Analytical Approaches to World Musics
ISSN 2158-5296
Zuckerberg et al. 2026
AAWM Journal 14/No. 1 (2026)
Hadza Public Vocal Music: A Preliminary Analysis of Rhythm and Meter
Gabriel A. Zuckerberg, Shani M. Mangola, Aniruddh D. Patel, Jason C. Rosenberg, and Alyssa N. Crittenden
Hadza, Tanzania, singing, meter, isochrony, language, archive, collaboration
This article introduces a music-focused, interdisciplinary project among scholars and Hadza community members in Tanzania. The greater project has the dual aims of studying Hadza music and preserving it for future generations of community members. Here, we provide a brief overview of existing literature on Hadza music-making, a topic that has been peripheral to over a century of scholarly publications on the community. This overview historicizes our work with existing documentation and previous recordings of Hadza singing. We then introduce our methodology for the current study, combining observations from ethnographic dialogue with discussions of rhythm, meter, and musical form. We also provide four detailed song analyses from our corpus of twenty analyzed song recordings. We chose to feature these four songs for their representation of singers’ age and gender, and of various rhythmic and metrical idioms. Rhythmic emphasis is a wide-reaching aspect of musical participation, expressed in singing through embellishing beats (clapping, tapping, etc.), in dance through footsteps, and in language through phonological stress patterns. Our transcriptions are structured by singers’ beat-marking in our initial recordings and, later, by supplementary recordings prompted by our initial analyses. Whereas the included analyses are not a large enough sample to make empirically robust generalizations about Hadza musical style, much can be learned about the challenges and opportunities in this style through the analysis of rhythm and meter in Hadza singing. Of the twenty song transcriptions, eight are in simple meter, eight in compound meter, two in alternating meters, and one in an additive meter; among them, at least ten are fully isochronous and four are non-isochronous. Most thematic to the coauthors during analysis was their observation, especially in compound meter songs, of beat-initial short-syllable stressing: that is, the placement of a short-duration syllable in a short-long syllable pair on the first pulse of a beat. We contextualize these patterns in relation to meter across our corpus, our current ethnographic data, and the Hadzane language. In addition, several songs stand out as having idiosyncratic meters and parallel phrases with variable durations. While cultural and linguistic context are not sufficient to explain musical phenomena, they help us understand how the apparent stressing pattern provides organization to musical “flexibility” and “flow.” We conclude each section of analysis by proposing future steps for further research, demonstrating the potential of the larger project.
Gabriel A. Zuckerberg is a doctoral candidate in Musicology and Ethnomusicology at Brown University.
Shani M. Mangola is the co-founder and director of the Olanakwe Community Fund.
Aniruddh D. Patel is Professor of Psychology at Tufts University and Fellow in the CIFAR Program in Brain, Mind, and Consciousness.
Jason C. Rosenberg is a composer and music cognition researcher.
Alyssa N. Crittenden is Professor of Anthropology at University of Nevada, Las Vegas.
Click for DOI, citation, PDF version
Background
History, Ethnography, and Societal Change
[1] Hadzabe[1] are an Indigenous hunter-gatherer community residing in northern Tanzania. Hadza lifestyle has been changing over the past century, as both evidenced and influenced by Euro-American, national, and local actors. While about 1,000-1,300 individuals self-identify as Hadzabe and speak Hadzane, only around 250 continue to forage for wild foods as a significant portion of their diet. The remaining majority of the community has adapted to a predominantly agricultural diet and mix foraging with other subsistence activities, including wage labor (see Marlowe 2010). Virtually all adult members of the community purchase, donate, or trade foods for at least part of the year, even in remote bush camps (Crittenden, et al. 2017; Pollom, et al. 2020a, b). Still, a hunter-gatherer lifestyle remains emblematic of Hadza folkways and has been key in structuring normative social roles.
[2] As is the case for most mixed-subsistence foraging and former-foraging populations across Africa, the wild food resource base (i.e., undomesticated plants and animals) in northern Tanzania is currently threatened by increased population pressure from neighboring communities, climate change, and expansion of wildlife conservation areas (Jodoin 2017; Gibbons 2018). European colonization and local land encroachment over the last century have constricted Hadzaland into the Yaeda Valley, south of Lake Eyasi (Moshi, et al. 2019). Recent studies reveal how the ongoing reduction of Hadzaland is intertwined with climate change, increased economic integration, and changes in lifestyle (Safari, et al. 2021; Joachim and Zahor 2022; Shadrack and Mwalilino 2022). This locale is situated in the borderlands among several communities: Datoga, Isanzu, Iraqw (Mbulu), and Maasai. Despite the rapid pace of ecological, political, and economic change that the Hadzabe are experiencing, their cultural identity and self-identified expressive culture persist.
[3] Ethnographic accounts of Hadza culture have been published since the early 1900s (see Bleek 1931a, b), and the history of this literature has been outlined in recent publications (Blurton Jones 2016). The majority of published ethnographic research over the past century has largely focused on how Hadza diet and ecology impact life history, health, and reproduction (Marlowe 2010). Some work has also represented their music-making, but it either is not devoted specifically to Hadza culture (e.g., Hatfield 2019) or is popular or touristic writing (e.g., Stephenson 2000; Peterson, et al. 2012). This writing primarily contextualizes Hadza music-making in life and ritual, but it offers scarcely anything about the audible elements of musical repertory and style. For example, Hatfield recounts overseeing a recording session with a Hadza singer and “choir,” including the characteristic antiphony (call-and-response between a leader and group) of the repertory they recorded. She mentions their use of ankle bells and a police whistle to accompany certain songs (Hatfield 2019, 59–60). The remaining account of the recording session focuses on the cultural values expressed in the songs’ lyrics (ibid, 61–4). Interestingly, Hatfield notes a group of young men performing a hip hop song and routine, observing that “the stylistic differences were apparent and clearly modeled after hip hop music they had likely heard and seen on a cell phone” (ibid, 63–4). Marlowe, who published an entire monograph on the Hadza, dedicates less than a page of prose to his accounts of music-making. He touches on the limited use of instruments as well as improvised singing and dance (Marlowe 2010, 67–8).
[4] There are many open questions about Hadza music-making, including how it has shifted over time, and particularly over the past three decades with increasing ethnotourism (tourism of “traditional” cultures and lifestyles). Since the 1990s, Hadzabe have been progressively making more of an income from a local ethnotourism industry, which is largely mediated by local non-Hadza Indigenous groups. Visitors travel to remote areas of Northern Tanzania to enjoy “‘traditional’ Hazda [sic] dance and experience shooting arrows” (Yatsuka 2015), at an increasing pace in the past few years. Given the more frequent interface with tourists, Hadzabe working in ethnotourism are facing a new set of demands, obstacles, and possibilities in how they represent their own expressive culture, in which music and dance play a central role. As such, Hadzabe are making ongoing decisions to maintain and emphasize certain aspects of their musical style and repertory for both themselves and tourists, and to adapt musical practice in their shifting social landscape. Meanwhile, members of the Hadza community have expressed enthusiasm to coauthors Alyssa Crittenden and Shani Mangola (a Hadza community member) for this current study of contemporary Hadza music-making. Therefore, this article presents Hadza music, its current developments, and what can be learned from a preliminary analysis of Hadza singing.
Linguistic Research
[5] Hadzane is widely considered to be a language isolate, meaning that it is unrelated to all other existing languages, including the Nilotic, Cushitic, and Bantu languages that are spoken near Hadzaland (Pennisi 2004). It is also one of the few languages unrelated to the Khoisan languages that uses click consonants (Sands, et al. 1996). Hadzabe live near the Sandawe, who speak yet another language isolate with click consonants (Sands 2013). Despite any such similarities, Hadzane is genealogically unrelated and mutually unintelligible with any other languages, near or far. Swahili is the second language of the community and is spoken by most adult members of the community and, increasingly, children. It is commonly spoken to interact with neighboring ethnic communities, researchers, tour companies, and Tanzanian government entities.
[6] South African anthropologist Dorothea Bleek published a brief cultural and linguistic account of Hadzabe in 1931. She transcribed and translated the lyrics of four songs, one of which is discussed in this article as part of an overview of our analytical corpus. Given no prior orthography for Hadzane, Bleek chose a transcription scheme based on the International Phonetic Alphabet (Bleek 1931a). In the mid-20th century, Margaret Bryan, Archibald Tucker, James Woodburn, and later Bonny Sands transitioned to an orthography designed for the Khoisan languages. Hadzane and the Khoisan languages’ phonetic commonalities made it convenient to standardize a single orthography. Yet, the failure to demonstrate a genealogical link between these languages in the following decades motivated further development. Founded on Bonny Sands’s fieldwork, Kirk Miller and Mariamu Anyawire wrote A Hadza Lexicon, introducing a new orthography based on Romanized Swahili. Guided by Anyawire’s intuition as a native speaker of Hadzane, it is meant to be “practical for the Hadza rather than intuitive to the linguist” (Miller and Anyawire 2007, ii). In consensus with these authors’ priority of accessibility, we decided to use their orthography in our writing and transcriptions.
Representation of Hadza Music
[7] Aside from a few published recordings and descriptions of songs in broader ethnographic works, there is no extant literature with a dedicated focus on Hadza musical practices. Indeed, ethnomusicologists seldom recognize Hadzabe’s existence among the multiplicity of surrounding cultures. When they do recognize Hadzabe, they designate Hadza music as a stylistic unknown embedded within broader cultural regions. When A.M. Jones (1959) sought to categorize African musical regions by their use of harmonic intervals, he placed Yaeda Valley in a patch of ambiguity, merely suggesting that nearby cultures use harmonic fourths, fifths, and unison (Jones 1959, facing p. 230). Similarly, when Alan Lomax (1968) attempted to categorize large-scale cultural regions in his Cantometrics theory, his boundary line between “Upper Nile” and “Northeast Bantu” renders the Yaeda Valley in a liminal space between the two (Lomax 1968). Thus, to our knowledge, the early African ethnomusicology literature offers no Hadza-specific precedent or insight relevant to the present study.
[8] The most basic historical written accounts of Hadza music-making, which we might compare to oral history and present-day ethnography, come from German and British colonial reports. British colonial officer F. J. Bagshawe observed that Hadzabe did not play any musical instruments during his visits between 1917 and 1923 (Marlowe 2010, 21). Thus, it would seem the voice is the only musical instrument native to Hadza culture that is still in use; yet, the “traditionality” of a few material instruments is disputable. One of our interlocutors, an elderly man, indicated that Hadzabe used to play a variety of traditional musical instruments, none of which exist today. With a few exceptions, Hadzabe continue to sing without instrumental accompaniment, and only a minority of individuals play instruments in other contexts. These instruments were all introduced by other local communities (ibid, 67) and are generally limited to touristic performances. The zeze, a one- or two-string gourd fiddle adopted from local Isanzu musicians, is becoming popular among Hadzabe. Some individuals play the mbira, though it is uncommon. James Woodburn attests to the zeze and mbira’s dissemination from the Isanzu to Hadzabe in a 1970 publication, but the decades-older history of their intercommunal trade provides reasons that this may have occurred earlier (ibid, 34–5). Finally, there are qqingilibi’i (singular qqingili), “bell-like seed-pod rattle[s]” that dancers wear on their ankles for special occasions (Miller and Anyawire 2007, 181) and sometimes for touristic performances. Hadzabe accompany a song with zeze or qqingilibi’i for the sole musical function of embellishment: the zeze follows the melody as it is sung, and qqingilibi’i amplify the beat as it is danced and/or clapped.
[9] While these earlier remarks on Hadza music may be intriguing, they do not discuss the methods used to collect data. Neither do they mention the colonial history of Tanzania and the long relationship between Hadzabe and the postcolonial research landscape. We contend that, in order to understand Hadza music-making in the present, it is necessary to understand the German and British colonial and American postcolonial legacy undergirding the extant body of academically published knowledge on Hadza culture. After the early, explicitly colonial era of Hadza ethnography, the bulk of continued scholarship remained congruent with its problematic past. Ethnographers of various disciplines—including biological anthropology, psychology, and cultural anthropology—have been driven by a focus on the biological and cultural distinctions between Hadzabe and other ethnic groups. We raise this issue not to reinforce a narrative of Hadzabe as permanently marginalized (see Fabian 1990), but to confront how scholarship has overwhelmingly approached Hadza life as a stand-in for human ancestry. Barrett (2022) terms this “the ancestral gambit,” a tendency to reduce living people to proxies for evolutionary speculation (see also Mangola, et al. 2022, 750, 725). Instead, our research collective emphasizes the importance of representing Hadza expressive culture on the community’s own terms, as a dynamic field of practice deserving attention in its own right. We hold that Hadza culture proves to be a worthwhile field for deeper ethnographic study and musical analysis.
[10] A historiography of Hadza music research becomes more explicitly precarious when we shift our lens from observations of material culture to musical performance and style. Though descriptions of material culture may be influenced by an ethnographer’s bias, descriptions of musical performance and style are especially vulnerable to a researcher’s habitual listening practice. Such habits are shaped by the listener’s previous musical knowledge and are guided by the aspects the listener attends to or actively listens for. The purported legitimacy of cross-cultural musical listening might be traced back to music criticism, or the aesthetic evaluation of music, and the positivistic use of empiricism in music research. Given the normative status of music criticism in Western classical music scholarship, aesthetic judgements may be accepted as legitimate interpretations of musical performances, even when the evaluator has no other experience with the musical practice from which the performance comes. In more conservative ethnomusicology, this is legitimized through the discourse of music as a human “universal,” or a homologous element across cultures that can be understood by anyone by virtue of being human. Thus, acts of musical listening would be treated as empirical data. While this viewpoint is no longer conventional in mainstream music research, it has arisen and has remained unaddressed in the case of Hadzabe.
[11] In the field of systematic musicology, this latent universalism has shaped scholarly knowledge of Hadza music-making. We see the issue of listener reflexivity in one of the few literary representations of Hadza music by musicologist Victor Grauer, Alan Lomax’s former collaborator on the Cantometrics analytical method. Grauer writes the following in his book examining music as a lens to the origins of human culture in Africa:
“To my knowledge no one has ever done a systematic study of Hadza music, so my impressions are based exclusively on [a] particular set of recordings, which might not be fully representative. On this basis, it would seem that Hadza singing, like Hadza autosomal markers, is highly distinctive — very beautiful indeed, but quite different from any other type of sub-Saharan African music with which I am acquainted.
Moreover, the Hadza tend to sing in relatively free, loosely coordinated, rhythms, and the rhythmic relation between the accompanying handclaps and the voices doesn’t appear to be very clearly coordinated — all of which is very unusual for Africa south of the Sahara, where rhythms and rhythmic coordination tend to be clearly defined and precise…” (Grauer 2011, 85–6).
Grauer’s description of Hadza music reflects a perceived aesthetic of disorganization. Listening to a particular set of sound recordings leads him to a suspect characterization of Hadza music as heterophonic. This characterization does not align with our analysis below. Grauer’s attention is motivated by a version of systematic musicology that prioritizes data collection and comparison over the social and pragmatic contingencies of music-making and sound recording. We acknowledge Grauer’s representation of Hadza music as an influence on the reality of public knowledge of Hadza music, but not as an academic precedent, given that no analyses of musical structure or context were presented.
[12] We also sound a note of caution about research that seeks to generalize about hunter-gatherer music via research on Hadzabe. Alice Rudge (2019) critiques the trend of musicologists to focus on Africa when sampling musics of hunter-gatherer societies, without comparison to hunter-gatherers in, for example, Southeast Asia. Conclusions showing a homology between lifestyle and musical style, such as an egalitarianism and vocal polyphony (ibid, 165), may be a result of such restricted sampling. Rather than being linked by definitive musical characteristics like polyphony, Rudge suggests that hunter-gatherers’ musics are comparable in how their egalitarianism lends itself to an “aesthetic of flexibility.” For such societies, there is a necessary balance between cooperation and autonomy, but the way this is achieved (i.e., musical “flexibility”) varies. For example, the Batek of Malaysia tend toward individual autonomy in music-making, while the Ju|’hoansi of Namibia practice communal medicine dancing (ibid, 178). This perspective is supported by a cross-cultural analysis of singing by Shilton, et al. (2023), which found no consistent relation between group singing and social structure (community size and social differentiation), in contrast to what would be expected based on earlier theorizing (Lomax, 1968; cf. Shilton, et al. 2025).
[13] Accordingly, we aim for this article to address questions specific to Hadza practices, rather than comparative questions. As a general stance, we position our work against generalizing Hadza music as an “African music” or “hunter-gatherer music,” both of which are problematic categories. This goal requires that we discuss Hadza lifestyle in its dynamic reality. Marlowe (2010) sets a positive example, acknowledging that Hadzabe have been undergoing cultural transition for a long time, with European colonization playing a central role. He consolidates the citations of documented outsider influence in his timeline of foreign documenters’ descriptions of the Hadzabe, including their trade practices and interactions with other communities (ibid, 20–7), and a timeline of cultural influences from other communities (ibid, 34–5). External influences persist in the forms of ethnography, Christian missionization, and increasing ethnotourism, activities that do not simply obfuscate a static tradition but are in ambivalent dialogue with a living tradition.
Methodology
Ethnographic Ethics
[14] Community approval and Human Research Subjects approval (from the University of Nevada, Las Vegas and the Tanzanian Commission for Science and Technology COSTECH) permitted the recording of a variety of musical scenarios: focus group interviews, informal daytime singing, and larger group celebrations. Each scenario demanded a flexible adaptation of the parameters for sharing content, ensuring in each instance that interlocutors and singers agreed to be recorded for the project and agreed to have the music shared. Interviews were run in Swahili, with clarification provided in Hadzane. Transcriptions of the interviews and songs were provided by coauthor Shani Mangola as well as long-time collaborators and translators who have worked with Mangola and Crittenden on previous transcription projects. We obtained both community-level consent upon entry at a residential encampment as well as individual-level consent before recording commenced. This project is one that the community specifically requested, with permission granted to the research team to record songs as they arise in daily life.
[15] This project’s archival material is meant to be available for use by Hadzabe, and its content is chosen and created by our community collaborators, team members, and interlocutors. The present article, concerned with the recorded music material, aims to be the first publication in a series of studies that will include community participation to varying degrees. We present close analyses of four songs, selected by community members and study participants as the focus of the first study from this collaborative effort. Future articles will detail our development of accessible archival resources and approach to musical repatriation. We hope the present article’s contiguity with successive publications is clear in retrospect, and that a preliminary close look into musical contents helps readers understand and appreciate the aesthetics of Hadza music-making.
[16] The songs and extramusical content discussed in the article are delimited by the overlap between the knowledge our interlocutors choose to share with the public and their boundaries of privacy. We only discuss songs and musical customs that are endorsed through collective and iterative consent, a process that involves reconsenting our interlocutors on specific modes of presentation and analysis. In opposition to some extant publications’ exposition of taboo subjects in Hadza music, we seek to build ongoing trust with the community by practicing the method of ethnographic refusal, broadly defined as “a method centrally concerned with a community’s right to self-representation” (Zahara 2016b; see also Simpson 2007; Zahara 2016a). Extending Hadzabe’s denunciation of certain musical publications disclosing taboo materials, and as a direct request from our community interlocutors for this article, we do not cite such studies. Whereas such material already exists, we can only amplify the community’s authority for self-representation in the present and future publications.
[17] The ethical and methodological issues of previous Hadza musical corpora, dating at least to 1998, include reporting on culturally private practices, mischaracterization of musical instruments’ role in their music, and staged scenarios of musical performance. In contrast, we sought to build a corpus that better represents everyday Hadza music-making, aimed at archiving and digitally securing the current state of cultural knowledge for future generations of Hadzabe. This is not meant to freeze Hadza culture or uphold a notion of “authenticity,” but to put a stake in the emerging—and precarious—representational discourse of Hadza music.
Building a Corpus
[18] Our present song corpus was recorded primarily during summer 2019 by coauthor Alyssa Crittenden at two residential camps, one located near the village system of Mang’ola and the other in a more remote location in the Gideru Hills. The corpus includes recorded video from Crittenden’s previous fieldwork with Hadzabe in 2004–5 and 2012 as well as personal video footage from coauthor Shani Mangola, all used with community permission. We are prioritizing the most recent recordings as to represent a synchronic view of singing. Likewise, we highlight moments in Crittenden’s focus group interviews with Hadzabe from her 2019 fieldwork. Figure 1 depicts one such focus group with several of Crittenden’s long-term interlocutors, who organically decided to start singing, akin to spontaneous singing in everyday social life, and agreed to record a particular song.
Figure 1. Focus group of women singing Hale Hale ’Ahabi’iteya in 2019.
[19] Presently, the construction of our corpus is the most collaborative aspect of the current project. Songs that were included in our corpus came out of typical dialogue among Hadzabe interlocutors in our focus groups, much like the way they instigate singing in everyday life. Though existing recordings of Hadza songs in performance contexts are valuable in other ways, our corpus is unique in its representation of casual musical preferences and decision-making. The corpus, furthermore, represents singers who are children and elders, women and men, alone and in groups. Using open-ended prompts with our focus groups, our interlocutors ultimately made the musical decisions. The exchange between our research questions and interlocutors’ personal and cultural boundaries led to our intentional exclusion of some sacred and private songs—a method of ethnographic refusal mentioned earlier. We include songs that the community specifically asked to be recorded and analyzed. These include public vocal songs with varied topics: songs for children and about interpersonal relationships, about relationships with plants and animals, and about land rights and other political issues.
[20] We also acknowledge that the corpus presented here was made opportunistically: this study takes place within a larger, collaborative project led by coauthor Crittenden, building resources for community-directed musical preservation. While a general musical system may underlie our examples, our focus is not to theorize it. Thus, we found it pragmatic to choose several songs for close analysis and survey rhythm and meter within and across songs.
Transcriptions and Music Analysis
[21] We primarily rely on staff notation for this preliminary study. Staff notation is a restrictive tool, and by no means do these transcriptions represent all the musical events of their respective recordings. Nonetheless, staff notation is an accessible method for documenting and indexing a corpus, and it can represent many of the significant features in multipart vocal music and lyrics. The transcriptions are meant to help us better understand our corpus by recognizing its parts in finer detail and to elicit new questions that warrant further ethnography or computational analysis. It should also be noted that the transcribers are non-Hadzabe coauthors Gabriel Zuckerberg, an ethnomusicology doctoral candidate, and Jason Carl Rosenberg, a composition professor and music cognition researcher. While cultural outsiders are prone to repeating the issues found above in Grauer (2011), these issues are avoided as much as possible through practicing self-awareness and studying Hadza sociocultural context. Therein is also the opportunity for outsider transcribers to take advantage of their distinct musical subjectivities in forming novel, creative questions.
[22] Our approach to transcribing Hadza songs prioritizes the rendering of rhythm and meter. Yet, we are aware of the bias for non-African scholars to study rhythm and meter in African music practices (see Agawu 2003). This largely stems from European music theorists’ historiography of tonality and its subordination of parameters other than melody and harmony. In contrast, we intend to be candid and deliberate regarding our emphasis on rhythm and meter, an emphasis we only decided after an initial round of transcription and our realization of these features’ structural importance. Unlike pitch-related features, temporal features have two common non-vocal markers in musical performance: dancing and clapping. In our corpus, videos of dancing and audio with clapping provide basic insight into Hadzabe’s rhythmic and metrical thinking. The videos consistently show that dance steps and clapping align with the same syllables. We call this alignment “beat-marking” and examine it closely in our analyses, since, given this apparent equivalence, we interpret these markers as signifying an internalized sense of beat in the Hadza musical idiom. Clapping data has thus been crucial in transcribing the structure of musical phrases.
[23] Our focus on rhythm and meter is also a matter of convenience, because as music scholars like Tara Browner and Anne Danielsen emphasize, “Western” staff notation is relatively strong in representing rhythm and meter over pitch (Stanyek 2014, 115), plus it can be easily complemented by extra markings to indicate microrhythmic “groove” (ibid, 130–1). This metaphorical “strength” may come from metrical constraints: when beat-marking (e.g., clapping) is customary in a given musical style, the beat can be understood as a recurrent unit that implies meter. In Hadza music-making, pitch does not have the same constraints (as it would if, say, fretted instruments were used), so pitch has no unit cognate to the beat. On the other hand, staff notation’s “strength” of rendering of rhythm and meter may be in its generalization of patterns and, thus, reduction of detail. Our analyses make use of staff notation’s efficiency and regularity while also with dealing with its shortcomings in representational precision, where we propose additional markings and alternative schemas.
[24] Additionally, our rationale for focusing on rhythm and meter extends to our archive’s repatriation aim. Song and dance, parts of the same cultural practice, overlap most greatly in their temporal structure. Tara Browner also notes, from her perspective as a dancer, that cyclic forms of music are not simply “repetitive,” and their variability may be important to analysis (ibid, 116). Thus, we transcribe recordings in their full durations to illustrate within-song variability. We hope that our analytical priorities are auspicious for future Hadzabe (re)learning a more comprehensive scope of their present-day musical practice. Scholars of musical repatriation can attest to the efficacy of using archival material in pedagogy. Patricia Shehan Campbell and J. Christopher Roberts argue that
“there is the potential, through the pedagogical application of archived recordings to supplement informal at-home experiences, for teachers to work across a continuum of repatriation acts to bring children into a full-fledged experience of their musical identities within the lessons they design and deliver in schools (Campbell and Roberts 2018).
Extending access to our archive to digital platforms accessible by Hadzabe will set the groundwork for a new avenue in their community-guided pedagogy of traditional song and dance, in any manner they choose to pursue it.
[25] Continuing with our decision to center beat-marking in our methodology, we conducted iterative ethnography to gather more clapping data. About a quarter of the original recordings feature clapping, and about a half feature some form of beat-marking. Thus, we sought to expand our interest in beat-marking across the remaining half of our initial corpus. Coauthor Shani Mangola conducted the iterative ethnography, taking several trips back to participating communities. He played previously recorded songs with no audible beat-marking to Hadzabe interlocutors, who were then recorded singing back the song with the addition of clapping. These new recordings were vital to identifying rhythmic stress and metrical organization throughout the songs in our corpus. Because these songs were performed by request rather than spontaneously, these newer recordings assist as a guide to intuited groove of the original recordings, rather than replacing them for our analysis. Whether or not Hadzabe regularly clap or dance to these songs in everyday life, our project aims to present songs in the way Hadzabe choose to present them.
[26] The beat-marking in our original corpus and iterative ethnography recordings disambiguates, at the very least, divisions of rhythmic pulse. These latter recordings revealed a difference in metrical categories: songs recorded originally with beat-marking are fully isochronous (regular, equally timed beats), and songs that prompted iterative ethnography have some element of non-isochrony. To remain consistent across these two categories, we aimed for a logogenic (lyrics-based) approach: we transcribed each pulse as the sum of the syllables sung within it. We then interpreted each cycle of pulses in non-isochronous songs as a measure and approximated each measure as a meter. Small variations in metrical duration are reflected in a measure’s time signature, for the sake of precision when another interpretation is not otherwise clear. Finding the dominant beats of isochronous songs required inference from multiple sources: when melodic phrases start, multiple kinds of beat-marking such as qqingilibi’i rattles, and/or concurrent accentuation of pulses in videos with dance.
[27] We approximate pitch, often using accidentals and slurs to represent patterned inflections and gestures. In future rounds of transcription, we aim to create figures that enhance learning alongside oral and archival transmission. These may include alternative paradigms that highlight certain musical elements better than staff notation, like Global Notation for its emphasis on pitch intonation (see Killick 2020).
Hadza Public Vocal Music
Content and Context
[28] Our present corpus of Hadza singing comes from over three hours of audio recorded primarily during summer 2019, made on a Zoom H1n digital recorder and, in a few cases, an iPhone. From this audio, we transcribed 28 song performances, which amount to at least 20 individual songs. The correspondence between our corpus and other extant corpora is shown in Table 1. Recording scenarios were prioritized to maximize representation across gender, age, and number of singers. Those most frequently recorded were groups of adult women and groups of adult men. In general, the rest of the transcribed corpus is equally distributed among groups of girls, groups of boys, solo singers of each age range and gender, and groups of mixed demographics.
Table 1. Correspondence between songs in our corpus (leftmost column) and other corpora.
[29] So far, we have only had a preliminary discussion about musical style and repertoire. Below are selected moments from our interview with a focus group of Hadzabe men, responding in general about singing in everyday life. All interlocutors’ names have been anonymized here and where appropriate for the article’s remainder (I stands for “Interviewer”).
R: The elders are saying that there is no such a thing as a ‘category’ because when they are building a house, they just build it. They build it as they normally would and they don’t necessarily sing. If they are going to the bush they just go—they don’t have a special song.
I: P, do you sing for your children?
P: Maybe if they are crying… Well, I sing to my child even if he doesn’t cry.
R: Elders are saying, “our songs have no categories, they only have one flow.”
I: How does it have one “flow”? What does that mean?
R: It means that if we are singing a Hadzabe song, it’s just Hadzabe(ne) (our culture), it’s the same.
Each of these men’s perspectives indicate the arbitrariness of a “song category,” cautioning against the imposition of an external framework on Hadza musical life. For the time being, these intuitions led us to refer to our corpus as a repertory of public vocal music rather than one or more delineated, unified genres. This interview also signals two potential avenues for future interchange. Firstly, is there a relationship between songs and the contexts in which they are sung (see discussion of solo versus group singing in Shilton, et al. 2023)? If not for an activity-specific purpose, do songs serve other functions in everyday life? Secondly, what is meant by Hadza songs sharing a “flow” (translated from Swahili mtiririko)? Is this notion shared widely, and if so, does it imply a particular style or aesthetic? We may return to these questions by investigating Hadzane’s concept of hazanokwa, meaning “in the Hadza manner” (Miller and Anyawire 2007, 89).
[30] Information about song provenance varies among the songs in our corpus. The following names are left non-anonymous, as they are public figures and part of shared, public knowledge. Most contemporary Hadza songs are attributed to Mzee (“Elder” in Swahili) Muyugu. According to one interlocutor, Mzee Muyugu’s songs “have been sung for generations and generations and singing them has become his legacy and a memory for the community.” Two other Hadza composers lived in the Siponga region: Malangasu, who composed Lambu Lambu and Mbulumbulu ’Ase’a, and Mzee Gichawi, who composed ’Ika Kwata. S. Murungunya, a community member who has contributed more recently, composed Magufuli Kweqepe’e, Ulimwengu ’Ishiako’o, and Zzi’eya Obe’e Yamu’aya. There is limited knowledge about the composition of other songs, such as Bahephe’e ’A Nxobabe’e, which is thought to have come from the Mang’ola area, and ’Uhuweya Titi’i from the Sungu area. The lyrics of public vocal songs reference a variety of topics concerning daily life, including hunting for animals, foraging for baobab fruit, the honeyguide bird and honey foraging processes, interpersonal stories, tourism, and politics. While these topics do not signify song categories, they exemplify a range of Hadzabe’s communal concerns.
[31] We now briefly discuss how the content and usage of songs raise questions of cultural change, focusing on two examples. Together, cultural memory and historical documentation attest to cultural change since British and German colonization of Tanzania, as well as the perseverance of particular songs. Mzee P responded to the following:
I: Is there no one who remembers the old songs, which are not sung at present—even you elders, K, G, and P?
P: No. New songs have dominated.
Soon after, Mzee P started singing Lambu Lambu (“Dry Wood”), a song that “depicts a beautiful woman sitting in an ostrich nest,” having used animal fat from a successful hunt for beautification, now a lost Hadza practice (Hatfield 2019).[2]
The interview continued,
T: Is this the old song?
G: Yeah man.
T: Do you know it guys?
K: We don’t know it. We have never heard it…
G: Why don’t we sing together? Let the elder lead and we will follow.
The entire group of men joined Mzee P’s lead and, with few signs of hesitance, sung a full and passionate rendition of Lambu Lambu. The men’s courage to perform a song with little guidance seems to indicate a robust musical intuition for structure and style. If the men present did know the song, K’s statement suggests an attitude of deference to his elder’s musical knowledge. In either case, this performance opens the possibility to inquire about norms of Hadza musical intuition, spontaneity, and values. Such intuition may also corroborate the elders’ suggestion of a Hadza “flow.”
[32] In contrast, the song Nqatekohina (“Stingless Honeybees”) is an example of strong cultural perseverance. Dorothea Bleek documented the following lyrics in a 1931 article:
Though Bleek’s IPA-based orthography is atypical for transcribing Hadzane, these lyrics are still recognizably sung almost a century later, transcending generations through oral tradition. Unfortunately, Bleek did not document the melody for Nqatekohina. Her description of Hadza singing and dance is rather racist, as is typical of all early ethnographies of Hadzabe. Still, Bleek’s article is evidence that Hadzabe have kept this song alive in oral tradition for decades, together with their long-standing tradition of honey foraging. A continued effort to find historical accounts of Hadza songs could help discern whether the survival and disappearance of songs are contingent on certain factors of cultural significance. In doing so, we could address whether it is a matter of lyrics depicting the current state of everyday life and whether certain songs are more persistent or transient.
[33] A final point on content and context concerns the concept of a song. Research on Hadza repertory should be cautious with the empirical usage of this etic terminology. Numerous cultural systems include the practices of music and singing in broader concepts, or with conceptual divisions different than in English, for example, the Ancient Greek mousikē referred collectively to music, dance, and poetry. Consistent with J. H. Nketia’s account of African expressive practices occupying a gray area between speech and song (Nketia 1974, 177–88), Hadzane does not have a root for song, but rather haggatoko, a feminized form of the word haggato (“dance”). Hadzane also uses the verbs dla’a (“to sing”) and thaka (“to play music” or “to dance”), among a broad musical vocabulary. Though this vocabulary may not constitute a coherent “ethnotheory” that can be used in musical analysis (see Agawu 2003, 181–3; Agawu 2017), the shared conceptual basis of music and dance inform the present, imperfect approximation between the meanings of song and haggatoko.
[34] Hukwashi’e (“Stand Up”), a popular, celebratory dance song, reveals possible ambiguities in translating the concept of haggatoko. We transcribed four variants of the Hukwashi’e melody: sung by a group of women and men in a touristic context, a woman alone, a 4-year-old girl alone, and a group of children. The adults’ rendition features not only the widest array of verses but also an introduction on the zeze. The first line states, hukwashi’e hukwashi’e, hukwashi’e shakwako wa’inasa (“Stand up…” 3x “Let’s all fetch water.”), and, with each line repeating twice, the singers ensue with verses about animals and collective dancing. The children’s renditions have modified lyrics as well as entirely different verses, including a variant of the aforementioned line: hukwashi’e hukwashi’e, hukwashi’e popho wa’inama (“Stand up…” 3x “Let’s all dance like an impala.”). Contrasting more so, the adult woman’s rendition has the line ’a’iya munahe ’a’iya (“The child is crying… She’s crying.”), expressing the flexibility of singing Hukwashi’e as a lullaby, a context far from the song’s typical celebratory character.
[35] Furthermore, renditions of Hukwashi’e (along with other songs) vary in melody, rhythm, accompanying harmonies, and even structure. Figure 2 displays the two constituent phrases of the full Hukwashi’e melody, of which both phrases are repeated in the group renditions and only phrase B is repeated in the solo renditions. Other than certain fundamental lyrics, all renditions are identifiable by their basic melodic structure and harmony placement (e.g., the major third harmony on the dotted quarter note in m. 3). Through further discussion with Hadzabe about these recordings, we hope to discern which musical and extramusical factors play a role in identifying a haggatoko. Such factors may concern lyrical content, whether one or more subphrases are sung, and emic parameters that we did not recognize. For the purpose of the present article, we use “song” as a heuristic for categories of recognizable lyrics and melodies, understanding that this concept is inexact.
Figure 2. First melody repetition of adults singing Hukwashi’e. A and B represent the two constituent phrases. Danced in three beat cycles with emphasis on the transcribed downbeat, hence it is transcribed in 9/8.
Ethnotourism and Missionaries
[36] Any discussion about Hadza life and music today would be remiss to not acknowledge the impact of ethnotourism and Christian missionaries. Regarding ethnotourism, two songs in our corpus exemplify dissimilar attitudes toward non-Hadzabe; they also illustrate different paths of external cultural expressions into Hadza public vocal repertory. Zzi’eya Obe’e Yamu’aya (“Our Land is Beautiful”) celebrates Hadzaland and its wildlife, calling for their protection and flourishing. This song features an interlude of antiphonal chanting between a man leading the song and a group of children: “Come, come,” followed by the reply, “and visit Hadzabe.” The interlude varies among renditions found in other corpora in lyrics, duration, and linguisticity (degree to which music resembles everyday speech), yet in all cases the leader swiftly returns to a reprise of the initial melody. The interlude’s lyrics suggest a touristic audience, despite outsiders having virtually no understanding of Hadzane. It would also be misleading to infer a touristic function of Zzi’eya Obe’e Yamu’aya, because it was recorded for a corpus of songs to preserve Hadza cultural heritage with no presence of tourists.
[37] Zzi’eya Obe’e Yamu’aya may be contrasted with a touristic song that is relatively enigmatic, called Nalidanganya Shilangela. It is one of the most frequent songs found in other available corpora (see Table 1) and is occasionally titled Tegeregaki. Nalidanganya Shilangela is also overwhelmingly frequent among touristic videos of Hadzabe on YouTube. According to one of our Hadzane translators, the lyrics of Nalidanganya Shilangela are nonsense words and thus cannot be translated into English, while another interlocutor suggests that it came from the Isanzu community. Liz Smith and Kirk Miller, in research documents from 2008, were also unable to translate these lyrics but suggested that they might originate from a Bantu language. Further research led us to Gunderson’s (2010) book on labor songs of the Sukuma, a large Bantu populace near Hadzabe in Tanzania. One song, titled Degelekagi (“Listen”), caught our attention for its resemblance to “tegeregaki.” Gunderson placed Degelekagi in a chapter on songs that celebrate the Tanganyika African National Union, the main anti-colonial political party that led toward Tanzania’s sovereignty from Britain (ibid, 342–5). In her master’s thesis, Hatfield writes that the incipit Tegeregaki also translates to “Listen,” but that the lyrics comprise warnings to Hadzabe about outsiders coming to tell them lies (Hadzabe 2018 and Hatfield 2019). According to our interlocutor who recognized the song as originally Isanzu, nalidanganya in Isanzu means “don’t say we lied,” and the song is often called “The Funeral News.” This at least corroborates the lyrics’ cautionary sentiment. Regardless of whether the melodies carrying these lyrics for Hadzabe, the Isanzu, and the Sukuma are related, we presume these songs are historical cognates.
[38] These quasi-touristic songs exemplify our gap in understanding of musical pragmatics: what it means for Hadzabe to sing for a touristic audience versus for themselves. To consider the pragmatics of Zzi’eya Obe’e Yamu’aya, we would ask what the music and lyrics offer Hadzabe in contemplating their recent accommodation to collectively working in the tourist industry. As for Nalidanganya Shilangela, we would ask how knowledge about the lyrics is distributed among the community, and if there is an intended irony of singing this song for tourists given its lyrical content about mistrust. If we cannot account for this song’s specific origin, we wonder whether it bears significance to Hadzabe as an outlier in their repertory.
[39] As for Christian missionaries, their influence is explicit in Dlozobe’ema Musa (“The Words of Moses”), which expresses living in an ongoing Biblical narrative. We are unsure when this song entered the cultural repertory, as missionaries have been attempting to change Hadza lifestyle for decades, starting in 1927 with a British colonial endeavor to pastoralize Hadza society (Woodburn 1979). There have been repeated attempts to settle the community, relocate them to village centers, and force them to take up agriculture (see Blurton Jones 2016 for historical discussion). None of these attempts has been successful—all are seen by the community as human rights violations. The community has largely resisted conversion, maintaining traditional cosmology at the center of their spirituality (see Apicella 2018; Lang, et al. 2019). When directly queried, community members often lament the loss of land and religious freedom, while simultaneously voicing gratitude for continued presence of missionaries when researchers and tourists are in short supply and they need material support (Crittenden and Mangola personal communication). Today, Hadzabe continue to negotiate between this external force and the incentives for the resources that they offer (Marlowe 2010, 32–3). Evidently, these resources are not only food and churches, but also an emergent corner of their musical repertory. In addition to Dlozobe’ema Musa, which was composed by a missionary, songs like Ulimwengu ’Ishi’ako’o (“Universe, Help Me”), composed by a Hadza native, hint at the internalization of Christian themes, or at least Euro-American spiritual themes, in everyday singing.
Analysis of Lyrics, Rhythm, and Meter
[40] A total of twenty song recordings were transcribed collaboratively by coauthors Gabriel Zuckerberg and Jason Carl Rosenberg. Below, we provide analyses of four songs that each led the transcribers to differing observations regarding the notation of the rhythm and meter. The project team and Hadzabe interlocutors agreed on four songs as a compromise between research interests and the community’s cautiousness with publicity, due to their prior exploitation and their desire to keep some music private and off of YouTube. The analyses are ordered in the difficulty experienced in transcribing and, relatedly, the degree of debate between the two transcribers about analytical representation. While the first two songs are perceived by both transcribers as isochronous and metrically consistent, the latter two are perceived as metrically complex and thus prompted different metrical interpretations. The transcribers’ debates indicate not only the relative perceived complexity of certain songs, but also the limits of staff notation and need for better analytical tools, and the analytical utility of more practice-based participant-observation. Nonetheless, the process of transcribing our larger corpus revealed some shared rhythmic tendencies existing among the use of distinct metrical patterns. Thus, we believe these four songs exemplify the stylistic constants and diversity among a greater repertory. Notable patterns include the presence of both simple and compound meter songs, alternation between simple and compound meter within two songs, and the metrical stressing of short syllables in short-long syllable pairs. These four songs also bring the relevance of extramusical themes to the authors’ attention through lyrics about childhood, national politics, and traditional folkways.
[41] Our first example is a relatively short song, sung by a four-year-old girl. One interlocutor explains that ’Iyaye ’Iyaye “is about a child singing about remembering her mother and sometimes being punished.” This song has simple meter, marked by quietly tapped beats in our original recording. In the transcription in Figure 3, the first staff line represents the basis of the whole melody, and its constituent phrases, starting ’iyaye and sayito, are themselves quite similar. The melody varies slightly in measure 9, in which accidentals are used to approximate expressive bends in pitch. Overall, the transcribers’ experience of ease and confidence transcribing this song allows us to present it as akin to a baseline, contextualizing the debates in transcribing the following songs on an increasing continuum of effort. This perceived simplicity of ’Iyaye ’Iyaye motivates us to inquire how Hadzabe think about musical difficulty and aptitude. Is the transcribers’ perception of simplicity consistent with Hadzabe’s views? Are specific songs considered difficult to learn or sing? What is the relationship between musical complexity and Hadzabe children’s acquisition of repertory and style?
Figure 3. ’Iyaye ’Iyaye (full duration 1:06), mm. 1–12 (0:00–0:37), sung by a four-year-old girl, as transcribed by the authors. See full transcription in Appendix.
[42] This rendition of Magufuli Kweqepe’e features a man leading a group of boys. Overtly political, community member S. Murungunya composed this song “urging Magufuli (and the Tanzanian government) to help the Hadzabe community,” according to an interlocutor. Controversial and autocratic former president of Tanzania John Magufuli died in March 2021 and was succeeded by the more democratic-leaning president Samia Suluhu Hassan. The lyrics of Magufuli Kweqepe’e may express the need for help, but they more clearly indicate political solidarity among Hadzabe and the despair they experience from decades of imperialism within Tanzania. Murungunya also signals Hadzabe’s deep identification with Hadzaland in his song Zzi’eya Obe’e Yamu’aya. In addition to the Hadzane word yamu’a, meaning “land,” Magufuli Kweqepe’e makes use of the synonymous Isanzu word tcalo. Coincidentally, yamu’a is hypothesized to be a cognate with the Iraqw word yaamu, meaning “earth” or “ground” (Miller and Anyawire 2007, 250). This apparent syncretism of local languages raises a few questions for further research: Does this song express solidarity with non-Hadza communities? Is it important that loanwords are used, and does their usage relate to this song’s recent composition and political context?
[43] Musically, Magufuli Kweqepe’e appears to be typical of Hadza public vocal songs, given its compound and fully isochronous meter. Though the precise timing of beats is more variable early on, the singers start to clap in the latter third of this recording, clarifying our interpretation of the meter. This allowed more regularity in our transcription of the entire recording, such as in the beginning in Figure 4. We chose to use a duple compound meter (6/8) out of convenience to delimit small phrases, and perhaps out of lack of embodied knowledge on the transcribers’ part, but a triple or quadruple meter might be equally valid. The tag ’ahi is especially variable and may also be represented as ending in a held note, but it fits the meter as is.
[44] The transcribers also experience a broader musical ambiguity related to affect and enculturation: notably, the dire contents of the lyrics seem to contrast with fast tempo and regular temporal structure, which might be heard as upbeat and danceable. Magufuli Kweqepe’e may serve as a basis to explore aesthetic and emotional associations with musical attributes and lyrics, or to what degree Hadzabe consider such associations important. Furthermore, it raises central questions about the option of dance in Hadza music: what makes music danceable, which social meanings are tied to a danceable song, and how might one distinguish the meaning attributed to specific performances (e.g., when and when not danced)?
Figure 4. Magufuli Kweqepe’e (full duration 1:24), mm. 1–25 (0:02–0:32), sung by a male leader and a group of boys, as transcribed by the authors. See full transcription in Appendix.
[45] An interlocutor explains, “this song was composed during the rainy season a long time ago by a famous Hadzabe composer, named Malangasu, after his calabash broke. He was thirsty and had to use a shrub plant which stores water in its vines to drink.” This shrub is known to Hadzabe as mbulumbulu, a word that resembles the Swahili root for the Iraqw, mbulu, and regional names east of Lake Eyasi, including the Mbulu Highlands and the settlement Mbulumbulu. The lyrics also speak to tlamakwahe, the thrush nightingale bird (Luscinia luscinia), which typically calls at midnight. The song asks why the bird did not call at night but instead in the morning.
[46] Mbulumbulu ’Ase’a comprises a few verses that can be repeated and pieced together at the song leader’s discretion. The primary two lines are mbulumbulu ’ase’a and kha’alomama zzifi tlamakwahe ’ase, each consisting of three or four metric cycles per repetition of the section. A children’s rendition of Mbulumbulu ’Ase’a opts for three repetitions of the mbulumbulu line, compared to the adult rendition analyzed in this article that opts for four repetitions of the mbulumbulu line. Both renditions repeat the kha’alomama line twice.
[47] Figure 5 depicts the first 16 measures of our transcription of Mbulumbulu ’Ase’a as sung by a group of men, led by a man named P. Each measure represents the antiphonal melody, interlocking between P and his fellow singers. P shifts his timing of the first two syllables (mbu- and –lu-) among the first and latter repetitions of the melody from a short-long (mbu–lu-Ø) to a long-short sequence (mbu-Ø-lu), keeping a steady duration. Later in the recording, at measure 53 in our transcription, a new song leader echoes this rhythmic flexibility, singing mbulu- again as a short-long sequence. At the end of each phrase, the antiphonal group response, ’ase’a (“sleeping”), leads directly into the leader beginning the next phrase. The call and response parts are contiguous in creating a metrical unit; they seem to be mutually reliant for singers to expect the beginning of each phrase.
Figure 5. Mbulumbulu ’Ase’a (full duration 1:56), mm. 1–16 (0:06–0:32), sung by a group of men, as transcribed by the authors. See full transcription in Appendix.
[48] Though each full antiphonal phrase is equivalent in duration, the transcribers interpreted the exact number of pulses differently. Initially, one transcriber reported hearing three three-pulse beats (nine pulses total), best represented in triple compound meter (9/16) to be consistent with our corpus’s many compound meter songs, while the other transcriber reported hearing five two-pulse beats (ten pulses total), suggesting representation in simple quintuple meter (5/8). After convening, the transcribers could deliberately shift to hearing each other’s intuition but realized that neither consistently matched the recording. Another recording of the song by a group of boys substantiated this uncertainty as, again, both proposed meters gave an imperfect yet near fit. These initial solutions both prioritized, and improperly assumed, that the beats are isochronous. To disambiguate, the transcribers requested a secondary recording with clapped beats, which rather supports a non-isochronous meter. The singers unwaveringly clap on the first and fourth pulses in each nine-pulse cycle, grouping together the first three pulses. This grouping of three-note durations is consistent with the compound meters found extensively in our corpus. Hence, this insight suggests understanding the song’s metric cycles to be additive, as they were finally transcribed: 9 = 3 (mbu–lu-Ø or mbu-Ø-lu) + 2 (mbu–lu) + 4 (‘a-Ø-se-Ø). It is possible that the second clap is syncopated, but only regular, isochronous clapping is heard elsewhere in our corpus. If the meter were completely compound and isochronous, we expect that the interlocutors’ clapping would have marked three beats per cycle, as is the case in other songs.
[49] It would be misleading, however, to assert that Mbulumbulu ’Ase’a concretely follows metric cycles of nine pulses. Indeed, a more precise view of P’s timing gives an impression of forward momentum to the transcribers. Similar to the antiphony and parallel phrase structure of a religious litany text, the exact onset of each phrase is flexible, seemingly determined by the leader’s rapid choice rather than a precisely fixed groove. The transcribers then surmised that the leader and group singers mutually push and pull when entering the antiphony, creating non-isochrony on the pulse level. This microrhythmic flexibility proves to be a challenge for notational representation, particularly with our goal of characterizing an overall meter. Therefore, we suspect that the pulse felt by our interlocutors may be better reflected by a more fluid descriptor of meter, shown in Figure 6 as 9~10/16. Despite aiming for descriptive accuracy, this compromise is not an ideally elegant depiction of the meter embodied in performing Mbulumbulu ’Ase’a. We speculate that a notation method that delimits measures but does not necessitate equivalent pulses within them would be better suited to represent and analyze this song, if not Hadza musical style in general.
Figure 6. Alternative metrical interpretation of Mbulumbulu ’Ase’a, mm. 1–6 (0:06–0:15), as transcribed by the authors.
[50] ’Ika Kwata is attributed to Mzee Gichawi of Siponga, translated by Gudo Bala (1998) with the title “Big Giraffe Song.” The central character in the lyrics is minyakalo, one of numerous words for a giraffe. The rendition of ’Ika Kwata analyzed in this article also uses kalo, a hunting name for female giraffes (Miller and Anyawire 2007, 106, 136, 259; cf. for alternate pronunciations and synonyms). This song is exemplary of topics in the Hadza oral tradition; Bala features a story in his booklet, titled “The Big Giraffe,” recounting a similar instance of hunters failing to kill a giraffe (Bala 1998, 16). One of our interlocutors clarifies that, in this tale, “the giraffe did not die because the poison was bad.” The historical significance of giraffes in Hadza diet (see Obst 1912) and mythology (see Matthiessen and Porter 1974) is well documented.
[51] Figure 7 shows an excerpt of our transcription of ’Ika Kwata as sung by a group of Hadza women. The woman whose vocal line is depicted in the top staff acts as a song leader, initiating repetitions of the song’s two melodic sections and singing lyrical variations across the repetitions. The bottom staff represents the remaining group of women—primarily one woman who solos later in a high register.
Figure 7. ’Ika Kwata (full duration 3:00), mm. 52–8 (2:35–2:55), sung by a group of women, as transcribed by the authors. See full transcription in Appendix.
[52] Analyzing ’Ika Kwata was a particular challenge for the transcribers, who made the same error as in Mbulumbulu ’Ase’a, expecting to find rhythmic and metrical uniformity. Due to limited beat-marking in this rendition, a group of interlocutors provided a secondary recording with consistent clapping to help discern rhythmic and metrical groupings. However, while the clapping is consistent at the beginning of each phrase, it alternates in tempo between the song’s two constituent melodies, labeled by their starting words: ’ika and dongolo. Much like in other songs, the beats marked by clapping in ’Ika Kwata tend to coincide with silences and short notes more than long notes: the first beat of the line ’ika is during a silence preceding the word ’ika, and the line dongolo is a short-long-short syllable sequence with the first syllable marked. Analyzing this song, the two transcribers only came to disagree in how to interpret these beat-markings after they received this insight. One transcriber attempted to maintain a uniformly simple meter, thus treating the alternation in clapping pattern as syncopation, and adjusting most measures to have non-isochronous endings. The other transcriber interpreted this alternation as metric modulation, or shifting between compound and simple (here, three- and four-pulse beats). An eighth note pulse would consistently represent the same duration, so a simple meter beat would have a longer total duration than a compound meter beat. Both of these interpretations would require the transcribers to acknowledge an additional, if not idiosyncratic, metrical possibility in Hadza music: syncopation or metric modulation. The latter interpretation prevails in our transcription, but it still requires phrase endings to be represented non-isochronously.
[53] Our interpretation of the metric modulation is generalized in Figure 8. Metric modulations take place between the two melodic sections,’ika and dongolo. A variable length, “x,” marks a cadential moment with a held pitch of variable duration. The meter of 9+5/8 breaks from the ’ika melody’s isochrony by shortening the syllable (hu)-pha by an eighth note, providing a metrical pivot into the 4/2 meter of the dongolo section. The singers occasionally repeat either melody a third time, marked by the meters to the right of the dotted line. In the recording we analyzed, the women vamp on dongolo until the leader initiates the ’ika melody, alternating for seven total repetitions of the song.
Figure 8. Metrical scheme used to transcribe ’Ika Kwata.
[54] The same phenomenon of a two-section form and regular metric modulation is also found in the song Tekemase Qunibe’ema (“They Stabbed a Giraffe in the Calf”), and even more clearly in its own secondary recording with clapping. Thus, we interpret Tekemase Qunibe’ema as sharing the feature of non-isochronous meters at the transitions between compound and simple meter sections. Further, the lyrics of Tekemase Qunibe’ema also narrate the hunting of a giraffe, so it is altogether reasonable to consider a meaningful relationship between these two songs, whether in origin, context in Hadza lifestyle, or merely in their strong musical and lyrical resemblance. These commonalities are noteworthy in that they seem to contrast with one interlocutor’s assertion that there are no “song categories,” indicating that categories may be implicit but not verbally recognized.
Across the Corpus
[55] As seen in Table 2, eight songs are analyzed as using a simple meter and eight as using a compound meter, out of the twenty transcribed songs. In the remaining four songs, two alternate between simple and compound meter, only Mbulumbulu ’Ase’a is best described, at the moment, as having an additive meter, and only Dlozobe’ema Musa is entirely uncertain, as our interlocutors declined to make a secondary recording with clapping. Dlozobe’ema Musa is the only song of these twenty with explicitly Christian lyrics, so it is neither surprising nor empirically useful that it is performed distinctively.
[56] Ten to fifteen of the songs are fully isochronous, and the non-isochronous songs clearly share metrical tendencies. Songs with non-isochronous moments, like Nqatekohina, have those moments at the end of otherwise isochronous compound meter phrases. Though it has no sequences of isochronous beats, the metrical organization of Mbulumbulu ’Ase’a is comparable to Nqatekohina, as the repeating phrases in both songs start with a three-pulse beat. These non-isochronous songs begin phrases with isochronous sequences and end them with flexible non-isochronous tags.
[57] All the songs have either a unary (one-section) or binary (two-section) form, occasionally with an interlude or outro, often a vamp on a simplified part of the melody. A few songs have unique formal characteristics. For example, Zzi’eya Obe’e Yamu’aya has two alternating melodies as well as an antiphonal interlude between repetitions of the song. And, in Mogela Taheneya’a, each repetition has a short introductory phrase sung by the leader. This could be conceptualized as another melodic section, but binary sections in other songs tend to be equivalent or similar in duration.
Table 2. List of songs in corpus, categorized by metrical and rhythmic attributes.
U Uncertain. I Interlude, outro, or vamp. V Variation in melody.
Discussion
Broader Analytical View
[58] Our analyses show that language, especially in its prosody, is important to Hadza musical style. A major focus for the transcribers has been the metrical stressing of the initial short syllable in short-long syllable pairs. The transcribers came to this realization through an initial misinterpretation of this rhythmic pattern as stressing the long syllable, quite similar to Stobart and Cross’s (2000) work with Quechua- and Aymara-speaking musicians. Stobart and Cross suggest a relationship between Quechua’s linguistic stress patterns, especially its high degree of suffixing and word-initial stress in trisyllabic words, and the apparent correlate in musical rhythm (Stobart and Cross 2000, 83–5; see also Patel 2008, 168–73). Likewise, Hadzane has a high degree of suffixing, often causing stress to be antepenultimate, or on the third-to-last syllable of a word (Miller and Anyawire 2007, Appendix; Sands 2013, 277). Figure 9 exemplifies our proposed impact of Hadzane’s linguistic stress pattern on musical stress, illustrating the near ubiquity short-syllable notes and silences accentuated by beat-marking, particularly at the beginning of phrases. This music-language relationship is not to be taken as deterministic, but Hadzabe’s shared affordances from the Hadzane language likely impact vernacular music-making and song composition.
Figure 9. Metrically stressed short-syllable notes and silences marked in red boxes in excerpts from A Magufuli Kweqepe’e, B Mbulumbulu ’Ase’a, and C ’Ika Kwata.
[59] ’Iyaye ’Iyaye is omitted in Figure 9, because it does not include short-syllable metrical stressing. In contrast to songs in compound meter that often present this stressing pattern, the presence of some songs in simple meter like ’Iyaye ’Iyaye divert us from suggesting that this pattern is a formal or necessary Hadza idiom. We noted that this song was easier for the transcribers to analyze, even with the challenge of working cross-culturally, which may be because it is meant for a four-year-old child to sing solo. If simple meter is relatively intuitive for both the transcribers and the young singer—and thus the transcription is an epistemically meaningful representation—this meter represents an element of musical thinking that is accessible across a notable cultural distance. The transcribers’ experience would then indicate that meaningful cross-cultural analysis cannot be a matter of cracking a code to what makes a musical system “different” from the analysts’ (often oversimplified as “Western”) musical perception. Rather, such an analytical project should then always consider “sameness” underlying the musical thinking of the analysts and the analyzed performers (see Agawu 2003, 168–71).
[60] In doing this, we can better contextualize this meter’s converse: compound meters with relatively more evident beat-initial short-syllable stressing. This pattern, appearing emblematic of musical style in much of the present corpus, may be better described as an aesthetically desirable motif rather than a subconscious tendency, given how the youngest Hadzabe children feel and think musically when singing. As an otherwise embodied tendency—as adults clap and dance on these short syllables—beat-marking is not an immediate matter of cultural belonging but accrued through each child’s enculturation.
[61] The degree to which language influences Hadza musical rhythm and meter should be underscored, given the community’s minimal use of musical instruments in everyday life. Language may be disproportionately important when instruments are absent, as musicians can generate competing linguistic tendencies that structure a musical style’s overall rhythmic and metrical patterns. Unfortunately, empirical research on this topic cannot yet support or refute our conjecture about this music-language relationship in Hadza music. One line of research on the music-language relation uses the nPVI (“normalized Pairwise Variability Index”) to measure variability of adjacent note durations, search for intracultural commonalities between musical and linguistic rhythm, and examine these patterns cross-culturally (e.g., Patel and Daniele 2003; VanHandel and Song 2010). However, this approach does not seem well-suited to address the music-language relations discussed here; regardless of low or high rhythmic variability durations in Hadza music, the core pattern found in the present analysis concerns the frequent occurrence of metrically stressed short syllables and silences in songs. Thus, an approach examining relations between phonological stress and short-syllable duration in Hadzane speech could be more productive in studying linguistic influences on musical rhythm in this case (Temperley and Temperley 2011).
[62] Notably, the majority of existing models for music theory and analysis of African Indigenous musics either focus on purely instrumental music or presume a community to have an instrumental tradition parallel to its singing tradition, neither of which are true for Hadzabe. This precedent is both an issue for the utility of our work to represent the particularities of Hadza music and, in doing so, for its representative validity of Hadza aesthetics. David Temperley (2000) argues that underlying metrical tendencies in “African music” (really, West African traditional drumming) and Western classical music are strikingly similar. Assuming those metrical tendencies apply more widely, as to include Hadzabe, Temperley still holds the importance of ethnomusicology and the study of emic knowledge, because the analysis of meter “tells us nothing, in itself, about why these things are important and how they contribute to the value of music” (Temperley 2000, 91-2). On the other hand, the most prominent analysts of African rhythm and meter—including A.M. Jones, J. H. Kwabena Nketia, John Miller Chernoff, Kofi Agawu, and David Locke—focus their efforts on music with stylistically integrated beat-marking, and thus their methods rely on the representative validity of clapping, drumming, or other percussion. The present study indeed relies on clapping, which appears to be more a restriction than an answer to the question of Hadza rhythm and meter, because much of the repertory has no beat-marking. This limitation in our initial analysis was not easily solved by our secondary recordings because of the degree to which they alter the music-making process. Indeed, several interlocutors went so far as to say that the addition of beat-marking “ruined” certain songs, and in some cases it would not even be permitted to try to clap. That beat-marking is an aesthetic or even ethical issue will be taken seriously as our research progresses.
[63] Given the basic limitations of applying models of Africanist meter scholarship to the present study, we look next to the study of “free rhythm” and microtiming. Martin Clayton (1996) wrote a crucial review on theories of free rhythm, or music without fixed meter, ultimately leading to a call for further study. In the process, Clayton delimits the concept of periodicity as a temporal structure with “pulse-based periodic organization” but not meter, and free rhythm as lacking periodicity but possibly having pulse (Clayton 1996, 329). He adds that it may be useful to clarify several subtypes of periodicity. It would seem that while some Hadza songs could be classified as having periodicity or even meter, many songs could be said to have some subtype of periodicity. The moments of non-isochrony in songs like ’Ika Kwata and Mbulumbulu Ase’a—especially phrase endings without beat-marking, even in our secondary recordings, and with variable phrase lengths across repetitions—indicate a loose or flexible periodicity. It may also be significant that singers clap, even in the most isochronous songs, only later in their performances, and there appears to be more rhythmic flexibility in the majority of those performances before clapping begins. Furthermore, there might be heuristic value in understanding the process of music-making through such metaphors as “flow.” Not only did one of our interlocutors relay his elders’ discussion of a Hadza musical “flow,” but this term has been used to describe “free rhythm” music as in the case of Jewish liturgical music (see Frigyesi 1993). Though “flow” is not itself explanatory, it replaces descriptors defined by absence such as “nonmetric” and “free” with something constitutive and visualizable (ibid, 67).
[64] Microtiming research may also offer a useful model in later stages of the present research. Two of the existing models suggest either an experimental psychology approach or a greater investment in musical ethnography. Much effort has been going into data-driven empirical studies on microtiming: in particular, experimental psychology, corpus analyses, and systematic musicology (see the recent review by Danielsen, et al. 2024). However, any such methodologies would require a disruption of communal life for data collection beyond the research team’s current agreement with the Hadza community, and are therefore untenable. On the other hand, though not nominally “microtiming” research, the work of Charles Keil on “participatory discrepancies” and with Steven Feld on “groove” indicate a fruitful path in the semantics and discourse relating to musical style and music-making among Hadzabe. Participatory discrepancies refer to the kinds of tension that are aesthetically idiomatic to a style of music-making, and groove, though more thematic than technical, refers to the feeling of participating in a musical process (see Keil and Feld 1994). Taking this approach would help clarify how “flow” is used in Hadza discourse, its associations, and/or how its use relates to social distinctions in Hadza life.
[65] Additionally, such idiosyncratic methodologies as that of Vijay Iyer (2002) may be adaptable. Iyer typifies “microrhythmic techniques” that are guided by constraints of the human body and cognition while diversified by cultural particularities (Iyer 2002, 387). As Iyer does with African-American music, it may be possible to classify idioms of microrhythmic variation in Hadza music through a combination of directed ethnography and further analysis. Following Iyer’s naming conventions for techniques—like “spreading,” in which note onsets are thought to be “spread” out from a note’s “ideal […] metric position” to a more easily embodied yet mathematically uneven note onset rate (ibid, 400–3)—it may be valid to say that non-isochronous moments in ’Ika Kwata and the “additive” meter of Mbulumbulu ’Ase’a are related to a technique we would term as “rushing.” If this hypothesis is accurate, these songs would be performed in an idealized, less expressive fashion and/or imagined as entirely isochronous. They would only become non-isochronous, or analyzable as such, through the conventionalized “rushed” entrance of leaders in antiphonal songs or in “rushed” repetitions of litany-style lyrics. It would not be surprising that such a convention would be explained as “flow,” as “flow” may be iconic of both the experience and aesthetic outcome of embodied microtiming.
[66] Further research questions should be more deeply informed by our ethnography, built on specific questions of Hadzane language and Hadza musical values. One possible direction is to focus on Hadzane’s rich onomatopoeic vocabulary. Daudi Peterson writes that many names for animate objects are “descriptive of behavior, form or function” (Peterson, et al. 2012, 90). Through compiling and translating lyrics, we have observed the breadth of Hadzane onomatopoeia and its grammatical flexibility. For example, Miller and Anyawire believe that the verbs ’uhu (“to cry”) and he’e (“to fall”) are mimetic of the actions they describe and may be inflected for “person-number-gender and tense-aspect-mood” (Miller and Anyawire 2007, 28, 89, Appendix p. 8). Given the linguistic role of onomatopoeia, we would like to investigate possible connections between sound symbolism in language and music. One crucial context for sound symbolism is hunting, where Hadzabe strategically use mimesis.
[67] Lastly, future research could explore the potential relationships between Hadza musical forms and sociocultural values. In this article, we represent songs as being flexible to the number of melody repetitions and adaptable to different contexts of musical participation (e.g., who is participating, whether they are dancing or beat-marking, etc.). Though flexible repetitions and adaptability may be common cross-culturally for short-form music, we did not yet fully account for meaning attributed to these variations. Following Rudge (2019), we intend for further research to address whether this apparent “flexibility” relates to the social organization of Hadza music. For example, in what ways and to whom might such flexibility be intentional? Does it reflect egalitarian values? Is it customary for participants to compose or improvise new lyrics, and if so, which participants? What musical choices and limits exist in adapting songs across cohorts and demographics?
Analytical Representation in Context
[68] The materials we are making public will now exist alongside the published albums of Hadza music-making and YouTube videos of ethnotourism performances. This preexisting backdrop is one of performances staged for the purpose of recording and listening or viewing, representing music-making quite differently from how it occurs in everyday life. Compared to everyday, communal music-making, these preexisting professional recording and ethnotourism settings create different social pressures, which prompt musical differences. Thomas Turino (2008) articulates the distinction between performance settings with different social-aesthetic values as “participatory” versus “presentational performance.” For example, presentational settings may inhibit performers’ tendency toward heterophony, in order to increase “clarity” of the musical material (Turino 2008, 57–9). Such stylistic adaptations are plausible in cases of presentational Hadza music-making: for example, the vocal timbre required of performances loud enough to accompany dance, or changes in ornamentation in larger singing groups.
[69] Our choice to focus our analysis on rhythm and meter reveals a critical issue in the overwhelmingly presentational public representation of Hadza music-making. All published albums have some form of beat-marking in many, if not all, tracks. This is especially the case in YouTube recordings shared by tourists, featuring dance (sometimes with qqingilibi’i), zeze playing, and/or clapping. Hadzabe ethnotourism performers utilize these elements to satisfy tourists’ aesthetic tastes and to actively engage tourists in dancing or singing along. Without accounting for the social pressures at hand, one might interpret these elements as simply “enhancing” the vocal line. Rather, they create different rhythmic demands than exist in everyday singing. These elements inevitably restrict the repertory that Hadzabe perform for tourists (songs that already feature beat-marking in participatory practice), and they may even compel Hadzabe to modify repertory to meet the aesthetic and energetic demands of tourism. The most apparent modification is the alignment of sung syllables to marked beats. The precise timing of syllables is noticeably different in the absence and presence of beat-marking: examples include the initial 54 seconds versus the final 32 seconds (with clapping) of our recording of Magufuli Kweqepe’e, our original recording versus iterative ethnography (with clapping) recording of Mbulumbulu ’Ase’a, and our recording versus the several published recordings (with qqingilibi’i) of ’Ika Kwata.
[70] However these modifications may take hold, our interlocutors already express a suspicion toward them, in so far as beat-marking can “ruin” certain songs. The microtiming of syllables may seem to be one small issue, but this example is symptomatic of greater nonconsensual representation and imposed cultural change. This is the case for the public media portrayal of Hadza culture and lifestyle overall. For the increasing population of Hadzabe moving to villages and cities and gaining access to digital media, the widening availability of touristic performances may bring about new questions of representation, creativity, ethics, and cultural change. Our present corpus is by no means unmediated, but it is more thoroughly consensual, less staged, and aims to be more self-representative of the community that produced it.
Conclusion and Broader Goals
[71] Despite decades of scholarly research on the Hadza people of Tanzania by anthropologists, there has been no analytical study of their music. Our two primary aims in this article are to introduce our ongoing initiative of musical preservation and repatriation of Hadza public vocal music, and to share our preliminary analytical insights. Informed in part by Western classical music theory concepts and methods, we describe Hadza public vocal songs as varying between simple and compound meters (sometimes alternating within songs) and between isochronous and non-isochronous pulses. The majority of songs in our corpus have a compound isochronous meter. Also, nearly every song stresses the short syllable in short-long syllable pairs.
[72] Our present study of Hadza music is preliminary to a methodology of grounded music theory and represents the first step into the realm of their musical practice. Rather than being a conceptual template, grounded theory emerges from a continuous dialogue between ethnography and analysis (see Bryant and Charmaz 2007). Though our initial analytical questions of musical structure serve the interests of the non-Hadza coauthors, they intend it as a dialogic action in the direction of generating a grounded theory of Hadza music. The questions posed in our analysis section are directed toward this continuing engagement with the Hadza community and in the creation of materials to repatriate. Discussions will help us curate materials that align with the self-image Hadzabe intend to foster, such as understanding the political valences of particular songs, and questions will help us curate materials in appropriate categories, such as those around music learning and danceability. We hope that the pragmatic use of these analytical questions sets a precedent for other scholarly pursuits of Indigenous community-inclusive music research and repatriation.
[73] Future writing will expand on our project goals, draw a broader anthropological view, and situate the role of music within a richer view of Hadza everyday life. It will also deepen our present methodology—its purposeful instability and emergent purposes—by discussing Sonya Atalay’s adapted concept of “braided knowledge.” In brief, Atalay’s braided knowledge is achieved by coalescing “community knowledge” and “archaeological data” (Atalay 2012, p. 27), which in our case would be combining community-inclusive ethnography and music analysis. This methodology helps to bridge epistemological divides between different interpreters (community versus ethnographers) and forms of knowledge (qualitative and empirical). It is our obligation to not conclude with clean generalities or formulaic knowledge, as this article both synthesizes disparate literature and directs the reader to many unanswered questions. We hope that readers learn to be at ease with analytical ambiguity, and that this article demonstrates the potential for music analysis to be an interpretive and collaborative form of scholarship.
ACKNOWLEDGMENTS
We would like to thank the Hadza community for our ongoing collaboration, Nani and Chris Schmeling for research support, Audax Mabulla and Ibrahim Mabulla for research facilitation, and the University of Nevada, Las Vegas Department of Anthropology for funding.
REFERENCES
Aang Serian Peace Village. 2003. Wahadzabe Voice of Ngorongoro: Songs from Tanzania’s Indigenous Hunter-Gatherers. Produced by Mohamed Yunus Rafiq.
Agawu, Kofi. 2003. Representing African Music: Postcolonial Notes, Queries, Positions. New York: Routledge.
————— 2017. “Against Ethnotheory.” In The Dawn of Music Semiology: Essays in Honor of Jean-Jacques Nattiez, edited by Jean-Jacques Nattiez, Jonathan Dunsby and Jonathan Goldman. Rochester, NY: University of Rochester Press.
Apicella, Coren Lee. 2018. “High Levels of Rule-Bending in a Minimally Religious and Largely Egalitarian Forager Population.” Religion, Brain & Behavior 8(2): 133–48.
Atalay, Sonya. 2012. Community-Based Archaeology: Research with, by, and for Indigenous and Local Communities. Berkeley: University of California Press.
Bala, Gudo G. 1998. Hadza Stories & Songs. Printed by the UCLA Graduate School of Education and Information Science and UCLA Document Services. Los Angeles: Friends of the Hadzabe.
Barrett, H. Clark. 2022. “Psychology Within and Without the State.” Annual Review of Psychology, 461–87.
Benenson, Bill, dir. 2014. The Hadza: Last Of The First. Produced by Benenson Productions.
Bleek, Dorothea F. 1931a. “The Hadzapi or Watindega of Tanganyika Territory.” Africa IV 273–86.
Bleek, Dorothea F. 1931b. “Traces of Former Bushman Occupation in Tanganyika territory.” South African Journal of Science 28(7): 423–29.
Blurton Jones, Nicholas. 2016. Demography and Evolutionary Ecology of Hadza Hunter-Gatherers. Cambridge: Cambridge University Press.
Bryant, Antony, and Kathy Charmaz. 2007. The SAGE Handbook of Grounded Theory. London: SAGE.
Campbell, Patricia Shehan, and J. Christopher Roberts. 2018. “Teachers as Agents of the Repatriation of Music and Cultural Heritage.” In The Oxford Handbook of Musical Repatriation, by Frank Gunderson, Robert C. Lancefield and Bret Woods, 283–302. New York: Oxford University Press.
Clayton, Martin R. L. 1996. Free Rhythm: Ethnomusicology and the Study of Music without Metre. Bulletin of the School of Oriental and African Studies, 59(2): 323–32.
Crittenden, Alyssa N., John Sorrentino, Sheniz A. Moonie, Mika Peterson, Audax Mabulla, and Peter S. Ungar. 2017. “Oral health in transition: The Hadza foragers of Tanzania.” PLOS One 12 (3).
Danielsen, Anne, Ragnhild Brøvig, Kjetil Klette Bøhler, et al. 2024. “There’s More to Timing than Time: Investigating Musical Microrhythm Across Disciplines and Cultures.” Music Perception 41(3): 176–98.
Fabian, Johannes. 1990. “Presence and Representation: The Other and Anthropological Writing.” Critical Inquiry 16(4): 753–72.
Frigyesi, Judit. 1993. “Preliminary Thoughts toward the Study of Music without Clear Beat: The Example of “Flowing Rhythm” in Jewish Nusaḥ.” Asian Music 24 (2): 59–88.
Gibbons, Ann. 2018. “Hadza on the Brink.” Science, May 18: 700–04.
Grauer, Victor. 2011. Sounding the Depths: Tradition and the Voices of History. CreateSpace Independent Publishing Platform.
Gunderson, Frank D. 2010. Sukuma Labor Songs from Western Tanzania: We Never Sleep, We Dream of Farming. Leiden: Brill.
Harota. 2013. The Hadza King of Zeze. Swahili Records.
Hadzabe. 2018. Tegeregaki: Listen with Hadza. Produced by Tara Hatfield, CD Baby.
Hatfield, Tara. 2019. Music, Musicians, and Social Advocacy: Environmental Conservation, Knowledge-Sharing, and Cultivating a Culture of Wisdom in Northern Tanzania. Master’s thesis, Urbana: University of Illinois at Urbana-Champaign.
Iyer, Vijay. 2002. “Embodied Mind, Situated Cognition, and Expressive Microtiming in African-American Music.” Music Perception 19(3): 387–414.
Joachim, Luhalala, and Zahor Zahor. 2022. “Hadzabe’s Transformation and Livelihood Dynamics: Livelihood Sustainability in the Changing Environment of Yaeda Valley.” Tanzania Journal for Population studies and Development 29(1): 44–64.
Jodoin, Sébastien. 2017. Forest Preservation in a Changing Climate: REDD+ and Indigenous and Community Rights in Indonesia and Tanzania. Cambridge: Cambridge University Press.
Jones, Arthur Morris. 1959. Studies in African Music. London: Oxford University Press.
Keil, Charles, and Steven Feld, eds. 1994. Music Grooves: Essays and Dialogues. Chicago: University of Chicago Press.
Killick, Andrew. 2020. “Global Notation as a Tool for Cross-Cultural and Comparative Music Analysis.” Analytical Approaches to World Music 8(2): 235–79.
Lang, Martin, Benjamin G. Purzycki, Coren L. Apicella, et al. 2019. “Moralizing Gods, Impartiality and Religious Parochialism Across 15 Societies.” Proceedings of the Royal Society B 286 (1898).
Lomax, Alan. 1968. Folk Song Style and Culture. Washington, D.C.: American Association for the Advancement of Science.
Mangola, Shani Msafiri, Justin R. Lund, Stephanie L. Schnorr, and Alyssa N. Crittenden. 2022. “Ethical Microbiome Research with Indigenous Communities.” Nature Microbiology 7: 749–56.
Marlowe, Frank W. 2010. The Hadza: Hunter-Gatherers of Tanzania. Berkeley and Los Angeles: University of California Press.
Matthiessen, Peter, and Eliot Porter. 1974. The Tree Where Man Was Born: The African Experience. New York: Avon.
Miller, Kirk, and Mariamu Anyawire. 2007. A Hadza Lexicon. Unpublished.
Moshi, Henry Sweddy, Asantiel Makundi, and Baltazar Mushy Reguly. 2019. “Land Degradation and its Impacts on Cultural Tourism Among the Indigenous Peoples in Tanzania: A Case of the Hadzabe.” International Conference on the Future of Tourism (ICFT). Arusha, Tanzania: The Open University of Tanzania. 1–16.
Nketia, J. H. Kwabena. 1974. The Music of Africa. New York: W. W. Norton.
Obst, Erich. 1912. “Von Mkalama ins land der Wakindiga.” Mitteilungen der Geographischen Gesellschaft in Hamburg 26: 2–27.
Patel, Aniruddh. 2008. “Rhythm.” In Music, Language, and the Brain, 96–179. New York: Oxford University Press.
Patel, Aniruddh D., and Joseph R. Daniele. 2003. “An Empirical Comparison of Rhythm in Language and Music.” Cognition 87(1): B35–B45.
Pennisi, Elizabeth. 2004. “The First Language?” Science, February 27: 1319–320.
Peterson, Daudi, Richard Baalow, and Jon Cox. 2012. Hadzabe: By the Light of a Million Fires (Book and CD). Dar es Salaam: Mkuki na Nyota Publishers Ltd.
Pollom, Trevor R., Kristen N. Herlosky, Ibrahim A. Mabulla, and Alyssa N. Crittenden. 2020a. “Changes in Juvenile Foraging Behavior among the Hadza of Tanzania during Early Transition to a Mixed-Subsistence Economy.” Human Nature 31: 123–40.
Pollom, Trevor R., Chad L. Cross, Kristen N. Herlosky, Elle Ford, and Alyssa N. Crittenden. 2020b. “Effects of a Mixed-Subsistence Diet on the Growth of Hadza Children.” American Journal of Human Biology 33 (1).
Rudge, Alice. 2019. “Flexibility and Egalitarianism: Musical Insights from Hunter-Gatherers.” Ethnomusicology Forum 28(2): 163–83.
Safari, John Godfrey, Aron Joseph Nkua, and Zacharia Samwel Masanyiwa. 2021. “Household Food Security among Hadza Hunter-Gatherers in Mkalama District, Tanzania.” Agriculture & Food Security 10(1): 1–12.
Sands, B., I. Maddieson, and P. Ladefoged. 1996. “The Phonetic Structures of Hadza.” Studies in African Linguistics 25(2): 171–204.
Sands, Bonny. 2013. “Morphology: Hadza” The Khoesan Languages, by Rainer Vossen. London: Routledge.
Shadrack, Seleman, and J. K. Mwalilino. 2022. “Hunter-Gatherers’ Coping Strategies on Climate Change and Prospect in Iramba and Mbulu Districts, Tanzania.” International Journal of Tropical Drylands 6(2): 77–89.
Shilton, D., Passmore, S., and Savage, P. E. 2023. “Group Singing is Globally Dominant and Associated with Social Context.” Royal Society Open Science 10(9), 230562.
Shilton, D., Patel, A. D., Hill, K., and von Rueden, C. 2025. “Why Collective Music-Making is Sometimes Rare: a Study of Four Indigenous Societies.” Evolution and Human Behavior 46(4), 106695.
Simpson, Audra. 2007. “On Ethnographic Refusal: Indigeneity, ‘Voice’ and Colonial Citizenship.” Junctures: The Journal for Thematic Dialogue 9: 67–80.
Smith, Liz. 2008. Hadza Songs. Unpublished.
Smith, Sophia Hersi, dir. 2016. The Way We Live Now. Produced by Royal Anthropological Institute.
Stanyek, Jason. 2014. “Forum on Transcription.” Twentieth-Century Music 11(1): 101–61.
Stephenson, J., 2000. The Language of the Land: Living Among the Hadzabe in Africa. New York: St. Martin’s Press.
Stobart, Henry, and Ian Cross. 2000. “The Andean Anacrusis? Rhythmic Structure and Perception in Easter Songs of Northern Potosí, Bolivia.” British Journal of Ethnomusicology 9 (2): 63–92.
Temperley, David. 2000. “Meter and Grouping in African Music: A View from Music Theory.” Ethnomusicology, 44 (1): 65–96.
Temperley, N., and Temperley, D. 2011. “Music-language correlations and the ‘Scotch Snap.’” Music Perception, 29(1): 51–63.
Turino, Thomas. 2008. “Participatory and Presentational Performance.” In Music as Social Life: The Politics of Participation, 23–65. Chicago: University of Chicago Press.
VanHandel, Leigh, and Tian Song. 2010. “The Role of Meter in Compositional Style in 19th Century French and German Art Song.” Journal of New Music Research 39(1): 1–11.
Woodburn, James. 1979. “Minimal Politics: The Political Organization of the Hadza of North Tanzania.” In Politics in Leadership: A Comparative Perspective, by William A. Shack and Percy S. Cohen, 244–66. Oxford: Clarendon Press.
Yatsuka, Haruna. 2015. “Reconsidering the ‘Indigenous Peoples’ in the African Context from the Perspective of Current Livelihood and its Historical Changes: The Case of the Sandawe and the Hadza in Tanzania.” African Study Monographs 36(1): 27–48.
Zahara, Alex. 2016a. Refusal as Research Method in Discard Studies. March 21. https://discardstudies.com/2016/03/21/refusal-as-research-method-in-discard-studies/.
————— 2016b. Ethnographic Refusal: A How to Guide. August 8. https://discardstudies.com/2016/08/08/ethnographic-refusal-a-how-to-guide/.
[1]. Here, we use the endonym “Hadzabe” to refer to the entire community or plural individuals, and we use “Hadza” otherwise as an adjective.
[2]. Hadzabe (2018) and Hatfield (2019) title it Tsi-kwa Tsi-kwa (“The Oil is Dripping”), or Ccikwa Ccikwa in our orthography.