next trip to Bali, I brought separate microphones for each of the four drumheads, so that their tracks could be panned in the mix to ease transcription. I also placed a small video camera in front of each drummer, capturing a clear view of both their hands. One additional camera took in the whole scene: drummers, beat-keeper, and cycle-marking instruments as well as the singer they were accompanying. The majority of the recordings I made of reyong norot, by contrast, were of a single musician playing alone, accompanied only by core melody. This, of course, made transcription easier, but it created something of a laboratory setting in which I could not truly assess collectivity. When wanting to look at the collective aspects of norot improvising, I had only a video recording taken with a single camera. Transcription from this recording was predictably more (p.72) difficult. The lower end of the reyong, louder by nature and closer to the camera, was easy to discern. The top two positions of the instrument—quieter, and in this session played by less experienced musicians—required a great deal of repetition, slow-down, and ultimately transcription as much by visual as aural cues.

Although recording sessions were a somewhat contrived context for music making, I did what I could to make them feel natural. Food, coffee, and cigarettes were provided for the musicians, and we took lots of breaks to hang out, chat, and jam. I set the basic parameters for each session, but then let the musicians lead them. I set up the recording equipment, but in soundcheck asked the musicians to provide feedback on levels. I expressed to the performers my basic goal for a given take, but let them determine its length and structure. My reyong norot sessions always began with Alit showing me a core melody to play on the lead metallophone ugal. I then cycled that melody while he played reyong, improvising for as many cycles as he wished at one position of the instrument before moving to the next. We then played reyong together, talking about aspects of coordination and interlocking. My arja drumming sessions in Bali were longer and more involved. Two drummers, a singer, and a beat keeper were all required, but other musicians wishing to make music with these masters would inevitably come out of the woodwork to play accompanying instruments like gongs and cymbals. I let this flow naturally without organization. For each take, I requested a particular emotion or character type from the singer—which would result in different gong structures and drumming styles—but left all other details up to the performers. Throughout, the musicians might ask to listen to the recording and make comments on its quality, sometimes suggesting changes in mic placement or asking to play certain songs a second time. The photo in figure 2.14 shows Pak Tama of Singapadu listening to the early part of an arja recording session. As he made comments and suggestions, the headphones were passed to other musicians in turn. I hoped this back-and-forth and shared control of the recording session would give the musicians a sense of agency and active participation in the research process.


While the formal interview is a time-honored research tool, valuable for gathering certain kinds of data, I found it to be only marginally useful for clarifying details about musical structure, communication, and intention. Often the musicians I was working with struggled to put into words their ideas about music, and I in turn struggled to find ways to ask about abstract concepts like models (p.73)

Finding an Unspoken ModelThe Boundaries of Reyong Norot

Figure 2.14. Soundcheck in Singapadu, I Wayan Tama.

(Photo by Chelsea Edwardson, 2011. Used with permission.)

and knowledge bases, idioms and their boundaries. I found guided interviews, where specific recorded musical examples were the focus, to be somewhat more fruitful. Ingrid Monson, in her study of improvisation in jazz rhythm sections, centered her interviews on recorded excerpts from the musician’s own catalog (see Monson 1996, 17–20).33 For a practice like arja, where recordings are not readily available, I often interviewed musicians immediately following a recording session, listening back to excerpts from the session and discussing them. Having something concrete and creative to talk about helped focus our conversation. Yet the most important thing, I learned, was to let the musicians guide the discussion. As Monson notes:

While I had a general plan in mind when I arrived in the interview situation, I played each one by ear: if a musician didn’t seem to enjoy hearing an example, or seemed to find some of my questions uninteresting, I tried to listen for topics that were of interest, encourage expansion upon them, and stay out of the way. On some days, not surprisingly, I was better at this than other days.

(Monson 1996, 19)

This process could sometimes be frustrating, particularly when I had very specific questions in mind. Yet, though I didn’t always get the information I (p.74) was looking for, this tack often brought surprising new ideas to the fore. What resulted was something closer to a “dialectical ethnomusicology” (see Kippen 1992 and 1985), where musicians could point out moments in an improvisation that worked well and those that didn’t, discuss issues that they considered important, and even comment on specific improvised patterns.


Insights gained through interviews and sessions still require unraveling. And once basic musical proficiency has been reached through listening, playing, and immersion, private lessons can solidify and expand a researcher’s understanding of a practice. As Monson notes, though she was already a proficient jazz trumpeter, “the knowledge [she] gained of drumming from [lessons with jazz drummer] Michael Carvin was absolutely critical to [her] understanding of the rhythm section” (Monson 1996, 17). I found the same to be true in my studies of reyong norot and kendang arja. Perhaps most importantly, my lessons gave insight into musicians’ thought processes about their own music. When we ask improvisers to teach improvisation, we’re asking them to make choices: decisions about which patterns or ornaments or variants are the most foundational, or the easiest, or the most important, or the most interesting. We’re asking them to show us the things they know consciously about their practice and to think about presenting that practice to a cultural outsider. This process can elucidate unspoken or unconscious models and help unearth the knowledge bases surrounding them.

As useful as the music I learned during my lessons was, it was the dialogue that flowed naturally out of the learning process that provided many of the most profound insights into the collectively improvised genres I was studying.

“But you must be following rules when you improvise, right? I mean you can’t really just play anything.” I’m sitting at the reyong with Alit, feeling both hopeful and nervous as I ask this question. We’re thirty minutes into our first norot lesson, and I can sense him trying to negotiate my direct and analytical questions with his own tacit knowledge of the practice. This is new territory for both of us.

“OK, it’s like this,” he begins. “When I play norot, it comes from the base. You cannot just go anywhere,” he asserts, “not just do what you want. … It’s from the base, what I give you … I feel from that.”

“What do you mean by ‘the base’?” I ask. This could be the answer I’ve been looking for! Alit sits quietly for a minute, characteristically thoughtful as he mulls over this new question. But he finally responds: “I’m sorry. I don’t know how to explain.” (p.75)

Finding an Unspoken ModelThe Boundaries of Reyong Norot

Figure 2.15. The “base,” Oleg Tumulilingan

Finding an Unspoken ModelThe Boundaries of Reyong Norot

Figure 2.16. Comparison: composite gangsa melody (top) and reyong “base” (bottom), Oleg Tumulilingan

“Could you maybe show me the base for Oleg?” I suggest.

“Ya,” he smiles. And from his position at the low end of the reyong, he plays the melody in figure 2.15.

Although Alit’s passive knowledge of the tradition means he can’t verbally articulate an answer to my question, he’s clearly able to show me a concrete example of “the base” in action. It seems there is a consciously known model for this improvised practice after all, unspoken though it may be. And my excitement grows as it dawns on me how similar this “base” is to the gangsa norot melody I’d learned with Sekar Jaya the summer before. The biggest difference, circled in figure 2.16, is likely due to the gangsa’s melody going into a range Alit’s reyong mallets could not reach.34

It occurs to me then, a hypothesis to be tested, that fixed gangsa melodies may provide models for reyong improvisation.

A Fixed Elaboration Practice: Norot on Gangsa

A Question of Influence

One can find examples in many music cultures of known models for one voice or instrument guiding the improvisations of another. Such influence often comprises large-scale formal or aesthetic guidelines, as in traditions where song melodies, lyrics, or structures inform musical choices of improvising instrumentalists. In Balinese arja, as we will see in chapter 6, the various cyclic structures used in songs for different character types affect the density of improvised drumming. In Brazilian pagode performance, instrumentalists working with well-known melodies and song forms improvise musical figures (p.76) that emphasize cadences, fill small breaks in sung phrases, and even support lyrical content (see Stanyek and Oliveira 2011, 135–43.). Yet such model borrowing can also occur as small-scale mimesis. Pagodeiros will “sometimes vary their characteristic patterns by matching their playing to local rhythms of the vocal melody” (Stanyek and Oliveira 2011, 136). Similarly, the improvising lead drummer of an Ewe dance-drumming troupe will sometimes mimic the rhythm of a supporting drum, or invent a new rhythm that highlights a relationship between two contrasting drums, in order to bring these elements to the forefront of the texture. In each tradition, an improvising musician uses the more fixed idioms of other voices or instruments in the piece for her model. I wondered if reyong players drew their norot models from gangsa melodies in similar ways.

My thinking from this point proceeded as follows. I knew that there were direct connections between reyong and gangsa figuration. The 4-note ubit empat technique used in the warrior dance Baris (see figures 2.22.4) was often played identically on the two instruments. And I knew also that there were certain figuration styles unique to gangsa. Ubit empat’s close cousin ubit telu uses an interlocking technique that necessitates sharing the middle of its three notes between two gangsa players, impossible on a reyong where the four musicians use a single instrument. Thus, ubit telu is not played on reyong. Nor is another interlocking gangsa technique called nyog cag, where partners cover a much wider range by leaping over each other in strict alternation (see Vitale 1990, 4–5; Tenzer 2000, 216–20). If played on reyong, “the disjunct nature of the parts would cause them to cross repeatedly between players’ positions. When gangsa play nyog cag the reyong must do something else” (Tenzer 2000, 217). Musicians have not devised new ways to play nyog cag or ubit telu on reyong; they simply do not use them. That both gangsa and reyong have a style of figuration called norot hints at a level of equivalence that would require parallel realizations on the two instruments.

Although Alit could not verbalize his understanding of the norot “base,” nor confirm or deny a possible foundation in gangsa figuration, his pedagogical approach in our one-on-one lessons provided additional insight. After teaching me a new reyong pattern, Alit would often move to the lead melodic instrument ugal. Playing its elaborated core melody together with my norot, he hoped, would help me feel the musical link between the two. But whenever I faltered, forgetting the reyong pattern he had taught me, Alit would abandon the ugal melody and begin playing gangsa norot on his instrument, staring at me with an intensity that said, “Listen to what I’m playing. I’m helping you!” I found the combination of musical and ethnographic evidence for a reyong-gangsa (p.77) connection in norot compelling, and it became a working theory. It would be fifteen years before Alit confirmed it.

We’re sitting in the World Music Room at MIT, talking about the varied naming practices for Balinese interlocking techniques. Alit has just informed me that, in his village, the term norot is not in fact used for interlocking gangsa melodies. His teachers instead used “noltol”—a Balinese term meaning to pick things up one by one, but which Alit likens to chickens pecking at the ground—to describe the regular alternation between interlocking partners seen in figure 2.13.

I feel pretty deflated by this casual revelation. Every book on gamelan I’ve ever read calls this technique norot, and I’ve been working for fifteen years under the assumption that these gangsa and reyong styles must be connected. Does Alit’s insistence that norot is the wrong term for this gangsa figuration negate that link? Will I need to throw out my hypotheses and go back to the drawing board?

“What about when it’s played slowly?” I ask, referring to a technique where all gangsa players strike every note in a melody instead of sharing it with a partner. “Did they call it norot then?”

“I don’t remember,” Alit responds. “But it must be,” he continues, before unceremoniously adding, “because that’s where reyong norot comes from.”35 A decade and a half after our first conversation about the norot “base,” I stumble into the answer while looking for something else entirely. The model for reyong norot is gangsa norot!

Making the Model

If fixed gangsa norot melodies provide explicit, consciously known models for reyong norot, an overview of just a handful of their particulars can help frame reyong improvisation. The reason the Sekar Jaya gangsa players were able to figure out their norot so easily is that, unlike models in many improvised traditions that are unique to a single piece, the norot model is entirely formulaic. It begins with the calung, the core melody instrument moving at the half-note density. A norot melody, playing four notes per beat, alternates back and forth between the current calung tone and its scalar upper neighbor. It alters from this pattern only to prepare for a new calung tone.36

Every moment in a calung melody is defined by one of two possible kinetic qualities. The simpler of the two is the ngubeng, or static quality, and occurs between two successive calung tones of the same pitch. In such cases, the gangsa norot will simply alternate between the core melody tone and its upper neighbor for the full two beats. In the beginning of the Oleg core melody, the gong (p.78)

Finding an Unspoken ModelThe Boundaries of Reyong Norot

Figure 2.17. Static (ngubeng) norot on dung (u).

Finding an Unspoken ModelThe Boundaries of Reyong Norot

Figure 2.18. dung (u) to ding (i) norot shift

tone dung (u) is followed by another dung two beats later, and the gangsas follow suit with static figuration. With structural tones that align with calung bold-faced and followed by a double backslash, the gangsas play (u)//a-u-a-u/a-u-a-u//per figure 2.17.

The other melodic quality for core melody motion is majalan: kinetic or moving. This occurs in any 2-beat cell between differing calung tones, as in the dung-ding (u-i) shift following Oleg’s static start. The first half of a kinetic cell stems from its first core melody tone, alternating between that tone and the one above it just like a static cell; norot for a melody moving from dung (u) to ding (i), as in figure 2.18, also begins (u)//a-u-a-u/. The second half of a kinetic cell prepares the new core melody tone, paying no heed to the old one. It will always comprise the new tone twice, its scalar upper neighbor once, and the new tone once more, timed to land together with calung. In the same dung-ding (u-i) example, this anticipatory pickup gesture of sorts is/i-i-o-i//. The rocking motion then begins anew, this time on ding (i).

This simple set of rules lays out a single-voice model for gangsa norot figuration: a model made from a single strand of melody. Yet, although norot on the gangsa is a fixed practice, playing it idiomatically still requires indoctrination into a knowledge base. We know that the eight gangsa players are divided into four pairs, each musician taking on one of two roles. As we saw in the gangsa’s Oleg melody in figure 2.13, when the tempo is sufficiently fast, partners work together to create the single-voice model through melodic interlocking.37 The fixed formula for interlocking gangsa norot dictates that the polos (“basic”) (p.79) players sound all the current core melody tones while their sangsih (“complementary”) partners play the scalar upper neighbors. The two share the 3-note anticipation, polos playing only the new core melody tone while sangsih plays both the new tone and its upper neighbor. This neatly prescribed idiom has the two strands playing identical rhythms offset by one 16th note. For the dung (u) to ding (i) shift, with rests notated as underscores (_), the polos plays (u)//_u_u/i-i_i//; the sangsih interlocks with (_)//a_a_/i-i-o_//o. Every note of the model is sounded, either by one or both musicians in the pair, and figuration for both polos and sangsih players is fixed.

In slower passages, by contrast, the polos players do not need help playing the full melody. The function of sangsih often then becomes sounding a parallel “harmony” note, called kempyung, consistently three scale tones above polos per figure 2.19.

The dung to ding (u-i) polos pattern (U)//A-U-A-U/I-I-o-I//, for instance, is complemented by the parallel sangsih pattern (o)//e-o-e-o/u-u-a-u//. Figure 2.20 shows the full Oleg melody in parallel norot figuration, with polos notated stems down and sangsih stems up.38

These formulas for actualizing parallel and interlocking figuration become the knowledge base informing idiomatic gangsa performance on the single-voice norot model.39 Once a gangsa player understands the model and knowledge base—static and kinetic, kempyung and interlocking—she may realize any piece without the aid of a teacher; she simply follows the calung’s core melody. Yet while norot is probably the simplest figuration style to master on gangsa, though they use the same model, it is anything but straightforward on reyong.

Finding an Unspoken ModelThe Boundaries of Reyong Norot

Figure 2.19. Kempyung parallel “harmony” tones

Finding an Unspoken ModelThe Boundaries of Reyong Norot

Figure 2.20. Gangsa norot in parallel figuration, Oleg Tumulilingan

(p.80) Fixed Elaboration Turned Flexible: Norot on Reyong

Knowing that the strict single-voice gangsa norot template is a model for reyong players, I examine reyong norot by degrees of improvisational freedom. On the more conservative end, it involves combining basic model notes, kempyung “harmony” tones, and rests to create a variety of realizations without yet “breaking the rules” of the model. These formulaic substitutions are comparable to the free interchanging of subordinate with main drum strokes seen in kendang arja in chapter 1, and thus fall under the rubric of interpretation. On the more extreme end are those patterns created from a broadened conceptual space for norot figuration delineating idiomatic deviations from the model. These more drastic departures are based on improvisatory processes of embellishment, recombination, and expansion. For the remainder of this chapter, I explore how the particularities of the reyong itself create both boundaries and freedom for improvisation, before then seeking to derive a grammar of guidelines for idiomatically actualizing these improvisatory processes in chapter 3.

The Big Picture

Much of the analysis to follow is cellular, examining 2-beat units of sound at one of four positions on the reyong. These must first be understood as part of a fuller picture. As Simha Arom cautions: “Once the diverse elements have been disassociated the one from the others, they may lose their meaning and even their identity. It seems necessary, therefore, to always have at hand a global image of the aural document being analyzed.”40 Watching four experienced reyong musicians improvise norot is aurally and visually exhilarating: a flurry of eight mallets racing together over the twelve kettle gongs of the instrument, playing and damping with the same action. Each musician is at some points playing an independent line, at other moments interlocking with the player above or below, or even with an imaginary partner just outside the instrument’s range.41 Their interconnected streams of music dance across the reyong in continuous motion; the global impression is both more intricate and less simply elegant than gangsa norot. A careful eye will catch a small 3-note ascending passage being shared seamlessly between two musicians, or an alternation of two adjacent notes distributed between partners, much like interlocking gangsa players. And always, kempyung tones are an ever-present enrichment of the music’s “harmonic” pitch makeup. (p.81)

Finding an Unspoken ModelThe Boundaries of Reyong Norot

Figure 2.21. Improvised reyong norot, Teruna Jaya (2 cycles)

A brief improvised excerpt, played by musicians from the Sanggar Çudamani arts collective in the summer of 2016, will illuminate some general features of reyong norot. Figure 2.21 shows two successive cycles from a moderately paced improvisation on one of many cyclic melodies from the dance piece Teruna Jaya (see Video 2. Video 3 shows a longer excerpt of the improvisation).

Like gangsa norot, reyong norot creates a texture of almost continuous motion. Two notable exceptions in this excerpt are the gong tone and the dung (U) at the cycle’s midpoint, both of which are almost always left empty by all four players. A common technique on the reyong also occasionally used by gangsa players, this rest creates space for the hanging gongs to ring through the texture, emphasizing their structural importance. Yet what becomes equally apparent is that, unlike gangsa players, these reyong musicians do not perform fixed rhythms or note combinations; rhythmic variety is central to the reyong’s norot idiom. “It’s not noltol,” Alit laughs, referring to the gangsa’s technique of strict alternation named for its likeness to chickens pecking at rice in the dirt.42 Thus, each time they play through the Teruna Jaya melody, which will be cycled several times before the composition shifts to a new melody, the four (p.82) Çudamani musicians all play unique variants on the model. Although partially dictated by instrument idiom, these are also determined by the stylistic preferences of individual musicians. The second position player, I Madé Supasta, prefers denser figuration, filling almost every subdivision in many 2-beat cells. Alit’s eldest brother I Dewa Putu Berata in first position, by contrast, employs more rests. And though the principle of octave equivalence means that the first and third position players, and those of the second and fourth positions, play similar figuration, they are almost never identical. The existence of both freedom and boundaries, and the negotiations between idiomatic options and individual preferences and choices are all in play here. These are informed by the physical limitations of the instrument, limitations that necessitate a refining and expanding of the norot knowledge base.

An Expanded Knowledge Base for Reyong Norot

The Influence of Instrument Construction

Instrument construction and playing technique are central to the knowledge base of any music tradition, improvised or otherwise. Which idioms or common motives may be seen as a choice between equally plausible alternatives; which are necessities or likely eventualities of instrument technology? The limits of an average vocalist’s range and the physical challenge of singing large intervals with precision means that many vocal traditions feature stepwise motion and small intervallic leaps in ranges little more than an octave. Construction of the heads for Indian tabla and mridangam drums encourages idioms with large timbral palettes; the use of finger strokes in their playing techniques allows for faster rhythmic passages than in Balinese drum genres, where musicians must rely on slower wrist and elbow joints. The construction and playing technique of the Shona mbira dzavadzimu likewise determines many of its idioms. Two left-hand manuals an octave apart are played with the left thumb only. The right thumb plucks the first three keys of the higher-pitched right-hand manual, while the remaining keys are played with the right index finger.43 The physical arrangement of the mbira’s keys encourages disjunct melodies as players alternate right-and left-hand strokes, while right-hand fingering techniques dictate a common formulaic variation in which index-finger harmony notes are added three keys above existing thumb notes.44 Many of the most basic features of reyong norot improvisation, too, are artifacts of instrument technology and playing technique.

(p.83) The Limits of Range

A gangsa player is able to realize uniform norot figuration on any core melody tone because she has sole control over her instrument’s two-octave range. A reyong player, by contrast, must share his instrument with three other people, each controlling only a few of the notes in the scale. As shown in figure 2.22, the first position penyorog player controls a 3-note range from deng (e) to dang (a), extending to ding (i) only when his neighbor is not using it. This latter player, the pengenter, controls the 3-note range from ding (i) to deng (e). The third position ponggang player often has the most abstract figuration, limited to just two tones: dung (u) and dang (a). Like her penyorog counterpart, she can only play the ding (i) above when the player to her right is not using it. This last player, the pemetit, has control over the widest range of pitches: the four notes from ding (i) to dung (u).45

Setting aside improvisational freedom for a moment, a few observations will clarify the effects of this reyong construction and practice. Because of the limited ranges available at each position, unlike on gangsa there are no fixed roles in reyong interlocking. Actualizing the (u)//a-u-a-u/i-i-o-i//contour of Oleg’s dung (u) to ding (i) shift requires a different kind of teamwork. Only the first and third position players have access to the dung (u) and dang (a) required for the first half of this pattern, while their neighbors control the ding (i) and dong (o) needed for its second half. Other patterns require sharing on a smaller scale, similar to interlocking on gangsa. Norot figuration for a static melody on dang (a) alternates between that pitch and the ding (i) above it. While the first and third position players control dang (a), it is the second and fourth that play ding (i). Performing the (a)//i-a-i-a/i-a-i-a//rocking motion of the model thus requires alternation between reyong partners. (p.84)

Finding an Unspoken ModelThe Boundaries of Reyong Norot

Figure 2.22. Ranges of the reyong

Finding an Unspoken ModelThe Boundaries of Reyong Norot

Figure 2.23. Flexible pairings on reyong

Yet, unlike in interlocking gangsa patterns, no two positions on the reyong interlock exclusively with one another. The second position pengenter player, for instance, interlocks at times with the third position ponggang above, at other times with the first position penyorog below. While a (a)//i-a-i-a/motion requires the second player’s ding (i) to complement the first’s dang (a) below, for an (e)//u-e-u-e/contour, her deng (e) instead interlocks with the third player’s dung (u) above. These patterns are shown in figure 2.23 with the two voices not under consideration notated using smaller diamond noteheads.

The octave equivalence so fundamental to reyong grammar is equally fuzzy. The low deng (e) of the first player and the highest dung (u) of the fourth player expand the ranges of these outer parts in relation to their inner counterparts. The fourth player’s high dung (u) is “often judiciously used in what might otherwise be inhospitable tonal situations as a way to boost the prominence of the reyong in the texture and add extra cross-rhythmic vitality” (Tenzer 2000, 216). Yet even while abiding by the model, these extra notes given to the outer players influence the contours of their elaborations. In figure 2.23, the (e)//u-e-u-e/figuration necessarily shared between second and third position can be performed alone by either of the other players. The reyong players’ limited ranges demand creativity, different for each of the four musicians. In the next chapter, I examine how these musicians apply methods of interpretation, embellishment, recombination, and expansion to their limited pitch palettes.

Comparatively Analyzing Across the Instrument

Even just a cursory look at the improvised Teruna Jaya passage in figure 2.21 reveals that, despite the relative freedom allowed reyong players in model-abiding norot, these musicians are still constantly playing patterns that diverge from (p.85) the model. In chapter 3, I analyze these more complex patterns in an effort to uncover the conceptual space that guides them. In order to facilitate comparison of idioms across the instrument, each pattern is conceived in something of a moveable do system. Because of the end-weighted conception of beats, the second of two core melody tones in any 2-beat cell, regardless of pitch, is labeled 0 (zero). All other notes in the 5-tone scale are labeled 1, 2, 3, or 4, in relation to the second tone and without reference to relative frequency distance or octave placement. The note directly above any core melody tone 0 is labeled 1; the note one scale tone below is labeled 4. Thus a melody moving from deng (e) to dong (o) is considered a 1–0 shift while one moving from dong (o) to deng (e) is a 4–0 shift.

Figure 2.24 shows all possible core melody shifts compressed into the one-octave range of the calung. Figure 2.25 shows the Oleg melody with each of its core melody shifts labeled for analysis, revealing many areas for potential comparison in 1–0 shifts across its range. (p.86)

Finding an Unspoken ModelThe Boundaries of Reyong Norot

Figure 2.24. All core melody shifts labeled for analysis

Finding an Unspoken ModelThe Boundaries of Reyong Norot

Figure 2.25. Oleg’s core melody shifts labeled for analysis

Finding an Unspoken ModelThe Boundaries of Reyong Norot

Figure 2.26. Comparing improvised cells analytically

This numbering system simplifies analysis by encouraging pattern identification based on the number of scale tones traveled as opposed to the actual pitches used. Because of the “movability” of norot contours, I can search for commonalities of design in reyong improvisation between players. I know, for instance, that the (o)//e-o-e-o/i-i-o-i//realization of a 1–0 dong-ding (o–i) shift is only possible at the second and fourth positions of the instrument where those three tones fall within range. A more inclusive analytical system allows direct comparison of these patterns with those of their first and third partners on other tones. Figure 2.26 shows two cells from Alit’s improvisations using only notes from the model; they are labeled with numbers for analysis. Although seemingly dissimilar, their numbering reveals structural parallels.

In Broader Scope: Working from a Known Model

From reyong norot to Ewe dance-drumming, Karnatak kriti to big band jazz, many improvisatory traditions are based on consciously known, overtly conceptualized models. Uncovering these models in oral music cultures can be a complex process, where ethnography plays as important a role as analysis. Once discovered, however, such models allow us to bypass the kind of early stage, trial-and-error hypothesizing I experienced in my search for kendang arja’s unconscious models discussed in chapter 5. Our task as analysts, then, becomes to work in partnership with our collaborators within the culture, both to decipher their knowledge bases and to formulate suitable categories of divergence from known or hypothesized models. This research should be approached through a flexible combination of discourse, observation, and musical analysis appropriate to the culture being studied.

(p.87) In his examination of improvisation on known jazz tunes, we saw Berliner (1994) first being given the concept of “levels of intensity” by alto saxophonist Lee Konitz, generating the insider categories of interpretation, embellishment, variation, and improvisation. Berliner then delineates the rough boundaries of these classes through examples and descriptions of jazz idioms. David Locke by contrast develops categories for improvisation in the Ewe music and dance form Gahu through “a synthesis of the playing styles of [his] teachers and [… his] years of exposure to performances of the piece” (Locke 1998, 3). In his 1998 Drum Gahu, Locke presents specific processes for generating “improvised variations” on both lead and response drums, proposing his own terminology for techniques he has observed. These include methods of interpretation, like “variation by accentuation,” as well as techniques of embellishment, like “variation by subdivision.” Locke also identifies recombination techniques such as “isolation and repetition” and expansion techniques like “rephrasing” and “metric modulation.”46 Although seemingly quite different studies, both Locke and Berliner begin their analyses with a known model. Musical processes are then discovered, and locally appropriate categories derived through a combination of hands-on experience with the music, conversation with and observation of its practitioners, transcription, and musical analysis.

In the next chapter, I explore the many ways that reyong norot players diverge from their known model and its knowledge base in improvisation. Drawing on local music theory concepts, both formal and informal, as well as on my transcriptions of improvised performances, the analyses look to create a culturally relevant grammar of guidelines for norot improvisation and provide a test case for identifying and examining improvisatory processes across cultures. (p.88)


(1.) I Dewa Ketut Alit of Pengosekan, Gianyar.

(2.) On these techniques, called kecekan and ocak-ocakan respectively, see Gold 2005, 64–67.

(3.) Often generically called kotekan, in different musical contexts, these characteristically Balinese interlocking techniques are variably termed kilitan, candetan, reyongan, and so on (see Bakan 1999, 53). The terms vary by instrument, technique, region, and generation. In the current case study, I generally avoid an overarching term, preferring to refer to specific techniques by name (norot, empat, noltol, etc.); this is the commonly accepted approach among most Balinese musicians (I Dewa Ketut Alit, personal communication April 2017).

(4.) While there are many different versions of the Baris reyong melody, each is built from the same strict principles of empat interlocking.

(5.) For those unfamiliar with gamelan, this small range lacking both phrasing and rhythmic variety might not look like a typical melody. However, it is often sung by Balinese musicians for pedagogical purposes and, though formulaically composed, is indeed considered “melodic.”

(6.) On the effects of familiarity and experience (as well as age and tempo) in detecting melodic alterations, see Dowling et al. 2008. Closely related studies on the cognition and psychology of musical expectation include Meyer 1957; Huron 2006. Pearce and Wiggins (2012) extend these studies to include computational and probabilistic modeling.

(7.) On partimento, see Sanguinetti 2012. On pagode, see Stanyek and Oliveira 2011.

(8.) The knowledge base of each of these traditions is much more detailed and nuanced; here I am only concerned with the basic model.

(9.) Most of the musical examples in this chapter and the next stem from recording sessions I did with I Dewa Ketut Alit in Vancouver, British Columbia, in the winter and spring of 2002. (See Tilley 2003 for full transcriptions and more detailed analyses.) In these sessions, Alit improvised norot for several pieces, both traditional and newly composed, at each position of the reyong. He then taught me a handful of patterns for each and we discussed their interlocking qualities. This research has been supplemented by lessons and conversations with Alit and other musicians in the interim, observation of performances in Bali, and a 2016 video recording session of a full reyong complement in Pengosekan, Gianyar. Thus contextualized, examining Alit’s earlier improvisations at the cellular level can offer insight into the unspoken rules that guide them. Yet no single study provides a complete picture, and research in other regions would doubtless yield additional results.

(10.) This ordering is partially inspired by challenges I faced in my reyong norot research. Coming to Step 5 far too early, I struggled with comprehension and interpretation of many basic musical facts that the order of events suggested here could have mitigated.

(11.) Baily (2001) outlines some of the weaknesses of this term, and Hood himself noted that “musicality” is more far-reaching.

(12.) This scale, selisir, is a mode of the Javanese 7-tone pélog scale. In notating gamelan thusly, I join a tradition going back to McPhee (1966). Yet the actual pitch distribution of selisir varies significantly among villages and gamelan makers.

(p.298) (13.) Gamelan makers tune pairs of instruments to be slightly “out-of-tune,” so that the difference tones between their frequencies create a beating effect. This is done exactingly across the ensemble so that every note in every pair pulsates at the same rate.

(14.) Blake (2010) discusses the benefits of various kinds of listening, from background listening to analytical dissection. See also Borgo (2005, 26–30) on different listening styles.

(15.) Bakan extensively explores the value of “learning to play” in his research on gamelan beleganjur (1999). I originally adapted the phrase from Baily’s (2001) “learning to perform” as a more all-encompassing replacement for the term bi-musicality. Another good option, used by Monson (1996, 131), is poly-musicality.

(16.) This is true whether physically performing or simply moving and gesturing while listening. See also Gibbs 2006; Leman 2007; Godøy and Leman 2009.

(17.) One may argue that too much embodied solidarity with musicians could lead to a loss of objectivity, but the general consensus is that the benefits of performance far outweigh its potential pitfalls.

(18.) It may also have cognitive benefits. Recent research in linguistics has found improved divergent and creative thinking, greater mental elasticity and metalinguistic awareness, and heightened communicative sensitivity among bilinguals (see Cummins 1998; Lazaruk 2007; Bialystok 2001). I propose that musicians with performing proficiency in multiple musical systems have likewise honed the mental elasticity needed to find patterns and understand structure in unfamiliar genres, to read musicians’ physical and aural cues, and to recognize multiple options for musical success. These skills can then be brought to new musical systems, increasing the potential for subtle and sensitive cross-cultural analysis.

(19.) Many researchers of Balinese music, after Jaap Kunst, term cyclic music punctuated by fixed patterns of gong strokes colotomic. In an effort to encourage cross-cultural observation, I prefer the broader term cyclic, and see the gongs’ strokes as part of a stratified polyphony.

(20.) This common gong structure is called Gilak.

(21.) On insights into diverse apprenticeship styles gleaned through practical studies in Afghani dutâr and rubâb, see Baily 2001.

(22.) Interestingly, students in Study Abroad programs showed less improvement in a number of fluency factors than those in Intensive Domestic Immersion programs. The authors credit this difference to the latter group’s dedication to consistently speaking, listening to, reading, and writing French outside of class.

(23.) Note: because both gangsa and ugal range from dong (D) to the ding (C#) two octaves above, their full range is transcribed O-E-U-A-I-o-e-u-a-i, with lower octave pitches notated using uppercase letters. Range is not differentiated, however, in reyong transcriptions. A ding (i) in the lower octave is considered functionally equivalent to one in the higher octave, and reyong players can actualize the same figuration regardless of contour or octave placement in ugal and gangsa melodies.

(24.) See Ellingson 1992; Marian-Bălaşa 2005; Seeger 1958; Stanyek. 2014; Tilley 2018. See England et al. (1964) for a revealing demonstration of the subjectivity and creativity of both transcription and analysis.

(25.) On microtiming, individual musical roles, and ensemble entrainment, see Polak, Jacoby, and London 2016.

(26.) Most scholars attempting staff notation of Balinese music use a similar system, but as previously mentioned, its beaming is problematic. Lisa Gold attempts a novel approach to the (p.299) end-weightedness of cycles if not of beats, placing the full beat of the gong tone before the repeat sign (e.g., Gold 2005, 61). Yet Balinese musicians do not consider the three 16th-notes following the gong tone to be part of the old cycle; they are the beginning of the new one. Staff notation thus seems incapable of being bent to an end-weighted conception. This is further complicated by gamelan’s rhythmic stratification. Where different instrument groups move at different rhythmic densities, each group could be said to “begin the new cycle” independently, when first articulating a pitch not belonging to the old cycle. While the music is not felt in this divided way, these density strata create notational complications: Gold is forced to tie the calung and jegogan’s gong tones over the bar line at an artificial point. Every notational approach has both benefits and intrinsic representational challenges. I hope that the breadth of approaches throughout this book will demonstrate their relative strengths and weaknesses.

(27.) Tenzer (2000) examines these musical relationships in depth. Here, only the details relevant to an analysis of reyong norot are discussed.

(28.) One may argue that, because each instrument derives its part from the same melody, the music is technically heterophonic. However, both perceptually and experientially it is polyphonic.

(29.) True to an end-weighted conception, the penyacah track the ugal every other beat (at the calung density), then anticipate each of these structurally important tones with the scale degree either above or below (compressed into a single octave), thus playing every beat. On this and other compositional techniques, see Tenzer 2000, chap. 6.

(30.) The imperfect alignment of the “A” in penyacah and calung/ugal represents a composer’s choice to sacrifice vertical convergence for a smoother, more interesting melody line in all instruments.

(31.) Also termed nyok cok (sometimes spelled njok tjok). On this and other interlocking techniques, see Gold 2005, 58–64; Tenzer 2000, 213–31; Vitale 1990.

(32.) On tacit versus explicit knowledge and its effect on learning and teaching, discussed in more detail in chapter 5, see Han and Ellis 1998; Hall 1980; Brinner 1995.

(33.) Monson owns that this technique was more successful with some musicians than others.

(34.) These and other note choices will be made clear in subsequent sections. Octave equivalence can be seen following the klentong (t) stroke: Alit stays in the low octave on reyong while the gangsas go high.

(35.) Conversation with I Dewa Ketut Alit, April 2017.

(36.) Sometimes a composer will opt for more static norot realizations, tracking the slower-moving core melody instrument jegogan (discussed in chapter 3). Tracking the calung, however, is the norm.

(37.) Interlocking melodies can be played at slower tempi, particularly in slower-moving sections of a piece or in older, more classical styles. In these instances, the choice between parallel and interlocking figuration is stylistic.

(38.) Note that sangsih players fall into unison with polos on the top three pitches of the instrument, where kempyung tones are unavailable.

(39.) Although gangsa norot in parallel figuration features two voices, because the same compositional model is used for interlocking, I consider it a single-voice model.

(40.) “Les divers elements … une fois dissociés, peuvent perdre leur sens et jusqu’a leur identité. II apparaît donc nécessaire d’avoir constamment sous les yeux une image globale du document sonore destiné à l’étude” (Arom 1969, 172).

(p.300) (41.) Interview with I Dewa Ketut Alit, February 15, 2002.

(42.) Conversation with I Dewa Ketut Alit, April 2017.

(43.) See Titon 2009, 125, for an image of the mbira with keys labeled.

(44.) Of course this is something of a chicken-and-egg theory; we don’t know whether melodic preferences influenced the structure of the instrument or vice versa, and it is as likely some combination of the two.

(45.) For other techniques, like empat, it is not uncommon for reyong players to go outside these ranges; the penyorog, for instance, frequently plays dong (o). This flexibility is not employed in norot.

(46.) See Locke 1998, 48–68 on response drum improvisation and 74–123 on lead drum improvisation.