In most of the world, through most of history, music has been conceived as a set of practices and techniques, expressive or religious behaviours, and adaptable repertoire resources, or as an integral part of some larger phenomenon such as theatre or healing. Over the last two hundred years, however, music has increasingly been reconceived as consisting of autonomous musical works. This has surely been one of the most important developments in musical history worldwide; yet it has been relatively little studied.

Girls’ carnival morris dancing holds a curious status in the canon of English folk performance. On the one hand, this highly competitive team-formation dance operates at a fundamental remove from the conventional spaces and narratives of the two English folk revivals with which most morris dancing is associated (Wright 2017). More closely linked to the popular “town carnival movement” of the late-nineteenth and mid-twentieth centuries than to the contemporary folk scene, girls’ morris dancing is practised in sports halls and community centres, predominantly in the northwest of England and parts of North Wales, and is rarely seen in public. The dancers—almost exclusively girls and young women from working-class communities—compete as members of morris dancing troupes within regional and cross-regional carnival organizations. They do not straightforwardly identify as “folk” dancers.

In “Culture on the Ground,” Tim Ingold states that “a more grounded approach to human movement, sensitive to embodied skills of footwork, opens up new terrain in the study of environmental perception, the history of technology, landscape formation and human anatomical evolution” (2004:315). In “Being Alive” he continues this argument, explaining that

On 17 March, the last day of Montserrat's annual St. Patrick's Festival, locals and tourists gather at the culminating Slave Feast event to commemorate the island's attempted slave rebellion on that same day in 1768. At the Slave Feast, festivalgoers enjoy local foods, buy handmade crafts, and watch artists perform music, dance, and poetry. Masquerade dancers form a circle among the crowds; the young performers wear grinning masks that dangle in front of their faces as they stamp and spin in the dusty street. Their bodies move polyrhythmically, shoulders pulsating up and down. The troupe's captain cracks a whip that bites at the dancers’ feet, urging them to step faster, higher; the performers become more frenzied. As their feet punctuate the pavement below, there is a sense that they are floating just above the ground, occasionally dipping their bodies toward the earth and momentarily disrupting the driving rhythm of the dance. Their hands gesture in toward and then away from their torsos, fingers pointing to their temples and then opening out into serving motions in front of them. The sound of the masquerade “boom” drum vibrates through the space, the high-pitched fife dictates the dancers’ steps and formations, and the goatskin kettle drum emits driving rhythms that become increasingly heated (see figure 1). Echoing the multiple voices that contribute to the community's cultural heritage, the fife layers European-sounding melodies onto African-sounding drum patterns. The discrete physical and historical elements of Montserrat's masquerade dance combine to “remember” a “dismembered” community, one that has been divided and dispersed over centuries of trauma inflicted by both man and nature. These performers dance the archive; these dancers are the archive.

As of 2011, an estimated 669,900 Aboriginal and Torres Strait Islander people accounted for 3 per cent of Australia's total population (Australian Bureau of Statistics 2015). Of that population, over 30,000 people from a uniquely large urban/rural area in the southwest of Western Australia—including the author of this article—identify as Noongar (also spelled Nyungar). This makes Noongar one of the largest Aboriginal cultural groups in Australia (SWALSC 2009; see figure 1); and yet, the Noongar language is critically endangered, with just 369 speakers acknowledged in the 2011 Australian census (Australian Bureau of Statistics 2015). Noongar language is not unique in this regard; the most recent National Indigenous Languages Survey indicates that only “around 120” of more than two hundred Aboriginal languages are still spoken and that “about 13 can be considered strong” (Marmion, Obata, and Troy 2014:xii). Music traditions, often strongly tied to language, are disappearing too: approximately 98 per cent of Aboriginal musical traditions have been lost since colonization (Corn 2012:240). As is the case with most of Aboriginal Australia, traditional Noongar music is primarily vocal, featuring lyrics in the Noongar language. This implies an inextricable link between Noongar language and Noongar song traditions, a co-dependency that is critical for the vitality of both.

In present times, it may be surprising to see the word “origins” in the title of an ethnomusicological article. Reviewing historical and current hypotheses on the origins of song, or music more generally, it becomes apparent that most authors interpret a possible bifurcation between (spoken) language and (sung) music as the most plausible point from which “music” evolved. Curt Sachs states polemically, “Music began with singing” (1943:21). Such statements beg questions about the definition of music, not just origins. I will, however, forgo such questions in this article, concentrating instead on the prehistoric and contemporary fluidity between speech and song. I understand these terms as poles on a continuum, ranging from everyday fugitive speech to formalized speech (such as invocations, recitations, chanting) and, finally, to song. In this article, I suggest that seeking to understand the moment in which this bifurcation, or rather spreading out, of utterances occurred is a productive task. Understanding the differences between speech and song is a starting point for unpacking how these distinct phenomena emerged. Alternatively, thinking about those prehistoric times when speech and song emerged may support insights about modern meaning and the function of utterances that occupy a liminal space between speech and song.

The Yezidis are a Kurmanji-speaking (northern Kurdish) religious minority living scattered across northern Iraq, Syria, the Caucasus (Armenia and Georgia), and western Europe. Among the Yezidis of Armenia, “speech” (axavtin), “song” (stran; pl. stranen), and “words about” (kilamě ser, pl. kilamen ser) constitute the three main categories of vocal production. This article is a detailed exploration of the acoustic characteristics of these categories of vocal production with a special focus on “words about.” At first sight, one may question whether “words about” (kilamě ser) actually amounts to a separate vocal category. These words lack any name that would set them clearly apart from axavtin (speech), which is clearly also made up of “words” (kilamen). Furthermore, the Yezidis never refer to an abstract category of “words about” but only to “words about” something: “words about the dead” (kilamen ser mirya), “words about exile” (kilamen ser xerîbîě), “words about the hero” (kilamen ser měranîě), and so forth. What these topics share in common is their association with tragic events and/or feelings of sadness and nostalgia; “words about” are thus lamenting utterances. When Yezidi people (especially women) say “words about” such things, they often resort to a special tone of voice. This tone may be described as a kind of chanting that combines characteristics of axavtin and stran. As with axavtin, there is neither metre, rhyme, nor steady beat. And as with stran, there are sustained pitches and an interactional salience. These differences are all addressed in more detail below.

The task set out for us in this curated section of the Yearbook is, from the perspective I present here, problematic. We are invited to consider utterances on the boundaries between speech and song, and I cannot help thinking that this is like being asked to consider bodies at the border between the air and Canada. Though the terms “speech” and “song” both have numerous meanings, speech generally refers to something relatively concrete: the use of the human voice to convey linguistic meaning. The term speech is like the term air; it refers to something intangible but still concrete. Song, on the other hand, is like Canada. It is a reification. How do we address the space between something concrete and something imagined? Song's borders lie at a variety of distinct perceived locations. Unlike with speech, we cannot objectively determine the line between song and non-song. Even if no one shares your sense of where the borders of song lie, no one has the authority to claim you are wrong. Others may be correct to deem your judgment as culturally inappropriate in a given context, but not objectively untrue. If I hear all speech as song, you cannot prove me wrong. If you see all running as dance, I have no solid ground to assert that it's not. We can quibble over intention and the importance of shared cultural conceptions, but ultimately there is no objectively verifiable way to confirm an utterance as song.