Oxus trumpets are shorter (ca. 10 cm in length) than modern trumpets, but like modern ones they have a flaring bell at the front and a mouthpieces at the back. The most common material is silver, but copper, gold, lead, and gypsum are also used. Some are decorated with human and animal faces of high artistic merit.


OXUS TRUMPET, an artifact mostly found in the border area between northern Afghanistan and southern parts of the former USSR, with a few stragglers further south in eastern Iran (see map, Figure 1). Oxus trumpets were used ca. 2200–1800 BCE.

They are shorter (ca. 10 cm in length) than modern trumpets, but like modern ones they have a flaring bell at the front and a mouthpieces at the back. The most common material is silver, but copper, gold, lead, and gypsum is also used. Some are decorated with human and animal faces of high artistic merit (Lawergren, 2003).

Figure 2. Extant trumpets of three types.

Figure 3. Unexcavated face trumpets.

Figure 4. The four trumpets of Figure 3.

Figure 5. Trumpets with a bulbous expansion between bell and mouthpiece.

The first excavated trumpets were found 1841 at Astarabad, Iran (Figure 2h–i; Rostovtzeff, pp. 17–18), later identified with Turang Tepe (Wulsin, p. 2). Both were made of gold. Three more Iranian trumpets were excavated 1931 at Tepe Hissar (two silver, one gold, Figure 2b–c, g; Schmidt 1937, p. 210). A copper trumpet was reported 1997 from Šāhdād, Iran (Figure 2k; Hakemi, pp. 245–46, 635).

But the scene switched to the South Bactrian oasis between Dawlatābād and Balkh when looters found many trumpets there during the 1970s (Pottier, nos. 314–15; Figures 4–5). But lately they have again turned up in legitimate archaeological digs in Margiana, a nearby region (Figure 2a, d–e, j; Sarianidi, p. 239; Lawergren 2003, p. 48). All belong to the Bactria Margiana Archaeological Complex (BMAC; see ARCHEOLOGY v. PRE-ISLAMIC CENTRAL ASIA).

Frequently the trumpets have a bulbous expansion between the bell and the mouthpiece (Figure 5), but the most spectacular ones have faces instead of bulbs (Figures 3–4). Mostly the faces are human, but one with three bison faces displays formidable craftsmanship. It is a double-trumpet with an outer layer of gold hammered to fit an inner copper trumpet. The gold faces are rendered with fine details, such as rows of body hair, a pointed beard, eyelids, lips and nostrils. Each bison has silver horns set into the gold layer. A few trumpets have plain conical shape without an obvious bell (Figure 2b–c).

The mouthpieces follow a design different from modern ones. Instead of pressing against the outside of the player’s lips, they are inserted between them without touching the very back of the fleshy lips. In playing, this inner part of the lip vibrates. The short mouthpiece expands toward to bell, whereupon it abruptly constricts to the diameter of the tube.

Are they trumpets, i.e., were they meant to be blown to give sound? Early investigators were ambivalent; Schmidt (p. 210), for instance, called them problematic pieces but two circumstances now point to a positive answer. First, the Yima story in Vendidād relates how the earliest man (King Yima) used a trumpet (suβrā-, cf. Pers. sru, sarun, “horn”) to call animals and humans and lead them in a certain direction (J. Duchesne-Guillemin). According to Boyce (p. 94), the legend has pre- Zoroastrian elements and may have arisen before 1200–1000 BCE (Lawergren 2003, pp. 94–96). The gap between the rise of the story and the archaeological demise of Oxus trumpets may be no more than a few centuries and oral traditions could have bridged it. The efficiency of the trumpet sounds is still attested in stalk hunting. The hunter mimics the sound of the prey attempting to lure it. If it sounds like the calls of female deer in heat, male deer may approach.

Second, the acoustical properties of the instruments are well suited for the production of sounds similar to those of female deer (Lawergren, 2004). The trumpets differ greatly from modern (post 1500 BCE) instruments in two major ways, and both cooperate in the sound production. The short length results in the high pitch of a deer, and the uniquely designed mouthpiece facilitates high lip frequencies. Unlike modern mouthpieces, Bactrian ones do not force the full lip to vibrate, but select a minute part. Its small mass allows high frequencies without excessive lip tension. Also, the short air column of the trumpet allows the fundamental pitch to be ‘lipped’ up and down.

The high-pitched sound (about two octaves above middle-c) lies near that of desirable prey (e.g., deer and gazelle) that inhabited the Bactrian region before it became a desert, and the ability to slide the pitch helps mimic their sounds. The trumpet could have function as a lure call (as in the Yima legend). They had a utilitarian use rather than a ‘musical’ one.



M. Boyce, ed., tr., Textual Sources for the Study of Zoroastrianism, Totowa, N.J., 1984.

J. Duchesne-Guillemin, “Cor de Yima et trompette d’Isrāfīl; de la cosmogenie mazdéenne à l’eschatologie musulmane,” in Institut de France, Académie des inscriptions et belles-lettres. Comptes rendu des séances (juillet-octobre) 1979, pp. 539-49.

A. Hakemi, Shahdad, Archaeological Excavations of a Bronze Age Center in Iran, Rome, 1997.

B. Lawergren, “Oxus Trumpets, ca. 2200–1800 BCE: Material Overview, Usage, Societal Role, and Catalog,” Iranica Antiqua 38, 2003, pp. 41–118.

Idem, “The Acoustical Context of Oxus Trumpets,”. in Proceedings of ISMA 2004, March 31st to April 3rd, 2004, at Nara, Japan, CD-ROM (ISBN4-9980602-4-4), M.-H. Pottier, Matériel funéraire de la Bactriane Méridionale de l’âge de Bronze, Paris, 1984.

M. Rostovtzeff, “The Sumerian Treasure of Astrabad,” Journal of Egyptian Archaeology 6, 1920, pp. 4-27.

V. I. Sarianidi, Margash. Ancient Oriental Kingdom in the Old Delta of Murghab River Ashkhabad, 2002 (texts in Turkmen [romanized], Engl. and Rus.).

E. F. Schmidt, Excavations at Tepe Hissar [Damghan], Philadelphia, Pa., 1937.

F. Wulsin, “Excavations at Tureng Tepe, Near Asterabad,” Bulletin of the American Institute for Persian Art and Archaeology 2, 1932, pp. 2-12.

(Bo Lawergren)

Originally Published: July 20, 2005

Last Updated: July 20, 2005