The Bullroarer Atlas

Atlas Essay — 2026-08-22

The Living Double

Rouku · Idoma country · the Urubamba · the Xingu–Tapajós · Xerente country · Torres Strait

At Rouku, Nakre Abia folds coconut leaf into a bow, ties on a fibre cord, and sends it around his body. It buzzes only while it clears his shirt, the ground, and the trees. The film calls the thing by name: frfr, a children's toy. But Komnzo has another word for another bullroarer. That one is wooden, secret, and born in a myth of blood. It belongs to initiated men. The two instruments have not replaced one another. They are alive together.

In brief

  • Living bullroarer systems often keep a child-safe double beside the restricted instrument: a second form separated by material, name, shape, scale, time of day, or permission.
  • Komnzo children make the leaf frfr; initiated men formerly guarded the wooden ŋofoŋofo. Idoma boys improvise an ofukùfukù from a ruler while an ancestor society sounds a separately named fish-shaped plaque.
  • Matsigenka evidence states the division as sacred and nocturnal versus children's and diurnal. Xipaya evidence suggests a shape boundary; paired Xerente passages show a stronger ritual object distinguished by custody, training, and the being expected to answer.
  • Mabuiag supplies a graded-open case, while Mbaya-Caduveo and Upper Tanana supply controls in which one object crosses between ritual and play. There is no universal evolutionary sequence.
Nakre Abia swinging a coconut-leaf Komnzo frfr in open air at Rouku.
Nakre Abia swings the Komnzo frfr at Rouku, 14 September 2012. The visible instrument is a children's coconut-leaf toy. The restricted wooden roarer belongs to another name and another social jurisdiction. Frame from Christian Döhler and Nakre Abia, Frfr a Kifikifi, Kaipuleohone/Zenodo 5653474.

Living bullroarer systems often keep a child-safe double beside the restricted instrument: a second form separated by material, name, shape, scale, time of day, or permission.

The toy is not always the last stop

The standard history runs downhill. A sacred instrument loses its god, taboo, and ceremony; what remains is a toy. European buzzers and buttons-on-a-string make that descent easy to imagine. The atlas contains genuine examples of ritual forms passing into play. But the downward line becomes a trap when applied to every toy. It turns coexistence into chronology. If an ethnographer records a children's bullroarer, we begin to invent a sacred past for it; if a cult instrument appears nearby, we assume the toy is its worn-out descendant.

The more interesting records resist that sequence because the senior and junior forms are documented at the same time. They may occupy the same village or the same museum accession. Their names, materials, hours, or operators differ. The child's instrument does not announce that the cult has died. It shows how a living cult distributes access to a dangerous acoustic technology.

“Child-safe” here describes a social fact, not a modern safety rating. Coconut leaf can cut. A swinging board can strike someone. The safety lies in what the object is allowed to mean. Its sound does not summon the forbidden being, confer the office, expose the myth, or place the child inside the penalty structure of the senior instrument. The physical mechanism is shared; the ritual consequence is sandboxed.

Rouku: translate the voice into leaf

Christian Döhler's Komnzo lexicon makes an unusually clean separation. Frfr is a “children's toy bullroarer,” made entirely from coconut leaf, bent into a bow and tied to a cord. Ŋofoŋofo is listed under its own headword as the bullroarer, and the wood of the süfr tree is specified as its material. A filmed demonstration supplies what dictionary entries cannot: construction, handling, and sound in a continuous public act. Abia builds two leaf instruments and explains that contact with body, clothing, or tree arrests the buzz. The instrument works by staying free.

Hands tying coconut leaf strips into a Komnzo frfr.
Making the frfr: leaf, fibre, and a few minutes of work. Its perishability matters. A child can reproduce the mechanism without acquiring the durable object, restricted material, or secret name. Döhler and Abia, Rouku, 2012.

The wooden roarer occupies a different world. Döhler's grammar preserves the story of a man who heard roaring inside his wife's belly. Birds entered her body to retrieve the object, emerged blood-red, and one finally stole it for the man. The story ends by establishing the wooden bullroarer as sacred and restricted to initiated men. A later linguistic-ethnographic account describes it as a thin wooden slat on a long twisted cord and says that across the Morehead district it was formerly secret.

Nothing here resembles one undifferentiated object sliding between seriousness and play. The system makes four cuts at once: leaf against tree wood, frfr against ŋofoŋofo, child against initiate, public demonstration against bodily origin myth. The junior form reproduces the aerodynamic trick but not the sacred biography. It is almost a translation: durable wood becomes disposable leaf, esoteric name becomes everyday name, dangerous interior becomes open air.

Idoma: an ancestral fish beside a school ruler

Roger Blench's inventory of Idoma instruments is only a few lines long at the decisive point, but those lines are taxonomically exact. He distinguishes three free aerophones. The òhúmhúm, also called òfúmfúm, is a fish-shaped wooden plaque sounded by the àlèkwu society and forbidden to women. The ofukùfukù is made by young boys from a ruler or scrap wood for amusement. A spinning disc of gourd or bottle-top is listed separately. This is not an outsider confusing two whirled toys. The source separates true plaque bullroarers from the rotating disc and then separates the two bullroarers from each other.

Printed page from Roger Blench listing Idoma free aerophones, including the fish-shaped cult bullroarer and boys' ruler bullroarer.
The Idoma division in print: a fish-shaped plaque for the àlèkwu ancestor society; a ruler or scrap-wood bullroarer made by boys; and a spinning disc classified apart. Roger Blench, “Idoma Musical Instruments,” African Music 6.4 (1987), p. 49.

The materials are historically suggestive. A school ruler is a straightedge, a standardized commodity, and an almost ready-made rotor. Boys can turn the institution of schooling into noise with a hole and a cord. Yet the modern material does not abolish the ancestor society's fish. The new object is absorbed under its own name. The result is not a fossil but a two-track technology: one shaped and guarded by the cult, the other improvised from whatever lies to hand.

This case also warns against making “miniature” the essence of the double. The Idoma child form need not imitate the fish-shaped senior blade. What passes across the boundary is the acoustic operation, not the iconography. The double can be mechanically homologous while visually refusing the sacred form. Indeed, refusal may be the protection. Boys learn how a flat thing on a cord turns air into a pulse without copying the ancestor society's animal body.

Matsigenka: the clock is the fence

Among the Matsigenka of the Urubamba, France-Marie Renard-Casevitz states the contrast in a parenthesis too sharp to ignore: “the sacred and nocturnal bullroarer” exists alongside others that are children's toys, diurnal. The senior instrument bears sangenari, signs connected with the Master of Thunder and with the spotted coat of the jaguar. Its rumble becomes Thunder's voice and calls great cats to the allied shaman. The child form belongs to daylight.

A broad engraved Matsigenka bullroarer collected by France-Marie Casevitz.
A broad engraved Matsigenka instrument from the Casevitz accession, the visual context for the sacred nocturnal form described in the accompanying research. Its marks belong to a sign-world tied to Thunder and the jaguar. Musée du quai Branly – Jacques Chirac, 71.1980.61.48.
A narrow red-stained Matsigenka child's bullroarer with cord.
Object 71.1980.61.49, from the same accession, is catalogued pour enfant—for a child. It is narrower, red-stained, and retains its cord. Musée du quai Branly – Jacques Chirac.
A second narrow red-stained Matsigenka child's bullroarer.
A second object in the accession, 71.1980.61.50, is likewise catalogued for a child. The paired museum records turn the ethnographic parenthesis into a small material series. Musée du quai Branly – Jacques Chirac.

This is the clearest temporal firewall in the corpus. Sound is licensed twice, but consequence attaches to the hour. By day, a child's rotor is a toy. At night, an engraved instrument opens a relation with dangerous nonhuman allies. The difference is not hidden in acoustics alone. It is assembled from time, mark, operator, and expected listener.

The sacred bullroarer is nocturnal; there are others that are children's toys, diurnal.

Renard-Casevitz 1980, p. 263 n. 2, translated from the French

Calling this “software” is useful if it is kept concrete. The wooden blade supplies a physical platform. The social program specifies when it may run, who may run it, which signs are loaded onto it, and what event counts as an answer. A child's blade does not become Thunder's voice because it happens to buzz. The society has not installed that interpretation on it.

Xipaya: geometry controls access

Curt Nimuendajú's 1919–20 account of the Xipaya, whom he wrote as Šipáia, gives a fourth kind of partition. After midnight a bullroarer called warī warī sámį joins the flutes in the all-night Zetáβia dance. People fear even speaking about it because the subject can bring illness or misfortune, and women and children may not approach. Nimuendajú thought the sacred form was probably round or oval. He immediately contrasted it with paddle-shaped examples used as children's toys.

A paddle-shaped crosshatched Xipaya children's bullroarer collected by Curt Nimuendajú.
A documented children's toy from Boca do Baú: paddle-shaped, crosshatched, and still carrying its plied cord. The collector, provenance, and subtype fit Nimuendajú's Xipaya discussion even though the museum record preserves a conflicting legacy ethnicity field. SMVK 1922.08.0052, collected by Curt Nimuendajú.

His identification of the sacred outline was cautious because the information and drawing available to him were unclear. That caution belongs exactly where the uncertainty occurs; it does not erase the explicit paddle-toy statement or the after-midnight restriction. The safest formulation is also the most interesting one: in this Xipaya record, the source itself uses shape to distinguish a probably round-oval restricted form from a clearly paddle-shaped child form.

The case should not be expanded to the Kuruaya merely because a later catalogue row inherited the compound label “Shipaya-Kuruaya.” The primary account is Xipaya. The surviving museum card adds its own noise by preserving a legacy Tembe ethnicity field, but the collector, provenance, and paddle-toy subtype line up with the Xipaya passage Nimuendajú actually wrote. That small correction strengthens the comparison. This essay is not plotting a vague South American motif; it is following a precise distinction made in one record and then matched by a closely aligned object from the same collecting context.

Xerente: the same buzz, a different interlocutor

Nimuendajú's The Šerente places small boys' “buzzing boards” among ordinary toys. Elsewhere in the same monograph, specialists carry a named bullroarer in an elongated lidded calabash through roughly six days of instruction. Its compound name means the “howling of Hieparo-wawe.” At the appointed place they whirl it, and the being appears. The contrast is less taxonomically airtight than Idoma because Nimuendajú does not explicitly identify the boys' buzzing boards as bullroarers. But the paired passages show the relevant transformation: an ordinary whirled buzz becomes a demon-call through a special object, name, container, curriculum, operator, and place.

A long narrow Xerente bullroarer with cord wound around its neck.
The named Xerente caller collected by Nimuendajú, its cord wound around the neck. The public toy passage supplies a buzzing board; this curated object carries the compound name, trained use, and supernatural addressee. SMVK 1937.27.0146.

The Xerente case reveals what a double need not share. It may not share a name. It may not be stored in the same place or taught by the same people. It may not be expected to speak to the same listener. From outside, both acts could be described as making a board buzz. From inside, one is child's play and the other is an address with a proper recipient.

Fort McDowell: one accession preserves both tracks

The Fort McDowell Yavapai material is unusually valuable because collector, place, date, and accession hold the functional split together. Object 1911.86.95 is a 26.5-centimetre medicine bullroarer cut from lightning-struck cottonwood. Zigzag lines repeat the lightning in graphic form, and the cord is cotton-seed fibre. Two smaller ual ual in the same 1911 accession are identified as schoolboys' toys. The contrast is not reconstructed from widely separated reports; it was present in the collector's own series.

A large Yavapai medicine bullroarer incised with lightning zigzags.
The Fort McDowell medicine ual ual ii'g agual: lightning-struck cottonwood marked again with lightning. It belonged to doctors or dreamers and was used in curing and to call the Kakaka mountain beings. Pitt Rivers Museum 1911.86.95.
A smaller plain Yavapai schoolboy toy bullroarer from the same accession.
A smaller schoolboy's ual ual from the same Fort McDowell accession. The museum series makes scale visible, but the decisive divide is function, material history, and ownership. Pitt Rivers Museum 1911.86.97.

This is not an initiation-secrecy case. Gifford reports that women and children might see the medicine instrument, so the senior form should not be made artificially forbidden to fit the pattern. Its boundary is occupational and cosmological: doctor or dreamer, lightning material, curing, and mountain-spirit summons on one side; small schoolboy toy on the other. A child-safe double can coexist with a senior instrument even when the senior is visible. Permission has more dimensions than concealment.

More ways to divide one mechanism

The same architecture appears in smaller atlas records. A Chemehuevi account separates plain wooden children's toys from a notched mountain-sheep-horn instrument used in rainmaking. Material does almost all the work: common wood for play, an animal substance and a marked edge for weather power. Oraon sources likewise distinguish the men's dormitory's sacred sounding objects from a generic toy tradition, although the secondary toy label is less securely pinned than the Komnzo or Idoma names. Vocabulary still blocks the temptation to treat every rotor as one social species.

Roy's plate of Oraon musical instruments, including whirled sound objects.
Roy's Oraon instrument plate gives a material frame for a source tradition that separates dormitory-linked sacred sounding objects from a more generic toy category. Sarat Chandra Roy, Oraon Religion and Customs (1928), illustration 16.

These cases suggest at least six independent controls: material, name, geometry, scale, time, and authorization. No society needs all six. Any one may be enough if everyone knows the distinction. The controls can also be redundant, as in the Matsigenka case where hour, engraving, use, and expected supernatural response reinforce one another. Redundancy is precisely what one would expect from a system guarding a high-consequence interpretation.

Mabuiag: the boundary can have a middle

A. C. Haddon's Mabuiag evidence prevents the model from becoming a binary diagram. The large painted bigu belonged to turtle-fishery ceremony and made a low, deep note. The smaller wainis was high and shrill, kept in the bush, and attached to the same ceremonial field. Women could see the instruments, and islanders reportedly described the practice as “half-play.” Large and small, low and high, ceremonial yet partly open: the system had a middle register.

Haddon's five painted model bigu bullroarers from Mabuiag.
Five model bigu from Mabuiag. Haddon paired the large, deep-sounding form with a smaller shrill wainis; the wider practice was visible to women and described as “half-play.” Cambridge Anthropological Expedition to Torres Straits, vol. IV (1912), fig. 237.

Mabuiag is therefore not a strict child-double case. It is more valuable as evidence that permission can be graduated. The opposition is not always sacred versus profane, secret versus toy, man versus child. A society can build an acoustic ladder with different pitches and publics on adjacent rungs.

The controls break the easy theory

A useful pattern must survive its counterexamples. Among the Mbaya-Caduveo, decorated bullroarers used at funerals were reportedly given to children as playthings. The same object appears to cross the boundary. Among the Upper Tanana, the common form was a child's toy, but Nabesna John mounted one on a long stave for séances and cabin purification. McKennan treated that ritual use as John's own specialization. Here context elevates an available toy rather than producing a separate junior copy.

These controls rule out two satisfying but unsupported claims. The child-safe double is not universal, and it is not proof of a single ancient institution diffusing intact around the world. Similar social problems can generate different solutions. One society duplicates the object. Another reclassifies the same object by occasion. Another creates a graded series. Another lets one specialist appropriate a toy. The comparative discovery is the menu, not one compulsory origin.

It also follows that the atlas flag “toy survival” cannot carry a chronological meaning by itself. It may describe the terminus of desacralization, a parallel child technology, a licensed rehearsal, the same ritual object released after use, or a toy promoted into ritual by an innovator. The row-level particulars decide which history is actually present.

A cult can publish a harmless emulator

What, then, is the junior instrument for? The records rarely answer in abstract language. They show children making, sounding, and recognizing it. It may train dexterity, familiarize the ear with pulsed sound, provide amusement, or mark a permissive outer edge of adult knowledge. Those functions can coexist. What matters is that the double allows the mechanism to circulate while preventing the senior interpretation from circulating with it.

This is a striking form of cultural conservation. Absolute secrecy makes a practice brittle; total publicity dissolves its restricted force. A harmless analogue solves both problems. The acoustic idea remains ubiquitous. Children learn that a blade and cord can make an impossible voice. Yet the cult retains the names, biographies, materials, hours, signs, and offices that turn that noise into a being. The public copy keeps the hardware familiar and the ritual program scarce.

The discovery also changes how diffusion should be investigated. A toy beside a cult object may preserve contact between populations, but it may equally arise by internal duplication. The strongest diffusion cases will be those where the boundary itself travels: the same paired names, the same material substitution, the same day-night split, or the same distinctive geometry. A generic resemblance between two buzzing toys is weak. A matched system of permissions is much harder to invent as coincidence.

Children get the air; elders keep the address

Living bullroarer systems often keep a child-safe double beside the restricted instrument: a second form separated by material, name, shape, scale, time of day, or permission. Rouku translates sacred tree wood into coconut leaf. Idoma country sets a ruler beside an ancestral fish. Matsigenka country partitions the sound by day and night. Xipaya evidence points cautiously toward geometry; the paired Xerente passages toward custody and supernatural addressee; Mabuiag toward a graded scale of pitch and openness. The counterexamples show other choices, which is why the pattern is useful rather than doctrinal. The toy is not necessarily the cult's corpse. Sometimes it is the living cult's public emulator: the same aerodynamic revelation, stripped of the dangerous address.

Key sources

  • Christian Döhler, Komnzo Lexicon (2022), pp. 34, 65, 78; A Grammar of Komnzo (2018), pp. 348–349; and Döhler with Nakre Abia, Frfr a Kifikifi (Rouku, 2012), Kaipuleohone/Zenodo 5653474.
  • Roger Blench, “Idoma Musical Instruments,” African Music 6.4 (1987), p. 49.
  • France-Marie Renard-Casevitz, “Inscriptions: Un aspect du symbolisme matsiguenga,” Journal de la Société des Américanistes 67 (1980), pp. 261–295, especially p. 263 n. 2; Musée du quai Branly objects 71.1980.61.48–50.
  • Curt Nimuendajú, “Bruchstücke aus Religion und Überlieferung der Šipáia-Indianer,” Anthropos 14/15 (1919–20), pp. 1027–1028; SMVK 1922.08.0052.
  • Curt Nimuendajú, The Šerente (1942), pp. 41, 85, 89–90; SMVK 1937.27.0146.
  • Pitt Rivers Museum accession 1911.86.95–97; E. W. Gifford, “The Southeastern Yavapai” (1932), p. 237.
  • A. C. Haddon, The Study of Man (1898), pp. 306–310; Reports of the Cambridge Anthropological Expedition to Torres Straits, vols. IV–V.
  • Alfred Métraux, “Ethnography of the Chaco,” Handbook of South American Indians, vol. I (1946), p. 343; Robert McKennan, The Upper Tanana Indians (1959), pp. 44–45, 102, 152–154.

More atlas essays