The Sacred Sound of the Steppes: A Case Study on the Significance of the Sound Produced by Golden Shamanic Headdresses

Shamanism and Sound

North Eurasian shamanism of Siberian origin includes extensive visual symbolism. One prominent symbol is the tree of life, which is located at the center of the universe and passes through and connects the three worlds or cosmic zones; upper, middle, and lower (Hasanov, 2016: 221). The tree is often represented along with an animal on each side, usually herbivores with large horns, e.g., goats, ibexes, or deer, although horses do also appear, often depicted with horns or antlers (Francfort, 2008: 36; Peterson, 2020: 56; Stein, 2017: 2). In Siberian and nomadic traditions there is a close connection between trees and deer antlers as their cycles resemble each other, growing in spring and falling in autumn (Stein, 2017: 2). Depictions of trees and antlers are often very similar in North Eurasia and their imagery seems to be almost interchangeable (Peterson, 2020: 55). The reindeer is also very important and often difficult to distinguish from deer or horses with antlers in ancient art (Farkas, 2000: 10). In addition to these animals, the tree of life is also often represented with birds above or in its branches. The birds have been interpreted as representing ‘the souls of the deceased […] waiting to be reborn’ (Stein, 2017: 2) or as ‘the renewal of life’ and rebirth (Pak, 1988: 47). In southern Siberian shamanism, the tree of life is sometimes represented as a ritual post topped with a knob in the form of a bird or a hooved animal (Hasanov, 2016: 207), and the motif of a bird sitting atop a pole has parallels in northern Afghanistan, Inner Mongolia, and Korea (Pak, 1988: 48).

The tree of life is also often connected to specific sound elements; that of bells and jingles. There exist myths and traditions among Siberian peoples where bells are hung on trees, where trees give shamans bells as a shamanic attribute, and where bells on a tree have flowers growing out of them (Hasanov, 2016: 207). Furthermore, bells are often hung from the zoomorphic knobs topping the ritual posts said to represent the tree of life. These traditions seem to have reached the Korean peninsula early in the first millennium CE, as in the states of Samhan, the precursors of the Korean Silla kingdom, there was a sacred place where drums and bells were hung on a large tree (Ham, 2013: 48). In these cultures, bells and jingles generally seem to have been used in burials so that the shaman could call on helping spirits for the deceased to have an easier journey from this world into the next (Hasanov, 2016: 202). However, there is little explanation about the meaning or representation of the jingling sound in itself, apart from one description from a shaman from the Altai who said that the seven bells on his costume represented the voices of seven virgins (Eliade, 1964: 153).

The importance of bells and jingles in shamanism is not surprising, as sound is very important in ritual and in creating certain atmospheres and feelings (Watson, 2001: 180). As such, it is usually an integral component of shamanism and extremely important when it comes to connecting humans with the spiritual (Mills, 2014: 53–54; Watson, 2001: 178). It is therefore very likely that ancient peoples considered sound, as well as acoustic effects, to be magical or somehow connected with the supernatural or the divine. Some of this magical quality may be attributed to the invisibility of sound and its behavior, which would have been inexplicable to people of ancient cultures (Watson, 2001: 180). This magical quality can extend to the person who controls the sound (Watson, 2001: 180), e.g., the shaman, as well as to the objects used to create the sounds, which in North Eurasian shamanism are commonly percussion instruments (Mills, 2014: 54; Needham, 1967: 611; Scarre, 2018: 117).

The function of sound is often said to be to contact spirits or for a transition of some sort (Needham, 1967: 611). However, it can have other functions, such as in the Altai where music is important because it pleases the spirits, which in turn leads to the wellbeing of humans (Díaz-Andreu et al., 2023: 351). It is especially significant that the spirits of the Altai favor songs which ‘imitate the sounds of the place they inhabit’ (Díaz-Andreu et al., 2023: 354), as this means that the sounds made by humans are mimetic and inspired by nature, i.e., not random or based on the individual’s fancy. Similarly, the many Mongolian cultures of Inner Mongolia share a musical tradition which is largely based on mimicking sounds in nature (Huang, 2020: 7). This special attunement to nature is not only present in ancient cultures, but is also an important component in the aesthetic approach to music in shamanic cultures still extant today, such as the Yakut, where irregularities found in nature are easily applied to music (Hodgkinson, 1996: 65).

Research on percussion instruments in North Eurasian shamanic rituals has always focused on those which are visually identifiable as musical instruments to the modern observer. This contains the assumption that the ancient culture being studied had the same definition and understanding of music and musical instruments as modern Western cultures do. However, the way humans sense the world is determined by culture (Classen, 1993: 1), and ancient humans may have had different definitions and/or perceptions of concepts such as ‘music’ or ‘musical instruments’, and/or categorized and defined the senses differently, meaning that an artefact could have multisensory functions (Classen, 1993: 36). Music, dance, and ritual might have been inseparable in some ancient cultures, with there even being a possibility that landscape and sites were used as musical instruments (Till, 2014: 294). Furthermore, if sound only becomes an important attribute after the initial visual assessment, then the research is firmly rooted in the visual despite having the aim of studying the aural. For sound and its meaning to be the main focus of research, this research should extend to any artefact capable of producing sound, as it should be the artefact’s sound-making ability which is used to designate an artefact as a ‘sound tool’, i.e., an object ‘that people in ancient times used for their sound production’ (Lund, 2012: 62).

Just like sight, sound is and always has been ‘a primary source of information about the world’ (Cross & Watson, 2006: 107). Furthermore, the individual is surrounded by, as well as a participant in, the auditory world and is therefore not as detached from the aural as from the visual (Rodaway, 1994: 91–92). The sounds which surround an individual at all times make up a soundscape, an acoustic or sonic environment which every hearing person senses (Schafer, 1994: 7–8). Of course, it can be difficult to separate sound tools made specifically for meaningful sound production and those which happen to produce sound. The study of sound therefore fits into the paradox that ‘the phenomena which it seeks to interpret are often susceptible of the most detailed scientific scrutiny yet their purposive nature remains largely unverified and their meanings inscrutable’ (d’Errico & Lawson: 2006: 41). However, the same can be said about visual images and symbols whose possible meanings are frequently studied and interpreted, and there is therefore no reason why research on the way an artefact relates to senses other than the visual and the meaning thereof should not be attempted as well.

In order to shift the focus away from the visual to the aural and multisensory, I have conducted a case study on golden shamanic tree-like headdresses from three different North Eurasian cultures – the Tillya Tepe people, the Tuoba Xianbei Confederation and the Silla kingdom – dated to the first half of the first millennium CE, to find out the possible cultural and shamanic significance of the sound they produce. I analyze the headdresses and then apply Hornbostel and Sachs’ (1961) type-classification system for musical instruments and Lund’s (1981; 2012) probability grouping to analyze their sound properties.

Selection and Methodology

The artefacts were chosen for the case study based on the following criteria: 1) They should be artefacts not commonly associated with music but capable of producing sound; 2) They should have visual representations of steppe shamanism or nomadic cosmologies; and 3) Their visual attributes should already be thoroughly studied and connected to each other. The artefacts should not be visually connected to musical instruments in order to focus on the possible aural effects not immediately apparent based on visual assessment, but at the same time the visual connections between the artefacts were considered important because it was deemed likely that an already established connection through visual elements would increase the possibility of a connection through other senses.

Tall headdresses decorated with golden vegetal and animal motifs were considered to be a very good choice for such a case study, because they are one of the longest-lasting symbols of steppe shamanism and nomadic cosmologies in existence. Such headdresses still exist today in crown-like antlered headdresses of shamans found, for example, among the Tungusic, Yeniseian, and Buryat people (Stein, 2017: 2–3). Among some ethnic groups, this headdress symbolizes the upper world or the power to communicate with the upper world. When placed on a shamanic headdress, the antlers could possibly be related to a very ancient deification or worship of the stag in Eurasia, the stag of the sky, and a symbol of the strength that the shaman receives from the sky (Hoppál, 2001: 83). While it is impossible to say how old such headdresses are, they do bear a resemblance to the Scythian-type headdresses found in the Arzhan 2 kurgan, dated to the seventh century BCE (Parzinger, 2008: 20–24; Shenkar, 2017: 167). As there exist well preserved and researched examples of ancient headdresses of this type, this facilitates further study from a different perspective.

Some of the headdresses most frequently connected through shamanic and cosmological attributes are the headdress and crown found at Tillya Tepe, two buyao attributed to the Tuoba clan of the Xianbei Confederation, and crowns from the Silla kingdom, two of which were chosen for this study (see Figure 1 for find locations).

Figure 1

I analyze each of the six headdresses, beginning with the context in which they were found and some general cultural connections to the steppes, after which I describe the artefacts in detail and discuss their visual connections to steppe shamanism and nomadic cosmologies. Unfortunately, physically examining the headdresses was not possible for this study, and instead I rely on high-quality photographs of the headdresses, both in their entirety as well as close-ups of details. Video footage of the headdresses was also used, when available, to examine how the different parts of the headdresses move. Based on their appearance and the materials of which they are made, it was possible to infer which parts of the headdresses were likely to have produced sound when worn and what they would have sounded like.

After analyzing the possible sounds made by the headdresses, the two systems used by Lund (1981; 2012) in her research on sound tools were adopted to categorize them. The first system is the comprehensive type-classification system for musical instruments created by Hornbostel and Sachs (1961), one of the most common systems in use today to classify musical instruments, and which is comprehensive enough to encompass instruments of different music traditions from all around the world. The headdresses all fall under Hornbostel and Sachs’ (1961: 15) definition of indirectly struck idiophones, as the sound is emitted by dangling objects when the headdresses are worn and so ‘the player himself does not go through the movement of striking’. The sub-categories of indirectly struck idiophones listed by Hornbostel and Sachs (1961: 15) relevant for the golden headdresses can be seen in Table 1. The headdresses are categorized into one or more of these sub-categories based on the results of their analysis.

The other system used by Lund (1981; 2012) and adopted for this case study is the ‘probability grouping’ created by Lund herself (1981: 247). This probability grouping is made up of five groups in which objects and artefacts are categorized based on how likely it is that they were used as sound tools (2012: 63). Lund’s (1981: 247) five groups are as follows:

1. ‘Artefacts which clearly are sound-producing devices (such as bronze lurs and bells).

2. Artefacts with a strong potential for having been used as sound-producing devices (such as certain pierced pig-bones […]).

3. Artefacts which are probably – to judge for example from ethnographic analogies – made to serve a double purpose, among which would have been sound production (e.g. necklaces of shells or of metal objects).

4. Artefacts which probably were not made for the purpose of sound production, but which, as a result of their design, can generate sound at the same time as they carry out their primary function, and thus can have had sound production as their secondary function […].

5. Artefacts whose function is unknown, but which can generate sound as a result of the way they are designed and which, from analogies, the find circumstances, one’s own and other people’s interpretations and/or other circumstances, can be thought to have had sound production as one of their functions (e.g. bone objects with carved-out notches)’.

It is important to note that as the significance of sound is almost entirely unknown when it comes to the headdresses studied here, their placement in Lund’s probability groups will always be speculative, at least until more evidence for sound production and its meaning and/or importance comes to light.

The Tillya Tepe Burial Site

Tillya Tepe is a burial site in northern Afghanistan, an area which belonged to ancient Bactria, dated to the first century CE. The site was located on the crossroads of several central communication routes via land and rivers and the people buried there clearly lived during ‘a high point in trans-Eurasian connectivity’ (Peterson, 2020: 50), as the artefacts unearthed there display a mixture of Hellenistic, Chinese, Indian, and steppe styles, artistry, and iconography, as well as local Bactrian influences (Hickman, 2012: 79). However, this intercultural mixture makes it difficult to assess the origins of the people buried at Tillya Tepe. Most agree that they had ancestral origins in the Eurasian steppes and nomadic traditions, although the close proximity of the Yemshi Tepe settlement site suggests that they may have been semi-sedentary rather than fully nomadic, possibly the rulers of Yemshi Tepe (Peterson, 2020: 49; Shenkar, 2017: 152). There is less consensus about their specific culture, but they were probably either Saka or Yuezhi (Francfort, 2012: 86; Shenkar, 2017: 152–154).

Tillya Tepe was excavated from 1978 to 1979. Seven individuals were buried in wooden coffins at the site, interred at the same time but buried in individual graves (Francfort, 2012: 89). Six of the graves were fully excavated and one partially, and of these six graves five were female and one was male. The site had not been looted before discovery and the amount of gold found in all the graves suggests that the individuals were elites (Francfort, 2012: 89; Peterson, 2020: 48). The individuals were buried on a mound – which itself dates to the Early Iron Age and gives the burial site the appearance of a kurgan – with the male placed in the privileged position by the top of the mound and the females buried around him. Based on this placement, as well as on the distribution of the burial goods, it can be deduced that the man was a ruler or perhaps a prince. This type of layout is similar to trans-Uralic, South Siberian, and, especially, Xiongnu burial traditions from the same time period, a connection which can likely be attributed to the Yuezhi (Francfort, 2012: 89; Shenkar, 2017: 153). Chinese accounts state that the Xiongnu forced the Yuezhi out of their homelands in Xinjiang in the second century BCE, causing the Yuezhi to migrate into Bactrian territory, with both peoples remaining in close proximity to each other (Britannica, 2016; Francfort, 2012: 89; Shenkar, 2017: 153).

The headdress and crown found at Tillya Tepe are among the clearest examples of connections between Tillya Tepe and Eurasian steppe cultures. The headdress (Figure 2) was found in Tomb IV, which belonged to the male. It is a golden miniature of a tree, nine centimeters tall, which had been placed close to the male’s skull, leading to the interpretation that it was part of the deceased’s headgear (Shenkar, 2017: 157). The tree stands on four small feet, each of which has a hole pierced through it, and its trunk is a hammered shaft to which the branches, made of wires, are attached in four tiers (Peterson, 2016: 187). The branches were attached to the shaft by beating one end flat and then wrapping that end around the trunk. The other end of each branch is looped and has chains hanging from the loop, onto which a pearl and a flat gold disk are fastened (Peterson, 2016: 187). On the top of the trunk, forming the fifth tier, sits a rosette whose six petals also have chains hung on them with a flat disk.

Figure 2

In addition to the golden tree, a small statuette of a ram (Figure 3), 5.2 centimeters long and four centimeters tall, was also found next to the skull (Shenkar, 2017: 157). It is possible that the tree and the ram were originally attached to the golden bowl the deceased’s head rested on, meaning that the tree would have hung over the male’s forehead as a symbolic headdress (Shenkar, 2017: 162–165). Another possibility is that the prince was buried with a cap made of leather or felt to which the tree and the ram were fastened. It is difficult to say exactly how these objects would have been placed on the cap, but the ram was likely facing the tree and one can wonder if the position of the ram and the tree would give the impression of the tree coming out of the ram’s head like a giant horn or antler when the cap was worn. As a hooved, horned herbivore, the ram does fit well into the animal symbolism which often accompanies the tree of life in shamanism of Siberian origin.

Figure 3

It is likely that the golden tree would have emitted a sound when worn because of the dangling disks, making it a stick rattle. The hammered shaft which makes up the trunk is not very wide and many of the disks are far enough from the shaft that they would not easily jangle against it or make much sound while doing so. Instead, the disks are more likely to make a sound banging against each other, producing a tinkling or jingling sound. The pearls seem to be placed a bit too high up on the chains to contribute much to sound production. According to the probability grouping, the headdress could belong in Group 4, as it would likely almost always make a sound when worn. It could also be argued that the headdress should be placed in Group 3, but it is difficult to confidently place it in that group as there is no evidence as of yet for sound production being one of its specific purposes.

The other headgear found at Tillya Tepe is a crown found in Tomb VI (Figure 4), belonging to a female. The crown is nine centimeters tall and 45 centimeters wide and was made by being cut out of gold sheets (Hickman, 2012: 84; Peterson, 2020: 52). It is made up of five symmetrical stylized trees, which could possibly be interpreted as antlers, all of which are decorated with openwork patterns, topped by a finial in the shape of a spade, and fastened vertically to a band or a diadem (Leidy, 2012: 114–115; Peterson, 2020: 52–57). Four of the trees are identical, with two birds facing each other on the upper branches and reptilian or piscine creatures by each trunk. The fifth tree is the central one and it is taller, broader, and cut differently compared to the other four, with no birds or creatures (Hickman, 2012: 84; Peterson, 2020: 52). Fifty-three flowers, which were inlaid with some sort of turquoise in the past and have curved petals, are attached to the trees and the band with twisted golden wires and gold disks hang from these flowers (Peterson, 2016: 187; Peterson, 2020: 52). The crown’s parts can be disassembled and folded together by removing the trees, making the crown easily portable, which is a further indication of a connection with nomadic steppe traditions (Leidy, 2012: 114–115; Peterson, 2020: 52).

Figure 4

The crown has many of the visual characteristics of traditional steppe headdresses. Although the tree imagery is perhaps more dominant than the antler imagery, a tree with birds perching on the branches and animals at its base is indicative of the tree of life connecting the upper, middle, and lower worlds. The creatures at the base of the trees are unusual, however, as the steppe perspective tends to favor horned herbivores. If the tree of life represents the three-fold categorization of the world, then the reptiles or fish could be interpreted as some sort of subterranean netherworld or abyss (Peterson, 2020: 63). The crown also has influences other than the nomadic, as it is lower than nomadic headdresses usually were, and it has an overall shape possibly more reminiscent of Iranian female crowns. Furthermore, its flowers resemble narcissi, which might indicate a Palmyrene and/or Graeco-Roman connection (Peterson, 2020: 52–53, 65).

Despite the flowers not being visually connected to the steppes, they are the crown’s most likely source of sound, as their petals and dangling disks are quite close to the frame and would have hit the frame when the crown was worn. As they are small and thin, they would likely emit a jingling sound, although probably a slightly different one compared to the disks on the headdress from Tomb IV because of the size difference. Considering that there are very many round disks fastened to the frame of the crown, the jingling would probably be quite noticeable. This would make the crown a pendant rattle, but as there are also disks dangling from the base of the crown, it could be argued that it is also a stick rattle. Furthermore, since the trees are detachable, they could possibly rattle in their slots when the crown was worn, making the crown a sliding rattle. Like the headdress from Tomb IV, the crown could fit into Group 4 of the probability grouping, as it is very unlikely it would not emit any sound when worn. The aural effects of the crown from Tomb VI would probably be quite strong, more so than of the headdress from Tomb IV, and so there is a strong case to be made for the crown to be placed in Group 3. The same limitations do, however, apply, and as evidence of sound production being one of the crown’s specific purposes is lacking, the crown cannot be exclusively placed in Group 3.

The Tuoba of the Xianbei Confederation

The Xianbei were a confederation of tribes that took over from the Xiongnu in the steppes north of China in the first century CE, and by the third century their confederation spanned the entire northern Chinese border (Watt, 2004b: 7). One of the two main branches of the Xianbei was the Tuoba Xianbei, whose main territories were north of modern Shanxi Province and who established the Northern Wei dynasty (386–534 CE) (Holcombe, 2013: 9; Watt, 2004b: 7–8). The steppe roots of the Tuoba Xianbei can, for example, be seen in their material culture, which was heavily influenced by the Xiongnu until the beginning of the fourth century CE, and they themselves left their mark on later steppe cultures in that the famous title ‘khan’, used by rulers of the steppe, may originate from the Tuoba Xianbei (Holcombe, 2013: 15–18). One type of artefact which very much connects the Xianbei to the steppes are buyao, golden headdresses or ornaments resembling trees with dangling leaves. Most of the steppe buyao which have been found have been attributed to the Murong Xianbei, the other one of the two main branches of the Xianbei, but when it comes to nomadic cosmologies and shamanism, examples of buyao tentatively attributed to the Tuoba Xianbei are of more interest.

In 1981, two pairs of buyao, one with a bovine-headed base and another with a horse- or deer-headed base, were found in Inner Mongolia, Xihezi, Baotou City, located in the Damao Banner (Laursen, 2011: 74–77). The buyao pairs were part of a small hoard of gold buried approximately half a meter in the ground. They have been attributed to the Tuoba Xianbei because their find location is known for Tuoba Xianbei activity and because of some differences when compared to the typical buyao attributed to the Murong Xianbei, although there is a possibility that the pairs are a continuation of traditions inherited from the Murong Xianbei and/or the Xiongnu (Laursen, 2011: 76–77; Watt, 2004c: 129). The specific dating for the buyao, however, is much more difficult than attributing them to a culture. Yangzhishui (2022) attributes the buyao pairs to the late third or the early fourth century while Watt (2004c: 128) extends this to the fifth century and Laursen (2011: 76) mentions that the excavation report assigned the entire hoard to the early sixth century.

As the two ornaments which make up each buyao pair are very similar and were presumably always worn together, each pair is considered here as one headdress. Both buyao pairs are quite symmetrical, regular, and precisely made. Their branches are quite thick, the end of each wire is looped, and a thin but relatively large leaf shaped like a droplet hangs from each loop (Laursen, 2011: 74–75). The pairs were probably made to be attached to caps, the bovine-headed one probably by being stitched or riveted onto a cap while on the deer- or horse-headed pair the sides have been pierced by the tip of the nose and by the ears, so it could have been affixed to a cap through those piercings.

The bovine-headed buyao pair (Figure 5, right) is almost 18 centimeters tall and each ornament weighs around 90 grams (Laursen, 2011: 74). The antlers, which protrude from between the bovine’s ears, resemble branches of trees. They are made with hammered wires and are detachable (Stein, 2017: 10). The bovine head is inlaid with semi-circled stones of different colors, many of which did not survive until the present day (Laursen, 2011: 74). There is granulation around the face and down the nose, as well as around the eyebrows and eyes, which were presumably also represented by stones in antiquity. The ears are long and pointed and attached to the base above the eyes by being inserted into holes. Like the antlers, the ears end in loops from which a single leaf hangs. Behind the base and hidden by the leaves, the ends of the ears and the antlers are joined together, and as they are not soldered together, they can move, although only slightly. There are four round attachments on the sides of each base, one of which seems to have broken off and been repaired in antiquity, although with considerably less skill than with which the rest of the ornament was made (Laursen, 2011: 260).

Figure 5

The deer- or horse-headed buyao pair (Figure 5, left) is 16.5 centimeters tall and each ornament weighs around 70 grams (Laursen, 2011: 75). Like the bovine-headed pair, the deer- or horse-headed pair has hammered wires as antlers reminiscent of branches which protrude from between the ears on the base with loops on the end from which droplet-shaped leaves are hanging. However, the leaves on the deer- or horse-headed pair are lined with granulation-like bosses and the antlers are completely symmetrical. Unlike the bovine-headed pair which has two antlers which then branch out, the deer- or horse headed pair has three antlers, two branching out and mirroring each other and one standing straight up from the middle of the base. Furthermore, the antlers are lined with granulation and inlaid with stones. Both ornaments originally had ten leaves but are now missing one. The base is much narrower on the deer- or horse-headed buyao pair compared to the bovine-headed pair, making the animal’s snout much narrower, and this is what has led to it being identified as a deer or a horse instead of a bovine. The nose of the deer or horse is made with granulation and an inlaid stone, and the face is further decorated with inlaid stones and granulation. The ears are fully fastened to the base, inlaid with pink stones, and there are no leaves attached to them.

The two Tuoba Xianbei buyao pairs are among the headdresses Stein (2017: 10) mentions as part of the tradition connecting trees and deer antlers based on the similarity of their annual cycles of being lost and regrown. However, to the Tuoba Xianbei, reindeer may have been more important than deer, and it has, for example, been hypothesized that the divine animal which Tuoba Xianbei legends say led them in their early migrations was a reindeer (Holcombe, 2013: 5). Historical, though not contemporaneous, Chinese records describe this animal as ‘a heavenly beast in the shape of a horse, which has an ox-like bellow’, and it supposedly led the Xianbei from modern Heilongjiang Province into Inner Mongolia (Leidy, 2012: 116). Interestingly, today there are Evenki people living in Inner Mongolia and Heilongjiang Province whose traditions include reindeer herding. The herders have a very close relationship with their reindeer, valuing them for the milk and transport they provide humans with in return for fodder, and these reindeer are not killed for their meat (Huang, 2020: 8–9). This close relationship extends to the sacred, as herders were occasionally buried with their reindeer so that they could ride it to heaven. Being hooved herbivores, reindeer seem to fit well among the animals usually connected with the tree of life in the steppes. They also look similar to horses, and it could be suggested that the deer- or horse-headed buyao pair is in fact reindeer-headed. As reindeer also lose their antlers annually, this would still fit into the tradition of connecting trees and antlers. However, despite this spiritual connection to the steppes, there are no historical records of buyao being especially or specifically worn by shamans, and they were instead worn by men and women of the Xianbei nobility (Laursen, 2011: 108–109). It can therefore be inferred that by the time buyao worn by the Xianbei appear in dynastic Chinese histories, these buyao had lost their spiritual meaning but retained their meaning as symbols of secular power.

The term used for the headdresses, buyao 步搖, indicates that they were meant to move, as the word translates to ‘step-sway’ (Laursen, 2011: vi), although the character 搖 can also be translated as ‘shake’. This could be interpreted as evidence that decoration was not the only purpose of the ornaments, and as they were most likely meant to move, there is no reason that their shaking movement should not also relate to sound production. Like the crown from Tomb VI at Tillya Tepe, the antlers on the bovine-headed buayo pair are detachable, and there is therefore a possibility that they would rattle when the buyao was worn. However, the main source of sound would probably be the leaves, as they would likely bang against the antlers and ears making a clanging or jingling sound, especially since the antlers and ears can move. These two different ways the bovine-headed buyao can produce sounds make it a sliding rattle as well as a pendant rattle. The deer- or horse-headed buyao is slightly different, as the antlers and ears are soldered to the base and their ends are not joined together. However, the leaves do dangle and the leaves on the top of the ornament are quite close to each other, so they could hit each other and the antlers. This would mean that the pair would produce similar clanging sounds as the bovine-headed pair, though likely quite quietly, but not the rattling sounds. The deer- or horse-headed buyao can therefore only be classified as a pendant rattle. When it comes to the probability grouping, the bovine-headed pair should be placed in a higher probability group as it has a higher possibility for generating sound. The bovine-headed pair could therefore be placed in Group 4 or possibly, though only tentatively, in Group 3. When it comes to the deer- or horse-headed buyao pair it is much less likely that it was made for sound production and it should therefore be put into Group 5, possibly Group 4 at the very most.

The Silla Kingdom

Silla was a kingdom which existed on the southeastern part of the Korean peninsula from 57 BCE to 935 CE, starting as a collection of town states and eventually becoming the first state to unite the entire Korean peninsula (Lee & Leidy, 2013: 3). Until the sixth century, Silla’s neighbors on the Korean peninsula were Goguryeo, Baekje, and Gaya, but by the mid-sixth century, Silla had absorbed all three states, although it never controlled Goguryeo’s territories beyond the Korean peninsula. While all the Korean states are assumed to have had steppe and nomadic origins, Silla seems to have retained some of its steppe culture for much longer than the others, being the last state to convert to Buddhism as well as the least influenced by China (Bailey, 1994/2008: 381; Nelson, 1993: 243).

In the mid-fourth century CE Silla entered its Maripgan period (356–514 CE), during which it became a unified kingdom for the first time. This coincides with the beginning of Silla’s extensive use of gold, possibly because at the beginning of this period Silla had a close relationship with Goguryeo and the nomadic Former Qin (351–394) through which it may have adopted the appreciation for gold which was so prevalent in the steppes (Ham, 2013: 33; Nelson, 2020: 64). Today, one of the most famous features of Silla culture is its use of gold, and the golden tree-shaped crowns found in Maripgan tombs in modern-day Gyeongju are some of the most famous archaeological artefacts ever discovered in South Korea. They are generally interpreted to mean that royal power and shamanic power were in some way combined during the Maripgan period, especially since other artefacts found in the tombs also hint at shamanic steppe connections (Rosén, 2009: 3).

The two crowns analyzed in this article are the crown found in the Great Tomb of Hwangnam and the one found in the Auspicious Phoenix Tomb. Of all the golden crowns found, only the crown found in the Great Tomb of Hwangnam has been incontrovertibly identified as belonging to a monarch – although the others certainly belonged to members of the royal family – and researchers generally agree that this crown dates to the early fifth century (Ham, 2013: 38–41; Nelson, 2003: 85). The tomb was excavated from 1973–1975 and it is the largest tomb of the Silla kingdom, the final resting place of a husband and wife (Nelson, 2020: 62). As such, it is ‘gourd shaped’, with two circular mounds overlapping each other and one spouse interred in each mound. The crown was found in the northern mound and belonged to a woman, evidenced by a belt which was among the grave goods and had the inscription ‘belt for milady’ (Nelson, 2003: 85). The Auspicious Phoenix Tomb was excavated in 1926 and is thought to date to the early fifth to early sixth century, possibly either 451 or 511 based on an inscription found on one of the artefacts in the tomb (Ham, 2013: 40; Lee, 2010b: 52).

Both crowns consist of a golden circlet to which golden antler- or branch-like uprights are connected (Bailey, 1994/2008: 368). Golden leaf-like spangles and jade and glass gogok, precious comma-shaped embellishments, are fastened onto the uprights with wire, and golden streamers, also decorated with various attachments, hang down from the circlet (Bailey, 1994/2008: 368; Nelson, 2020: 65–66). Both crowns have five uprights affixed to the circlet with spade-shaped finials, perhaps signifying unopened flower buds; three connected branch-like ones in the middle and one antler-like upright on each side (Lee, 2010b: 25, 55). The crowns were made individually, and testing has shown that the gold used in making them is very close to being pure, with any impurities likely being natural (Nelson, 2020: 63). As the crowns are quite flimsily constructed it has been suggested that they may have been made for funerary purposes only, but they do show some signs of wear and tear, so it is likely that they were also used in rituals or ceremonies in life (Bailey, 1994/2008: 368; Portal, 2000: 61–62).

On the crown found in the Great Tomb of Hwangnam (Figure 6) the uprights, decorated with leaf-like spangles and gogok, are fastened with triangular-shaped gold rivets on the circlet’s inner side, and they are heavier on the top than they are on the bottom (Lee, 2010b: 26). The crown is 27.3 centimeters tall (Ham, n.d.). There are six streamers, three on each side, attached to the circlet with thick rings which are hollow on the inside, and thin leaf-like spangles and gogok hang from the streamers (Lee, 2010b: 27). The crown found in the Auspicious Phoenix Tomb (Figure 7) also has small leaf-like spangles and gogok attached to all five uprights (Lee, 2010b: 55). It has only one streamer on each side decorated with leaf-like spangles ending in three elongated leaves, attached to the circlet with a thick ring. There is also a dangling intermediary decoration attached to the circlet. The crown has an unusual feature in that it includes an inner headdress made up of two gold plates, with small leaf-like spangles attached to them, which form a cross on the top of the head, creating the shape of a hat. On top of the cross is an upright in the shape of three birds, possibly phoenixes, sitting on branches of a tree (Figure 8). The crown is 36 centimeters tall and 18 centimeters wide (Kwon, 2023: 107).

Figure 6

Figure 7

Figure 8

The tree symbolism in the Silla crowns is generally considered to be similar to that of Siberia, where the tree connects the shaman to a heavenly god (Lee, 2010a: 115). The antlers represent the deer which leads souls to heaven after death, a role which Lee (2010a: 115) links to Siberian traditions as well. Both the tree and the deer could have been important sacred symbols for the royal family, used to strengthen its lineage. While the gogok decorating the Korean crowns are not present in the Xianbei buyao or Tillya Tepe headdresses, they are nevertheless considered to have shamanic meaning. Grayson (1984: 188) points out that gogok bear a resemblance to claws of bears or tigers, which modern Korean shamans use for animistic power. Lee (2010a: 115–116), on the other hand, discusses two different interpretations. The first is based on the resemblance between the shape of the gogok and that of a fetus, or some other life form in a primitive state, making the gogok a symbol of life, rebirth, or a wish for new life. Lee’s other interpretation is that the gogok may represent fruit which hangs from a tree, symbolizing the continuation of existence, as well as the continuation of the royal family.

The crown found in the Auspicious Phoenix Tomb is different from the other golden crowns found in Korea in two ways: the inner headdress and the upright coming out of it with birds on top. While this type of inner headdress is unusual in Korea, it does have parallels in Siberian cultures, for example in Ket shamanism where it symbolizes the four cardinal directions and the sacred center (Taksami, 2001: 17). The birds, however, are not unusual for Korea, despite generally not being present on the tall golden crowns. Birds in Korean culture have been interpreted as messengers between earth and heaven, but they also play a large role in Korean folk tradition, as some villages have a wooden pole topped with a bird standing at their entrance in order to guard against evil spirits (Pak, 1988: 48). This could perhaps be related to birds sitting atop the ritual posts representing the tree of life in Siberian shamanism. Additionally, the Record of the Three Kingdoms, written in China in the third century CE, states that the Silla people’s predecessors had the tradition of including birds’ wings in graves (Nelson, 1993: 252).

Although both crowns are covered in spangles and gogok, these are unlikely to emit any sound, as the gogok are fastened quite tightly to the uprights, and the wire fastening the spangles is twisted too far from the uprights for the spangles to be able to bang against them. The streamers and intermediary decorations, however, are a different matter. The spangles on the streamers are attached close enough to be able to clang against the chain and each other, and the streamers themselves and the decorations could jingle and tangle with each other as well. Both Portal (2000: 62) and Nelson (1993: 252) mention how tinkling would have been part of the dazzling effects of the crowns when worn. The spangles on the steamers fit the description of strung rattles while the streamers themselves can be classified as pendant rattles. The crowns could fit the criteria for Probability Groups 4 and 3, but it is difficult to argue which group would fit better. There is no evidence to support the claim that one of the crowns’ specific purposes was sound production. However, the sound-producing nature of the streamers and dangles likewise makes it difficult to argue that the crowns were intended to be silent.

Possible Meanings and Interpretations

Before discussing shamanism and sound properties, it is useful to briefly compare some of the metalworking techniques which connect the headdresses, as well as their visual imagery, in order to understand the multiple ways in which they are connected. The goldsmithing techniques used in the production of the crown from Tomb VI at Tillya Tepe and the Silla crowns are remarkably similar in the way the leaves are made of thin, cut gold and attached to the branches with twisted wire (Pak, 1988: 46). This is different from how the leaves are fastened on the buyao pairs, where the ends of the branches are looped to hold the leaves in place, and it is also different from the headdress found at Tillya Tepe in Tomb IV, where the leaves are hung from chains. Another part of craftsmanship to consider is inlay, a technique which has been connected to the steppes (Pak, 1988: 52; Stein, 2017: 19). The buyao have inlay and the Tillya Tepe Tomb VI crown shows signs of having had inlay in antiquity as well. While the Silla crowns do not have inlay, perhaps the gogok decorating them are an echo of this inlay tradition. Finally, granulation is present both in the buyao pairs and the Silla crowns. This is especially interesting as granulation is not thought to have been invented in multiple cultures, but rather to have originated in one steppe culture, possibly with the Scythians, and then spread from there (Stein, 2017: 19–20).

The headdresses all share the visual imagery of trees or branches and dangling leaves. The tree imagery is perhaps clearest on the headdresses from Tillya Tepe Tomb IV and the bovine-headed buyao. While the imagery on five of the headdresses could be interpreted as antlers instead of branches, the Tillya Tepe headdress from Tomb IV does not have an antler shape. Additionally, in the Tillya Tepe crown from Tomb VI the antler imagery is quite vague and overshadowed by tree imagery. There are land or sea animals at the base of four of the headdresses (from both Tillya Tepe tombs and in both buyao pairs) and birds on the top of two of them (from Tillya Tepe Tomb VI and the Auspicious Phoenix Tomb).

When it comes to the headdresses’ classification according to Hornbostel and Sachs’ (1961) system, the only headdress which does not fit under their classification for pendant rattles is the one from Tomb IV at Tillya Tepe. It can be concluded that instead of the headdresses having different sound properties because of different classifications, the difference in sound can be traced to the size of the leaves, how many leaves there are, and how closely together they have been placed. This difference in sound volume and how easily the sound is produced is what led to the different placements in the probability groups, with the Tillya Tepe and Silla headdresses being placed in higher groups than the buyao pairs. There is therefore a higher possibility of the Tillya Tepe and Silla headdresses having been used as sound tools. The different features of the headdresses, both aural and visual, are compared in Table 2.

The headdresses can all be assumed to make similar jingling or tinkling sounds, presumably resembling bells, as the leaves bang against the branches they are strung from or, in the Silla case, against the streamers they are suspended from and the streamers against each other. Jingling sounds are very important in North Eurasian shamanic soundscapes and the most common visual imagery associated with jingling is the tree of life. These jingling sounds could be a symbolic representation of the sounds leaves make in nature, fitting the cultural practice of having music imitate natural sounds, as jingles and bells used in modern shamanic rites in Korea and Mongolia sound similar to the sounds of wind and rain coming into contact with leaves (for recordings of shamanic bell sounds see, e.g., Mongolian Ways, 2020; Penn Museum, 2013; 스브스뉴스 Subusunews, 2019). This interpretation would fit with the jingling sounds made by the crowns: tree-like sounds made by artefacts with tree-like imagery. Additionally, as the sound of jingles and bells is known to be important in ‘help[ing] the deceased in the journey to the other world’ (Hasanov, 2016, p.202), it fits quite well with the meaning of the headdresses’ visual imagery of rebirth and contact to different worlds and even with the possible meanings of gogok. Perhaps the jingling sound also had this same meaning of rebirth and renewal of life.

This connection between jingling sounds and headdresses fits with examples of bells being fastened to shamanic headdresses in Siberian Kacha culture (Hasanov, 2016: 198). It can be presumed that these bells would likely make sounds similar to the headdresses studied here. The sound made by the headdresses would presumably be produced as the wearer moved, which fits with more examples of sounds and movement being connected, as jingles or bells were sometimes fastened to sleeves, hats, the edge of cloaks, or even the footwear of shamanic costumes in Siberia (Hasanov, 2016: 198). In this way, the headdresses hint at a multisensory cultural perception of movement, aurality, and vision. Because of the way the leaf-like spangles are made and because of the reflective properties of gold itself, the headdresses shimmer and catch the light differently depending on movement. This suggests a relationship between movement and sight. Furthermore, the Chinese name for these headdresses, buyao, step-sway or step-shake, suggests that the wearer’s movements were an important part of allowing others to appreciate the ornaments. Sun (1991: 55) even specifically mentions that the leaves on the two Tuoba Xianbei buyao pairs can move, indicating that this was an important aspect of the ornaments. As jingling sound would have been produced as the wearer moved, this suggests a relationship between movement and sound. Finally, there is a link between the meaning of the headdresses’ imagery and the sound they produce, as the sound of bells helps with transition to the next world and its closest related visual symbol is the tree which is the connection of the three different worlds which make up the universe. The tree is, in turn, seen as almost interchangeable with antlers, which signify rebirth, and often has birds in its branches, who represent souls waiting to be reborn or act as messengers between worlds.

However, the jingles and bells used in shamanic rites are quite loud while the sounds the headdresses produce would likely be much quieter, and it could be assumed that as the sound becomes quieter, its importance and sacredness diminish. This interpretation would mean that the headdresses examined here are not all equally sacred, as some of them are presumed to have produced a louder sound than others. This difference in sacredness could be explained with the difference in the strength of steppe influences in the cultures that produced the headdresses. The people who are buried at Tillya Tepe are considered to have had very strong connections with the Eurasian steppes, and during the Maripgan period in Silla the kingdom was experiencing very strong steppe influences. However, despite having steppe origins, the Tuoba Xianbei had much closer ties with China than the people of Tillya Tepe and the Silla Maripgan period ever did, even founding the Northern Wei dynasty, and ruling over all of North China by 439 CE (Watt, 2004a: xix). It is therefore not surprising that the makers of the buyao pairs would be the least concerned with the sound symbolism of the steppes. This might suggest that as the cultures turn away from steppe influence, the jingling sound properties lose their importance, which, in turn, could be an indication that while the Tillya Tepe headdresses and Silla crowns were ritual objects as well as decorative, the buyao pairs did not have a sacred purpose. This is further evidenced by the dynastic Chinese descriptions that do exist of the buyao, which do not mention shamans or sacredness and instead say that those who wore buyao were members of Xianbei nobility (Laursen, 2011: 108–109).

In addition to the sound properties of the headdresses being connected to the steppes, it can be suggested that they were also connected to rulership. As discussed earlier in this article, the person who controls sound also controls its magical quality, and if sound is considered sacred then it adds to the sanctity and divine authority of the rulers of states who have this control over sound. The people buried at Tillya Tepe are presumed to have been the rulers of the nearby settlement of Yemshi Tepe and the Silla crowns were found in royal tombs. Those who have researched the headdresses of Tillya Tepe and crowns of Silla, as well as other similar ones found throughout the steppes, have agreed that they are symbols of authority, most likely both secular and spiritual (Rosén, 2009: 4; Shenkar, 2017: 156). The evidence for this is their visual symbolism, such as trees, antlers, and birds, as well as the gold of which the headdresses are made. As these visual attributes have been connected to royal authority, the sounds this type of headgear produces should be taken into consideration as possible evidence for authority as well, as the control of sound could have been restricted to those who had some sort of authority, either spiritual or secular or both. As for the buyao, although it is difficult to say anything for certain about their owners, they are usually discussed as ornaments for decoration, not as symbols for sacred authority (see, e.g., Laursen, 2011: 94–121; Xinhuanet, 2022), which is further supported by their quiet and restricted sound properties compared to the Tillya Tepe headdresses and Silla crowns.

Unfortunately, considerations about the importance of the movement of these six headdresses and the sounds they might produce can only remain speculative, as it is impossible to know how the cultures of the people of Tillya Tepe, the Tuoba Xianbei, and Silla sensed the world or if they did so in ways similar to each other. Additionally, it is impossible to definitively conclude what the meaning of the headdresses’ sound was, as intent can depend on a culture or even the individual (Zubrow & Blake, 2006: 122), and artefacts which to the modern eye seem similar, such as the headdresses studied here, may have been made with completely different intentions. Perhaps some of them were made with the intent to be used as sound tools while others were not, despite all of them having similar sound attributes. Furthermore, even if the sound is intentional, it may not have a deeper meaning. However, the sound produced by the headdresses cannot be completely accidental, as those making the headdresses must have known they would produce sound, based on former experiences with such objects. Additionally, even if their sound did not originally have a deeper meaning, it would still have contributed to the atmosphere of the rituals where the headdresses were worn, likely giving the sound at least some importance through practice. The headdresses were undoubtedly part of the soundscape the ancient peoples of the steppes experienced and studying them provides modern societies with a fuller understanding of the world in which these ancient peoples lived.

Conclusion

Based on the large role musical instruments and acoustic sites play in shamanic rituals of Siberian origin (see, e.g., Díaz-Andreu et al., 2023; Hasanov, 2016: 194–202), it seems quite clear that sound was an important element in these rituals. Studying possible sound tools other than those which can easily be designated as musical instruments based on their visual attributes is one way to try to circumvent modern Western conceptions of the ways artefacts should have meaning and be connected, and thereby possibly help us gain a better understanding of ancient societies. The headdresses studied here are connected in many ways through both visual imagery and metalworking techniques, and if visual shamanic symbolism and craftsmanship can be transmitted from one culture to another, then there is a possibility that the sacred meaning of sound could be transmitted as well.

Considering that sounds in shamanism are known to sometimes represent nature and that jingling sounds are commonly connected with the tree of life, it seems likely that the sound emitted by the headdresses was made to represent the sounds leaves make in nature. Additionally, jingling bells are used in shamanic rites to help souls transition to the other side, and so this practical function of jingling also fits well with the meanings of rebirth, renewal of life, and communication with worlds other than our own so frequently associated with the tree of life and the animals and birds which most often accompany it. This interpretation connects the headdresses’ aural and visual meanings, which indicates that the headdresses may have been perceived in multisensory ways by the ancient cultures that made them, an indication which is further supported by the shimmering visual effects the headdresses would have produced when worn. Movement could be suggested as an additional part of this multisensory perception, because the sound and shimmering effects would be produced when those wearing the headdresses moved. Finally, if it is assumed that the sacredness of the sound produced by the headdresses was connected with rulership, like the headdresses themselves, then this sacred sound was probably a cultural element from the steppes which gradually lost its importance when steppe cultures became more Sinicized. This is evidenced by the buyao pairs being presumably the quietest and being secular rather than shamanic, their sound properties indicating that the buyao had gone through a transition from being an indicator of sacred power to an indicator of secular power or nobility. A further indicator of sacred sounding headdresses not being part of the culture of central China is the fact that such headdresses disappear from royal Silla tombs after increased contact with China and with Buddhism becoming a state religion (Ham, 2013: 51–53).

Based on the case study presented in this article, it seems that the sound produced by the golden tree-like headdresses from North Eurasia, dated to the first half of the first millennium CE, was possibly significant to ancient steppe cultures and shamanism in several different ways. It is, however, important to note that the theories and conclusions presented here are general interpretations and that the meaning and importance of sound likely varied. There must have been cultural differences and nuances between peoples of the steppes, especially considering the large time frame and vast geography. This could, for example, have presented in some interpretations being more important than others depending on the culture and time period, or traditions being preserved in some cultures and forgotten or replaced in others. There are also, of course, limitations to this study, the most obvious of which is that it was not possible to examine the headdresses physically. This means that doing direct testing on the sounds produced by the headdresses was not possible, and that the types of sound they may make are deductions based on the material of which the artefacts are made, as well as the size, shape, and proximity of their components. This invariably affects the interpretations and conclusions of this case study, which will have to be corroborated with further research on the headdresses directly. Another limitation is that the six headdresses studied here are only a small example of the amount of tall gold headdresses which have been found in North Eurasia. This leaves ample room for future research on the significance of the sound produced by gold Eurasian headdresses by including more headdress examples, through cooperation with the museums where the headdresses are housed, or even including ethnographic evidence from the nomadic cultural groups which still exist in North Eurasia for better social context and comparisons. There is still much we do not know or understand about ancient cultures, but this article has hopefully added a new perspective on how the ancient peoples of the steppes may have experienced their world.

Competing Interests

The author has no competing interests to declare.

This article was originally written as an MA dissertation at the Institute of Archaeology, University College London.