Sign up for our Flute Newsletter


Previous PageNext Page

Flutes of Gilgamesh and Ancient Mesopotamia

The Epic of Gilgamesh, the greatest literary work of Ancient Mesopotamia, talks of a flute made of carnelian, a semi-precious stone. The passage was recently identified on cuneiform tablets written in Akkadian, an ancient semitic language.

When I happen to read about this new passage in Gilgamesh, it got me interested in flutes of the Ancient Near East. What were the style of flutes? How were they tuned and how were they played?

After hunting down the musical artifacts from that time, as well as literary references, the use of words in various languages, and images in sculpture and other visual art, we can get a composite image that tells us a great deal about the flutes of Ancient Mesopotamia. And amazingly, Archaeomusicologists identified a handful of ancient Mesopotamian texts written in cuneiform on clay tablets as precise tuning instructions for their stringed instruments. The information on these various “musical texts” dates from about 1800-600 BCE and has been the subject of intense analysis and debate for the past 45 years.

This pages follows the evidence from visual art, artifacts, replicas of instruments, the languages of the time, written literature, and the set of musical texts. This composite view leads to some amazing correlations between flutes of Ancient Mesopotamia and early Native American cultures.

Gilgamesh mourning Enkidu - from The Revenge of Ishtar ©1993 by Ludmila Zeman

The Flute of Gilgamesh

Gilgamesh, the great king of Uruk, embarks on a quest for fame and glory. He travels to the Forest of Cedar with, Enkidu, a mortal created by the gods to be Gilgamesh's friend and counterpart. The two heroes conquer the ogre Humbaba and the firey bull of the goddess Ishtar. For this latter deed, the gods give Enkidu a vision of Hell and condemn him to death. Gilgamesh holds a lavish funeral for Enkidu, and the process comes to yearn for a greater quest than fame and glory: immortality.

Gilgamesh wanders to “beyond the edge of the world” in search of the survivor of the great flood, who holds the secret of eternal life. There he finds the immportal Uta-napishti, who tells the story of the Great Flood. To the dismay of Gilgamesh, he also explains how the destinies of the immortal gods and mortality of mankind cannot be changed.

The Epic of Gilgamesh is known to us from cuneiform tablets scribed in Ancient Mesopotamia. The written record is fragmentry and has been pieced together over the last 160 years from tablets from various sources, locations, and timeframes between 1700 BCE and 600 BCE. A number of missing passages have recently been found, including a description of the funeral of Enkidu.

As Gilgamesh prays to the gods of the Netherworld he names the gifts he is burying with Enkidu for his journey in the afterlife. Here is my version of the new text, based on the translation from [George-AR 1999] (also [George-AR 2003] and [George-AR 2003a], pages 67–68), from Book VIII of the Epic, lines 144–149:

He displayed to the Sun God a flask of lapis lazuli
   for Ereshkigal, the queen of the Netherworld:
"May Ereshkigal, the queen of the teeming Netherworld, accept this,
   may she welcome my friend and walk by his side!"

He displayed to the Sun God a flute of carnelian
   for Dumuzi, the shepherd beloved of Ishtar:
"May Dumuzi, the shepherd beloved of Ishtar, accept this,
   may he welcome my friend and walk by his side!"

History of the Text

This newly found section of text comes from the latest and most complete version of the Epic, known in antiquity as “He who saw the deep” (Akkadian: Sha naqba imuru), and now known as The Standard Babylonian Gilgamesh Epic. This version comes from the library of the Assyrian king Ashurbanipal (668 - 627 BCE). However, the history of the poem stretches much further back in history. Most Assyriologists believe that the final form of the poem was written by a scholar (and professional exorcist) named Sîn-liqe-unninnī in about 1200 BCE.

Sections of three earlier versions of the epic poem from about 1700 BCE are similar to each other and similar to the Standard Version, but not identical. The 1700 BCE tablets were know as “Surpassing all other kings” (Akkadian: Shutur eli sharri). One version from that era, a single section from the Quest for Immortality chapter of the epic, was assembled by joining fragments in museums in London and Berlin. It cotains a famous scene of an old goddess who dispenses wise advice from her “tavern at the end of the world”.

The Epic itself appears to be an expansion of the collection of five Poems of Gilgamesh, written in Sumerian and probably composed for Shulgi, second king of the Ur III Dynasty (2094-2047 BCE). We have one of the poems, Gilgamesh and the Netherworld, because it was translated into Akkadian and added to the last tablet of one of the version from the Standard Version ([George-AR 1999]).

The first texts were excavated by Austen Henry Layard and Hormuzd Rassam at Nineveh in 1850 and 1853, and presented a huge tasks for the new field of Assyriology - coming to understand new languages and deciphering hundreds of thousands of tablets housed in museums around the world. Much of the early work was considered arcane research until George Smith, after having discovered the portion of the Epic of Gilgamesh dealing with the Great Flood, gave a public lecture. He drew parallels between the legend from anicient Iraq and the Biblical story of Noah, and received worldwide press coverage. It became clear that the legend of the Flood goes back to at least 1750 BCE and contained source material for the Book of Genesis.

For more background information, see the SOAS web site on the Epic of Gilgamesh. For parallels between the Epic and the Bible, see [Heidel 1963].

However, the Epic of Gilgamesh is not a static text. Researchers continue to find new fragments of text for the Epic as well as the five Sumerian poems that mark the distant past of the 4000-year history of the story. At present, we have about 65% of the text, including some lines that are damaged and only partially readible ([George-AR 1999]).

The World of Ancient Mesopotamia

There are many sources of background information on the early literate civilizations in Ancient Mesopotamia, Babylonia, Assyria, and the Fertile Crescent. In the way of context pertaining to this article, here is a map that highlights many of the key areas and issues discussed:

  • Ancient Near East - Topographic
  • Ancient Near East - 2250 BCE Trade and Resources
  • Ancient Near East - 2000-1550 BCE
  • Ancient Near East - Present Day

Ancient Near East - Development and Key Sites More information


Mesopotamia in the Bronze Age (2900-1100 BCE) is considered to be the cradle of civilization. It was home to a range of languages, spoken and written, that form the core of our understanding of the culture and the music. The texts that pertain directly to music and flutes were written in three ancient languages:


The Sumerians were a non-Semitic, non-Indo-European people who lived in southern Babylonia from 4000-3000 BCE. They invented cunieform writing, and their spiritual beliefs influenced all successive Near Eastern religions, including Judaism, Christianity and Islam. They produced an extensive body of literature, among the oldest in the world ([Kramer 1961]).

Sumerian was the language of the ancient city of Sumer and was spoken in Southern Mesopotamia from the early 4th millenium BCE (4000-3500 BCE) until about 1800 BCE. It continued to be a used as a classical language until about 100 CE, in the way that Latin and Greek have been used in recent and present-day Western traditions.

It is a language isolate, having no known relationship with other natural languages.

Tablet MS 3029

Tablet MS 3029 Larger image


The languages of Ancient Mesopotamia were written in cuneiform, a system of inscribing on wet clay tablets using a wedge-shaped stylus made of reed. The tablets were dried in the sun or baked in kilns to create a permanent record. The system of cuneiform writing had been used as early as 5300 BCE in the proto-writing found recently on the Tărtăria tablets ([Merlini 2008]).

Sumerian was the first written language. Beginning in about 3500 BCE with a primitive writing system during the “proto-literate” period, it gradually developed into a system of logographic signs inscribed in clay tablets, and later added Sumerian syllables to the writing system.

The tablet shown at the right, MS 3029 from The Schøyen Collection, was inscribed in Sumer in the 26th century BCE. It was written by an expert scribe and is the earliest manuscripts describing a religious practice.

We know quite a bit about Sumerian and ancient Semitic languages, in part because of the study of the 500,000 clay tablets and fragments that we have found (estimate of the Cuneiform Digital Library Initiative). Many of the texts are legal and accounting documents, but there are thousands of literary and mathematical documents, and a handful of documents known as the “music texts”.

The chart below, based on a drawing in [Kramer 1988], demonstrates the development of cuneiform writing. Each sign was a picture of one or more concrete objects and represented a word whose meaning was identical to, or related closely to, the object depicted. Drawing was a slow process and the set of syllables grew as writing became more complex. The system evolved into gradually more abstract symbols, using the wedge-shaped reed stylus to make impressions in the clay, and began substituting phonetic values instead of ideographic pictures.

Development of cuneiform writing

Development of cuneiform writing, based on [Kramer 1988] Larger image

The columns represent progressive development through time. The rotation of symbols in column 2 was adopted for convenience of writing, so the tablet was often read with the sign in this position. Row A represents the symbol for “mouth” (Sumerian ka) and also “to speak” (Sumerian dug), with the meaning determined by context. Row B is a bowl for holding food.

Tablet MS 3025

Tablet MS 3025 Larger image


The Akkadian language was the earliest Semitic langauge, a family of languages that includes many ancient Semitic languages as well as modern Semitic language such as Arabic, Hebrew, and Aramaic, now spoken by about 270 million people. It began at least as early as 2800 BCE in the Mesopotamian Valley and was written for about 3,000 years ([George-AR 2007]).

Akkadian and was written in cuneiform as a primarily syllabic systems, and the earlier logographic nature became secondary. Since many people were bilingual in Sumerian and Akkadian during the third millenium BCE, many tablets are written side-by-side in Sumerian and Akkadian.

Assyrian and Babylonain, previously treated as separate languages, are now considered dialects of Akkadian.

The tablet at the right, MS 3025 from The Schøyen Collection, is a section of the Epic of Gilgamesh from about 1900-1700 BCE that corresponds to Book IV of the Epic.


The Hurrian langauge was spoken by the Hurrian people (Khurrites), who inhabited Northern Mesopotamia form about 2300-1000 BCE. They are believed to have originated in the mountainous regions of present day Armenia, and spread through Anatolia (present day Turkey) and formed the Mitanni kingdom in Northern Mesopotamia as well as settlements in present-day Syria.

A Flute of Carnelian

Three tablets containing new text of Gilgamesh

Three tablets containing new text of Gilgamesh Larger image

The original text from the Epic of Gilgamesh that contains the reference to “a flute of carnelian” comes from the last 5 lines of column 3 of the front of a cuneiform tablet. Portions of the tablet were composed from three clay fragments in the British Museum: BM 36909, BM 37023, and F 235. A drawing is shown at the left, with lines 145-150 of Chapter VIII of the poem highlighted.

The original transliteration from cuneiform into Akkadian by A. R. George (provided courtesy of the School of Oriental and African Studies, University of London):

145:   a-na d┌ereš-[k]i-g[al šarrat erṣeti dUTU uk-tal-lim]

146:   lim!-ḫur dereš-k[i-gal] š[ar-rat erṣeti     ]

147:   a-na IGI i[b-ri-ia] lu-[ú ḫa-da-at-ma i-da-a-šú lil-lik]

148:   em-bu-bu šá ┌na4GUG x[     ]

149:   a-na ddumu-zi SIPA na-ram [     dUTU uk-tal-lim]

The key word is em-bū-bu, Akkadian for “flute”. We will look at the word em-bū-bu and many related Akkadian topics below.

The full phrase is “a flute of carnelian”. Carnelian is a semi-precious reddish-brown stone that was mined and processed in the East. It was used as early as 4000 BCE in Mehrgarh, in present-day Pakistan and the site one of the earliest centers of agriculture and herding in South Asia, beginning in about 7000-5500 BCE ([Sharif-M 1999]).

Akkadian Grammar and the word “Embūbum”

The grammar of Akkadian is complex: nouns change form based on their declension, based on their case, grammatical gender, and number (singular, dual, or plural). The word em-būb-um is always masculine, so the variants are (based on [Caplice 1988], page 11):

  • em-būb-um — Nominative (grammatical subject), masculine, singular (eg. “The em-būb-um is made of reed”).
  • em-būb-im — Genitive (grammmatical possession), masculine, singular (eg. “The em-būb-im('s) finger holes”).
  • em-būb-am — Accusitive (grammmatical object of a sentence, among other uses), masculine, singular (eg. “The song is played on the em-būb-am”).
  • em-būb-ān — Nominitive, masculine, dual.
  • em-būb-īn — Genitive and Accusitive, masculine, dual.
  • em-būb-ū or em-būb-ānū — Nominitive, masculine, plural.
  • em-būb-ī or em-būb-ānī — Genitive and Accusitive, masculine, plural (nominitive, masculine, singular in early Sargonic Akkadian).

Grammar also changed over time, so there are additional possibilities from early (“Sargonic”) forms of Akkadian ([Hasselbach 2005], page 161):

  • em-būb-u (Sargonic nominitive, masculine, singular),
  • em-būb-a (Sargonic accusitive, masculine, singular),
  • em-būb-ūt (Sargonic nominitive and accusitive, masculine, plural), and
  • em-būb-ūti (Sargonic genitive, masculine, plural).

From all these choices, I generally use the nominative singlular form of embūbum in mixed Akkadian / English contexts.

Flutes in Visual Art

Next we look at the “iconography” — the visual art of the time that depicted flutes, both alone and with other instruments. Of particular interest are images showing flutes with stringed instruments, since there was a close relationship between flutes and lyres at the time of the development of the musical texts.

There are many challenges to interpreting ancient visual art in the context of music. The evidence is often fragmentary (literally, involving broken pieces of clay tablets), the accuracy of the image is often questionable, and the intent of the author is open to conjecture. Another challenge is that many aspects of music relate closely to fundamental mathematical ratios, making it difficult to distinguish the intent of a written text: does it refer to music, or is it just the fundamental mathematics behind pleasing art? For one such example, see Venus of Lespugue.

However, similar depictions of musical instruments, at different times and from different locations, can give us a sense of what was prevelant at the times. These examples are in approximate chronological order, and include some examples from Egyptian cultures, for context and comparison.

Egyptian Two-Dog Palette, 3200-3050 BCE

One of the earliest representations in visual art comes from the Two-Dog Pallate of Ancient Egypt.

Detail of the reverse side of the Two-Dog Palette

Detail of the reverse side of the Two-Dog Palette More information

See the description of the Two-Dog Pallate as well as a detailed view of the image.

Woman Playing Flute, 2600-2500 BCE

A Sumerian culture ornament made of shell, from Nippur, about 2600-2500 BCE, showing a woman playing flute. This Artifact is in the collection of the Metropolitan Museum of Art, accession number 62.70.46, funded by the Rogers Fun, 1962. The ornament is 2.5″ × 1.25″ (6.3 × 3.1 cm)

Woman playing flute, 2600-2500 BCE, Metropolitan Museum of Art

Woman playing flute,
2600-2500 BCE,
Metropolitan Museum of Art Larger image

BM 102417 Cylinder Seal, 2400-2200 BCE

BM 102417 Cylinder Seal, 2400-2200 BCE Larger image

Cylinder Seal, 2400-2200 BCE

Cylinder seals are stone, glass, or ceramic cylinders engraved designed to roll an impression of a scene onto a two-dimensional surface, typically wet clay. They were invented in about 3500 BCE and early examples have been found in Susa, present day Southwestern Iran, and Uruk in Southern Mesopotamia.

The image at the right is a mondern impression from a cylinder seal at the British Museum. It is extremely worn, presumably from use, but does show an early image of a flute player.

  • BM 102417
  • AN420479001
  • Registration number: 1906,1110.7

Located at The British Museum. Description from their web site:

Date: 2400-2200 BCE

Dimensions: 3.2 cm high × 2.2 cm diameter (1.44″ × 0.94″)

Brown calcite (limestone) cylinder seal; a figure wearing a skirt holds a cup in his right hand and a staff over his left shoulder; he is seated in front of a reed building, facing left. His feet rest on a crouchant lion. before him is a rampant lion and above him is a crescent moon. In the field in front of him are various scenes; from left to right and from bottom to top these are as follows: a kneeling figure disposing five circles (cheeses?) on a mat before him, a flute-player, an eagle (?), a lion pacing towards the left; two pots on a stand, a seated figure holding a large tilted pot (probably a churn); a file of animals facing right and consisting of two sheep and one markhor-type goat which is long-haired and has a bird perched on its back; seal is extremely worn.

From [Rimmer 1969], page 19:

The long vertical flute played by a seated figure on a stone cylinder seal of the Akkadian period (BM 102417) is one of the rare depictions of this kind of instrument from ancient Western Asia. Rim-blown vertical flutes are shown in Egyptian wall-paintings of the Old and Middle Kingdoms [2686-1690 BCE], and this type of flute survives in north Africa, the Middle East and the Balkans. It may well have been common in ancient western Asia, though perhaps of too humble association to be depicted often.

Nubian Double Flute Player, Egypt, 1700-1400 BCE

This is a rendition from the incredible 1832-1840 publication, I Monumenti dell'Egitto e della Nubia ([Rosellini 1840]), showing a fresco from the Beni Hassan tombs. As described in [Southgate 1891] On a Pair of Ancient Egyptian Double-Flutes, page 19:

In addition to the double-flute player, there are three girls marking the rhythm of the dance by clapping their hands … about an inch before the inverted /\-shape tubes enter the mouth, the brown pigment with which they are painted stops, and the rest of the tubes are white. Evidently this represents the two short straw reeds with which they were supplied, and furnishes a certain proof of how they were played.

You will notice that in playing the hands are crossed, the right hand dealing with the pipe on the left side of the performer, and the left hand with the pipe on the right. This peculiarity is found in several other instances of the frescoes in the Egyptian tombs.

Bas-relief from Karatepe, circa 850 BCE, showing lyre and double flute players

Flute player from the Beni Hassan tomb fresco,
1900-1700 BCE, from [Rosellini 1840] Larger image

Flute and Lyre scene from the tomb of Djeserkara Amenhotep I

Flute and Lyre scene from the
tomb of Djeserkara Amenhotep I More information

Egyptian Double Flute, 1490 BCE

Many Egyptian musical scenes exist, often showing double-flutes. From [Bromiley 1986]: “Depiction of end-blown flutes, single and double clarinets, harps, and trumpets appear in tombs of the Old Kingdom (ca. 2686-2040 BCE), alongside singers and, often, dancers. By the New Kingdom (ca. 1558-1085 BCE) soloists and ensembles illustrate a variety of new instruments, including the lute (a Mesopotamian import), several types of lyres and harps, the double oboe, and various sorts of drums, tambourines, clappers, and sistra.

The painting on the right is an excerpt from a scene on the wall of the tomb of Djeserkara Amenhotep I (1515 - 1494 BCE), from [Badawi 1960], which would date to about 1490 BCE.

Keratepe, 850 BCE

A bas-relief from Karatepe, contained in the present-day Karatepe-Arslantaş National Park in South-central Turkey. It is part of a two scenes showing musicians at a banquet and is dated to circa 850 BCE (originally from [Cambel 2003] plates 50, 51, 144, and 145, taken here from [Collon 2008], figure 13a).

Bas-relief from Karatepe, circa 850 BCE, showing lyre and double flute players

Bas-relief from Karatepe, circa 850 BCE,
showing lyre and double flute players,
from [Collon 2008], page 56 Larger image

BM 118179 Ivory Box 900-700 BCE

BM 118179 Ivory Box 900-700 BCE Larger image

Ivory Box, 900-700 BCE

  • BM 118179
  • AN105331001
  • Registration number: N.973

Located at The British Museum. Description from their web site:

Date: 900-700 BCE

Dimensions: 6.7 cm high × 9.5 cm diameter (2.6″ × 3.7″)

Carved ivory, possibly made in Phoenicia. Excavated at Burnt Palace, Nimrud (Kalhu), present-day North Iraq by Sir Austen Henry Layard. Badly burnt. Acquired in 1856.

Circular ivory box or pyxis: with a continuous frieze of carved decoration. It is badly burnt, but shows musicians playing double pipes, zithers and the tambourine. They stand amidst palm and lotus trees and behind a goddess seated on a throne. In front of the goddess is a cross-legged table piled with food delicacies. Behind this are two ladies, one of whom is clearly an attendant. On the underside is an inscription written in West Semitic script. This is a fragment repaired with others of same Big Number (BM number) to form whole.

Nineveh, 645 BCE

A bas-relief carved circa 645 BCE in Room 33 of Sennacherib’s South-West Palace at Nineveh. It shows a procession led by musicians playing vertical harps of various sizes and a double flute player (from [Collon 2008], figure 20a. This also appears on the cover of [Kilmer 1976]).

Bas-relief from Karatepe, circa 850 BCE, showing lyre and double flute players

Bas-relief from Karatepe, circa 850 BCE,
showing lyre and double flute players,
from [Collon 2008], page 56 Larger image

North Palace of Ashurbanipal, Nineveh, 645 BCE

Detail from the upper-right corner of a large series of bas-reliefs carved about 645 BCE. It is from the North Palace of Ashurbanipal at Nineveh and depicts an orchestra playing in the garden of Ashurbanipal's harem. It was originally published in [Barnett-RD 1976], plate 63f and also shownas plate 2 in [Dumbrill 2005], identified as BM 124922 59, from the sixth-seventh BCE.

North Palace at Nineveh, circa 645 BCE

North Palace at Nineveh, circa 645 BCE,
showing lyre and double flute players,
from [Collon 2008], figure 21c Larger image

This image is from [Collon 2008], figure 21c, and is described there:

The musicians are women and beardless men (as this is the harem, they must be eunuchs), and they play a total of six vertical harps, one lyre, one lute, one long drum, three short double pipes and six long double pipes (the long double pipes are all held by beardless men). A fragment in Berlin shows a female flautist.

Limestone Flute Player, Cyprus, 575-550 BCE

A limestone sculpture excavated in Cyprus, from about 575-550 BCE, showing a woman playing flute. This artifact is in the collection of the Metropolitan Museum of Art, accession number 74.51.2517, funded by The Cesnola Collection, Purchased by subscription, 1874–76. The statue is 15.75″ × 13.75″ × 8.75″ (40 × 34.9 × 22.2 cm)

Limestone Flute Player, Cyprus, 575-550 BCE, Metropolitan Museum of Art

Limestone Flute Player,
Cyprus, 575-550 BCE,
Metropolitan Museum of Art Larger image

The Artifacts

This section looks at musical artifacts, both of lyres and flutes.

The Great Lyre of Ur, University of Pennsylvania Museum

The Great Lyre of Ur,
University of Pennsylvania Museum Larger image

Stringed Instruments

C. Leonard Woolley with the plaster cast of the Sumerian Queen's Lyre, 1922

C. Leonard Woolley with the plaster cast
of the Sumerian Queen's Lyre, 1922 Larger image

In 1929, C. Leonard Woolley led an excavation that recovered four lyres and a harp from Ur in Ancient Mesopotamia, present day Iraq ([Woolley 1934] Ur Excavations, Volume 2: The Royal Cemetery - Text and Plates). They are the oldest existing string instruments, dating to about 2750 BCE.

The instrument shown at the right, from [deSchauensee 2002], is the restored Great Lyre of Ur. It was recovered from the grave site of King "Lugal" Abargi, dated about 2685 BCE. The grave was ceremonially guarded by six soldiers wearing copper helmets and carrying spears. A dozen men armed with their weapons laid close to the bodies of richly adorned women, supposedly singers and a harpist. Close to their heads the remnants of two musical instruments were found. They may be associated with the ceremonial burial of the king.

The restored instrument is on display at the University of Pennsylvania museum. The head of the bull is covered with gold leaf and the beard and eyes are fashioned from lapis lazuli. The lyre’s wooden structure has been reconstructed from the detailed measurement made by pouring the plaster into the impression left by the disintegrated wood, similar to the photo on the left of the plaster cast done by Woolley for the Queen's Lyre, now at the British Museum in London. The eleven strings fastened on the rectangular sound-box are modern. The front of the sound-box is decorated with the mosaic plaque, trapezoidal in shape and set in bitumen. In one of four scenes, depicting mythological creatures, a seated animal – onager or bear – plays a similar lyre.

A Catalog of Flutes

A comprehensive list of early instrument artifacts from Mesopotamia, Egypt, North Syria, and Souteast Asia Minor was assembled in [Mitchell 1992], The flutes that were cataloged were classified in two types:

"Vertical Flutes (end blown like the modern recorder)": 4 from Mesopotamia — 3 from 2330-1600 BCE and 1 from 900-539 BCE.

"Double-pipe": 6 from Mesopotamia — 2 each from the period circa 2500 BCE, 2330-1600 BCE, and 900-539 BCE; 6 from Syria from 900-500 BCE; 2 from Cyprus from 900-500 BCE

The Silver Pipes of Ur

In one private grave, “PG 333”, they found “bars of silver wantonly twisted and bent” that was later discovered to be a silver double-flute, described in [Krispijn 2008]:

Silver flute from tomb PG 333. Ur ± 2500 BC. Beside the stringed instruments flutes were excavated from the royal tombs. The instrument might have had a reed mouthpiece. It certainly covers a diatonic scale, possibly from C-D-E-F-G-A.

[Krispijn 2008] provides two photos: the recovered silver fragments from the original excavated and a reconstruction of the double-flute:

Excavated fragments of the Silver Double-Flute of Ur

Excavated fragments of the Silver Double-Flute of Ur More information


Reconstructed Silver Double-Flute of Ur

Reconstructed Silver Double-Flute of Ur More information

The original excavation by by Sir Leonard Woolley in the early 1930s are described in detail in [Woolley 1934] Ur Excavations, Volume 2: The Royal Cemetery - Text and Plates, pages 258–259:

In PG/333 there were found what seemed to be bars of silver wantonly twisted and bent. These were scientifically cleaned in the University Museum and proved to be of great interest. The apparently meaningless mass consists of silver tubing, with a total length of 0.408 m. (16.06″); it is broken into five pieces, but may originally have consisted of two parts each of an approximate length of 0.260 m (10.24″). Along one side of each there are five (?) holes 0.006m. (0.24″) in diameter placed at intervals of 0.025 m. (0.98″); the last hole comes at 0.025 m. (0.98″) from the end of the tube, and the first at 0.14m. (5.51″) from the unbroken end which may be the mouthpiece. At 0.07 m. (2.76″) from the (complete) end of one tube there is a double incised band, and a similar band on the second tube close to its broken end. U. 8605.

There can be no question but that we have here the remains of one of the double pipes figured on Sumerian carvings, e.g. to take a late instance, on the great stela of Ur-Nammu; the slenderness of the pipe suggests that it is directly inspired by its original, the reed of the marshes; the intervals may help to throw light on Sumerian music as a whole.

The Silver Pipes of Ur are the oldest existing wind instruments from the Near East region, more than 500 years older than the oldest surviving Egyptian instrument, a set of Middle Kingdom flutes made of reed ([Lawergren 2000], page 123). See the reconstruction of his instrument below by Bo Lawergren.

The Babylonian Ocarina

The Babylonian Ocarina, from [Engel 1864] Larger image

The Babylonian Ocarina

The small ocarina with two finger holes was execavated at Birs Nimrûd, near Babylon, in about 1860. From this it appears likely that the Sumerians knew the principle of crafting a whilstle mechanism.

The image at the right is from [Engel 1864], page 76, which is the only surviving image of the artifact, which has since been lost. The original image was vertical, with the mouthpiece at the bottom. I believe that the upper finger hole shown in my diagram is the “left” finger hole in this description from [Engel 1864]:

The instrument next deserving of notice is a little pipe of baked clay which was found by Captain Willock in the ruins of Babylon, Birs-i-Nimroud, and which has been presented by him to the Museum of the Royal Asiatic Society. It is about three inches in length, and has only two finger-holes, situated side by side, and consequently equidistant from the end at which it is blown. The opposite end has no opening: the instrument in this respect resembles a whistle. If both finger-holes are closed, it produces the note C; if only one of them is closed, it produces E; and if both are open, it produces G.

Besides these notes, one or two others are obtainable by some little contrivance: thus, by blowing with unusual force, the interval of a fifth, G, may be raised to that of a sixth, A. But the fixed and natural notes of the instrument are only the tonic, third, and fifth. Moreover it is remarkable that the third which is obtained by closing the left finger-hole is about a quarter-tone lower than the third which is obtained by closing the right finger-hole. Perhaps it was intended for the minor third. It may have been originally more flat, and might perhaps be restored to its former pitch, if it were advisable to submit the pipe to a thorough cleaning.

The accompanying engraving exhibits the instrument full size. That it is a genuine Babylonian relic admits, in my opinion, of no doubt. It resembles, in material and workmanship, several other articles known to be of Assyrian manufacture; and several little idols have been found embedded with it, which are similar to those obtained from the Assyrian mounds.

This is, as far as I am aware, the oldest musical instrument hitherto discovered which has preserved its original condition; yet it is constructed of so fragile a material that were it to fall from the hand to the ground it would most likely be destroyed for ever. But its notes cannot have been clearer two thousand years ago than they are at the present day. They constitute the intervals of the common chord, either major or minor. No doubt the feeling for musical concord is innate in man, like the feeling for melody. It probably caused the Babylonians to adopt for their little wind-instrument those intervals which together constitute the harmonious Triad, and which, even when heard in succession (arpeggio), produce an effect similar to that most consonant chord.

The shape of this instrument appears to be intended to represent the head of an animal. It is singular that the little flageolets and whistles of the ancient American Indians, of which many have been found in tombs, especially in Mexico and in Central America, arc also of pottery formed to represent animals, and bear besides, in other respects, much resemblance to the Babylonian pipe.

Replicas of Artifacts

How were ancient flutes tuned? Were they tuned precisely? To a particular relative scale? To body measurements? To be played with other instruments?

One approach to answering these questions is to construct replicas of the artifacts. In many cases, especially for flutes with a vibrating reed for a sound mechanism, we do not have the complete instrument. So, the size and shape of the reed sound mechanim is crafted from present-day examples of similar instrument, depictions in art, or our imagination. And, even if you have a complete artifact, they are notoroiusly difficult to replicate accurately, as noted by Dayton C. Miller ([Miller-DC 1935], page 100):

One cannot even approximately calculate the length of a flute tube which will sound a given note [musical tone]; one cannot by theoretical calculation locate any finger hole on a flute tube which will produce a given tone.

However, the field of Archaeomusicology, which applies methods of archaeology to the study of music ([Hickmann-E 2001]), draws from more than the instrument artifacts. It combines excavated visual art, literary texts (and their interpretation based on our understanding of the source languages), and comparisons between cultures to help place our understanding of music in a historical context.

The Southgate Egyptian Replicas

In 1890, Thomas Lea Southgate did extensive analysis on an Egyptian flute excavated from a tomb. He created replica instruments and measuring their pitches. See the reference [Southgate 1891] On a Pair of Ancient Egyptian Double-Flutes for a complete analysis, including a reconstruction of the flute with a measurement of the pitches.

The Silver Pipes of Ur Replica

Many researchers have attempted to reconstruct the Silver Pipes of Ur over the years ([Galpin 1937], [Sachs 1943], page 73, [Rimmer 1969], page 36, and [Collinson 1975]), but were hampered by various issues of identification of the segments and the confusion surrounding subsequent pipes and segments found at later dates.

Bo Lawergren did an exhaustive survey in 1998-2000 of the past work and examined the artifacts first-hand, as well the various historical photographs ([Lawergren 2000]). His analysis produced this reconstruction, redrawn here from [Lawergren 2000], figure 1f:

Reconstruction of the Silver Pipes of Ur, by Bo Lawergren, 2000

Reconstruction of the Silver Pipes of Ur, redrawn from [Lawergren 2000] Larger image


We have already looked at references to flutes in the great literary work The Epic of Gilgamesh. There are a few other litererary texts that shed light on flutes in Ancient Mesopotamia.

Tablet K 162

Tablet K 162 Larger image

The Descent of Ishtar

A passage in a separate Babylonian literary work, known as The Descent of Ishtar into the Netherworld, mentions a flute of another semi-precious stone, lapis lazuli, in two locations.

The original tablet is K 162 in the Kouyunjik Collection of the British Museum, shown at the right ([DEAA 1902], plates 45-47, [Jastrow 1915], pages 453–461, [Horne-CF 1917] pages 235-241, and [Heidel 1963], pages 119–128). It is from 600–700 BCE.

The myth descends from an earlier Sumerian myth about Inanna, the Sumerian counterpart of Ishtar. According to the Sumerian myth of Inanna, she can only return from the underworld if she sends someone back in her place. Demons go with her to make sure she sends someone back. However, each time Inanna runs into someone, she finds him to be a friend and lets him go free. When she finally reaches her home, she finds her husband Dumuzi (Babylonian Tammuz, [Wiggermann 2010]) seated on his throne, not mourning her at all. In anger, Inanna has the demons take Dumuzi back to the underworld as her replacement. Dumuzi's sister Geshtinanna is grief-stricken and volunteers to spend half the year in the underworld, during which time Dumuzi can go free ([Wolkstein 1983], pages 52–89).

The two references to flutes are on lines 49 and 56 of the reverse of the tablet, translated here from [Heidel 1963], pages 127 and 128:

49:   Clothe him with a red garment (and) let him play upon the flute of lapis lazuli [....].

56:   On the day that Tammuz greets me with jubilation(?), that with him the flute of lapis lazuli (and) the ring of carnelian greet me with jubliation(?)

Heidel suggests the possibility that “the ring of carnelian” could be a tambourine inlaid with carelian. The transliteration of line 56 is provided in [Vanderburgh 1908], in footnote 1 on page 19:

56:   ina ȗ-me ilu Dumu-zi el-la-an-ni malil abnu ukni abnu it-ti-šu el-la-an-ni

The word used here for flute is ma-li-l, different from the em-bū-bu used in Gilgamesh. See the explanaition in The Pious Sufferer, below.

The Pious Sufferer

The epic poem The Pious Sufferer mentions a flute ([Jastrow 1915], pages 474–483). In a passage about the healing of Tabi-utul-Enlil during a great storm (translation from page 483):

The tongue which was swollen so that I could not move it,
He took away its coating (?) so that speech returned.
The throat which was compressed, closed up like that of a corpse,
He healed so that my breast resounded like a flute.

In this version, the Akkadian word ma-li-liš, is used for “flute”. As analyzed by [Jastrow 1915], page 78, footnote 199:

ma-li-liš explained in the commentary ma-li-lu = im-bu-bu, apparently the more common word for “flute,” but also introduced here to avoid a misinterpretation of ma-li-lišlike a flute,” since the malilu is ordinarily the instrument of “lamentation,” and not of rejoicing, as the context here requires.

So the Akkadian word for flute is either em-bū-bu or ma-li-lu, with different connotations and shades of meaning for each.

Shulgi Hymn B

This hymn is dedicated to King Šulgi (in English: Shulgi) of Ur (about 2029-1982 BCE). It boasts about the King's musical abilities, and also gives us insights into the music of the times. The Hymns of Šulgi generally speak about his achievements in the area of social behaviour and religion. He is depicted as making social justice, law and equity central to his reign.

Here is the translation by Theo Krispijn from [Krispijn 2008]:

154:  I, Shulgi, king of Ur,
         also have engaged in music.
         Nothing was too complicated for me.
         I learned the depth and the range of the tigi- and adab-compositions, the perfect musc.
         When I fasten the tuning pegs on the shukarak-harp/lute?, which always stays in the memory, I do not break its arm.
160:   I have established the plan of tuning it (the tuning pegs) up and down.
         On the instrument with the eleven string bows, the zamin lyre, I know the fine tuning.
         The lute with three strings and ‘sound box of music’ I know how to pluck.
         I pluck the lyre from Mari, that makes the house speechless.
165:   I know the fingering technique of the horizontal algar harp and the harp from Sabum, of royal quality.
         The lyre of Urzababa, the harhar lyre and the Anatolian lyre, the lion-lute and the lute ‘Stake of the Captain’ I caused to make a noise like fire.
         Like a skilful musician, although I had never heard its sound, whenever someone brought it to me, I know how it works.
170:   Like something that I had held in my hands before, I can handle it.
         When I am tightening, loosening or rastening (the strings) for tuning, they do not slip out of my hand.
         The double oboe I never cause to sound like a shepherd's flute,
         but I know how to play a sumunša song like its permanent accompanist.

Another translation of lines 170-174, from The Electronic Text Corpus of Sumerian Literature (ETCSL) reads (as of January 3, 2012, with translations and various editorial changes between August 10, 1998 and June 1, 2003 by JAB (Jeremy A. Black), GZ (Gábor Zólyomi), GC (Graham Cunningham), ER (Eleanor Robson)):

170:   I am able to handle it just like something that has been in my hands before.
         Tuning, stringing, unstringing and fastening are not beyond my skills.
         I do not make the reed pipe sound like a rustic pipe,
         and on my own initiative I can wail a šumunša or make a lament as well as anyone who does it regularly.

… along with a transliteration of those lines from the same source and authorship:

170:   niĝ2 ud-bi-še3 šu-ĝa2 ĝal2-la-gin7 ba-e-de3-ĝa2-ĝa2-de3-[X]
171:   ad pad3-de3 gid2-i tu-lu gen6-na šu-ĝa2 la-ba-ra-e2
172:   gi-di gi sipad-gin7 nu-um-me
173:   šumun-ša4 mur ša4-e i-si-iš ĝa2-ĝa2
174:   šaĝ4 gen6 saĝ us2-bi-gin7 i3-zu

So we see that a specific distinction has been made between the sound of two distinct instruments:

  • the gi-di (“reed pipe” in ETCSL and “double oboe” in [Krispijn 2008])
  • the gi sipad-gin7 (literally “reed” (of the) “shepherd”, translated as “rustic pipe” in ETCSL and “shepherd's flute” in [Krispijn 2008])

Types of Flutes and their Names

You'll notice that in the descriptions above, it is often unclear exactly what kind of flute is being described. Terms such as “reed pipe ”, “double pipe ”, “double reed ”, “rustic pipe ”, “vertical flute”, “double flute ”, “double oboe”, “short double pipes”, and “long double pipes” are used, but there is a general lack of consistency to how the terms were applied, and it is often unclear what kind of instrument they were describing. This does little to help with the question: What kind of flute was the embūbum in the Epic of Gilgamesh?

This section looks at the various type of flutes, going back to the source documents where possible for clues as to how the flutes were constructed.


In general, this web site has a specific set of terms when classifying flutes. In particular, the term “flute” on this web site is used for any instrument where the player's breath is directed either by the player or by the instrument against a splitting edge that causes the air to vibrate (class HS 421 Edge-Blown Aerophones). This differs from some other contexts that restrict a “flute” to embouchure flutes where the player's lips direct a stream of air to the splitting edge.

One class of instruments that is not a flute is anything that uses one or two “lamella” — vibrating membranes often called “reeds” — which periodically interrupt the airflow and cause the air to be set in motion. These are in class HS 422 - Reed Aerophones - instruments such as the oboe, clarinet, Armenian duduk, Great Highland bagpipes, and the Chinese bawu. And this is a general source of confusion, since the body of many of the wind instruments of Ancient Mesopotamia were made of bamboo or river cane reed.

Martin West describes the difficulty of naming this class of instruments, referring to the Greek instrument called aulos, which is made of reed, and has vibrating lamella, and are often played in pairs. From [West 1994a], page 82:

What, then, should it be called? ‘Pipe’ is unobjectionable but uninformative. ‘Double reed pipe’ may be thought better, but the phrase is fraught with ambiguity and may mislead the innocent. It is not automatically clear whether ‘reed pipe’ means a pipe made from a hollow reed or a pipe equipped with a vibrating reed; and incidentally, this so-called ‘reed’ in the mouthpiece is not necessarily made of reed. Again, it is not clear, unless a hyphen is inserted, whether ‘double’ applies to the reed or to the pipe. For reed mouthpieces fall in two categories, generally known as single reed and double reed. In the single-reed type, a cut is made down the side of a cane or straw (the top end of which is stopped), so as to create a narrow blade or tongue which, when excided by blowing, will beat against the opening that it covers. In the other type, the mouthpiece terminates in two thin blades that beat against each others; they can be, and in antiquity were, made simply by flattening the end of a hollow stem. In both types of mouthpiece the beating part is completely enclosed by the player's mouth.

Type of Flutes in Ancient Mesopotamia

A comprehensive survey of Ancient Mesopotamian instruments, The Music of the Sumerians, Babylonians and Assyrians, was done by Francis W. Galpin, the British organologist (one who studies the history, construction, development and use of musical instruments). This summary is quoted from chapter 2 of his work ([Galpin 2011]), augmented as noted. This description is based on and supported by many of the artifacts and literary references we have seen above.

(a) Flutes

The flute par excellence of these early ages is the vertically-held simple reed-tube, sounded by blowing across one of the open ends. In Mesopotamia it was generally called ti-gi (Akkadian tigu, tegu) or sometimes ka-gi (literally “mouth-reed” in English) and was highly esteemed in the temple ritual.

It had three finger holes, equidistant from one another, and, by the help of the first and second registers of the harmonic series, could produce a diatonic scale of seven notes with a sharp or tritone fourth: this scale could be extended into the second octave.

The flute ti-gi is therefore called also the imin-e (literally “the seven-note” in English). In a poem in praise of the Temple of Enki at Eridu (c. 2200 BCE) we read “the musician on the seven-note brings forth a plaintive sound”, and agian, “into the hallowed fourcourt let the musician duly bring the drum and seven-note” … Other descriptive names are also employed, such as “the long-reed or instrument” (gi-gid, gi-bu, giš-sír) or “the large reed ” (gi-dim) as contrasted with the much shorter reed-pipe.

The use of tīgῡ and gi-gid are supported by [Sachs 1940], page 72.

Ashmolean Prism

Ashmolean Prism Larger image

The use of the term imin-e relating to flutes may be significant to our understanding of these instruments. In analyzing a copy one artifact in the collection of a dealer in Paris that is a copy of the Ashmolean Prism (Ashmolean Museum, AN1923:444, shown at the right), Stephen Langdon translated the Sumerian Liturgy of the Cult of Kéš ([Langdon 1923], pages 48–59). In footnote 9 of page 56, Langdon give the possibilities of translation for nar-balag-iṃin as “the seven flutes” or “the flute of seven notes(?)”. This is further discussed in [Langdon 1923a], but I have yet to locate this paper.

Galpin goes on to discuss the Babylonian Ocarina described above, and offers a possible name for the instrument:

As to its actual name, we have just this possible reference on a tablet of the Greek Period (c. 300 BCE), now preserved in the Ashmolean Museum, Oxford, containing a bi-lingual Ishtar epic. According to Mr. Sidney Smith, the God of Heaven, Anu, is addressing Ishtar in her capacity as a goddess of battle and says: “Let the arrow pass through, piercing heart and entrails like an uku”. The cuneiform ideograms employed for uku mean “to pile on a flute”, and Mr. Smith adds “may we translate — like a blast (on a pipe)?” But the ideograms seem to imply also a sound “striking” or “overpowering” and “high in pitch”. Only a short instrument, of the whistle type, could give these effects. It is worth while noting that the Bantu word uku is used for “blowing a wind instrument” and that the sound of the whistle blown by the African tribes in war and incantations is believed by them “to penetrate the heart (of the evil-doer), who is frightened and will sink and die”. Here we appear to have the counterrpart of Anu's appeal against the foe.

Moving on to class HS 422 - Reed Aerophones instruments:

(b) Reed Pipes

The earlier instruments included under this section may have been of the single-beating reed or clarinet type … The difference between this instrumetn and the ti-gi or longer vertical flute is distinctly shown by the angle at which the instrument is held: this is necessary for the insertion of the vibrating tongue of the reed within the mouth. … The name among the Sumerians appears to have been (in Akkadian nabu)

In the Hymn to Ishtar, quoted under the previous section, its pleasant sound is linked with that of the flute. But its plaintive character appears to have pre-dominated and we find the names gi-er-ra or gi-ir-ra (nabu, “to lament”; nababu, qan bikiti, “the wailing reed”) given to it. … It is recognizable too under the name giš-gù-sìr or “the instrument with the crying voice”. In a Sumerian penitential hymn the suppliant says “like the reed , I am in sadness”, and in a lamentation by Ishtar for Tammuz she wails “the reed of lamentation (gi-er-ra) is in my heart”.

We find that a special class of ceremonial hymns was known as er-šem-ma or pipe songs, for, like the single-pipe, the šem was played in penitential processions.

Relating to an instrument that is prevelant in Southern China, the hulusi:

A third interesting type is the kitmu, on which the reed was probably “covered” (Akkadian katāmu) by a cap of rush, gourd, wood or horn, an improvement on the earlier practice of placing it inside the player's mouth. It is mentioned in a list of late Semitic songs on a tablet found at Asshur and forms the accompaniment to seventeen love-songs to be sung by women.

Another pipe, fitted probably with a single-beating reed, was called pîtu. The Sumerian ideogram, DUN, implies something “bored” or “hollowed out”, but with it we find associated a word kippatu, which means “curved”. The pîtu therefore was probably of the Phrygian type with a semicircular up-turned end or bell.

And, finally, Galpin describes the flute as named in the the Epic of Gilgamesh:

The last reed instrument of which we must speak is the imbubu, certainly a double-beating reed pipe, tapering in shape and played either singly or in pairs. It was also called malilu, which corresponds to the Latin tibia (a leg bone) and may have been used loosely in the generic sense. The pipe no doubt was introduced from Syria where its name was Abuba (Arab imbub) and its female performers were known in Rome as the Ambubajae.

Karatepe Relief

Karatepe Relief Larger image

The variant of the word imbubu used by Galpin for the word embūbu was noted above in the translation of The Pious Sufferer , based on [Jastrow 1915], page 78, footnote 199.

In the case of a single musician playing a pair of woodwinds, the first issue is how they were tuned to each other. [Krispijn 2008], page 9 and slide 37 makes a conjecture (citing the bas relief shown at the right, from his slide 37):

A relief from Karatepe (end of the 8th century BCE) shows an offering scene above, and below it shows an orchestra, with hand drums, two lyres and a double oboe. It is very plausible that the two tubes of the double oboe (gi.di.d = embūbusounding reed”) differed by one-fifth, since in musical theory the embūbu, “sounding reed” interval of one fifth is attested.

The word “Embūbum”

Here are some other sources for the for the use of the term embūbu in the context of the instrument (as opposed to the tuning of scales, that we will see in the musical texts, below).

The Chicago Assyrian Dictionary ([CAD 2011] The Chicago Assyrian Dictionary (CAD), 21 Volumes (26 total parts)), sixth printing (2004) of Volume 4 (initially published in 1958) on page 137 defines the Akkadian term embūbu (variants enbūbu, ebbūbu) as “flute”. This definition is echoed and supported by:

  • [Kilmer 1965], page 267 (shown as “embubu”),
  • [Dumbrill 2005], pages 406–407 (listing an additional variant enbunu),
  • [Brown-JP 2001], page 375,
  • [Bromiley 2007], page 443, which equates the Akkadian ebbūbu and embūbu to the Sumerian GI.GÍD (“long reed”) and GI.DI.DA (either “single reed” or “reed to be played”).

Also, from [Kilmer 1971], page 137:

Among the names that were read with certainty from the outset was the Akkadian word embūbu, well known as the name of a flute, and used to describe the interval of a fifth extending from the third string to the seventh. In a rather vigorous effort to find more passages using the same interval names, I failed albiet with the help of the exhaustive files of Chicago's Oriental Institute's Assyrian Dictionary project, to locate a single one. I had, of course, in consulting the file cards, repeatedly come across the word embūbu, listed as the name of a flute in numerous contexts, together with other named instruments; but the lists of instruments were not fruitful for my purposes. … The composition appears to have been written in honor of the patron of music, the god Ea.

[Yamauchi 1980], page 13, notes that [Ellermeier 1970]has made a detailed study of the distribution of the Near Eastern double flute (or oboe), which has been found in Mesopotamia, Syria, Palestine, Egypt, Cyprus, Crete, and Greece. He notes that the Syrian embubu passed into Latin as ambubaiae, a word which designated both the instrument and the Syrian girls who played them.

And, in some other languages:

  • In Hurrian, another ancient Semitic language: um/em-būb-e ([Dumbrill 2005], page 127).
  • The Eblaite form of the word, another ancient Semitic language spoken in the third millennium BCE in the ancient city of Ebla, at Tell Mardikh, in present-day Western Syria: na-bu-bù-um ([Franklin 2009], page 44, footnote 191, which notes that the Akkadian and Eblaitic forms should descend from a common root).
  • In Ugaritic: ḥalil ([Franklin 2009], page 87, footnote 408)
  • In Latin: ambubaiae ([Ellermeier 1970], as noted by [Yamauchi 1980], page 13).
  • In modern Eastern Assyrian: abbueba or shabeba (AINA web site).
  • In modern Western Assyrian: abubo (Christians of Iraq web site).

The Musical Texts

Of the half a million recovered tablets and fragments, we have found seven texts written in Akkadian that relate to music. This may seem meager, but consider that there are we have found no writings from ancient Egypt that discuss music ([Dumbrill 2005], page 23). Also, consider these thoughts on the status of music in the culture and what would cause them to use the cumbersome process of inscribing into wet clay. From [Michalowski 2010]:

In early Mesopotamia, musical knowledge was learned by apprenticeship, not in the establishements that taught reading and writing, and the “music texts” were never part of the standard school curriculum. These were not practical instructions, for which writing was superfluous, but rather theoretical exemplifications of certain limited lexicographical and mathematical knowledge associatesd with strings of instruments. Therefore, I suggest that the narrow scholastic tradition that is represented by these “music texts” is a marginal one that is associated primarily with mathematical and lexical scribal practice rather than with professional musical knowledge and performance, although its origins may lie in Akkadian language liturgical contexts. The terminology, as well as the setting in which it was used, has nothing to do with ‘Sumerian’ music; indeed, it might be better to avoid such ethnic distinctions altogether and to speak of a variety of common as well as regional musical forms and traditions, that commingled in early Mesopotamia.

Despite these limitations, it is exactly the “narrow scholastic tradition” that is so interesting with these music texts … since they relate to tuning and scales.

Here are the central musical texts, in the order of the age of information that they contain (some physical texts are copies of older texts). For a key to the tablet designations, see Designations for Cuneiform and Ancient Mesopotamian Clay Tablets. General refereces for the musical texts are [West 1994] and [Dumbrill 2005].

MS 2340 - Early Sumerian Music Tablet

MS 2340 - Early Sumerian Music Tablet More information

MS 2340 - The Earliest Sumerian Musical Text

The description of MS 2340 on the Schøyen Collection web site provides general information and background (the source of the image at the right and from which the description below is derived), but there seems to be, as yet, no detailed analysis of the text by Archeomusicologists.

A 26th century BCE tablet written in Sumerian, this tablet contains the earliest known record of music and musical instruments in history. It lists 9 types of musical strings and 23 types of musical instruments and music. The name of one of the stringed instruments is a Semitic word, ki-na-ru, the later kinnaru known from the Mari letters and Ras Shamra texts (13th century BCE, cfr. MS 1955/1-6), and still later the Biblical Hebrew kinnor.

For more information, see [Civil 2010], pages 203–214.

MS 5105 - Early Musical Notation

MS 5105 - Early Musical Notation Larger image

MS 5105 - Early Musical Notation

This tablet is of unclear provenience and has recently been acquired by the Schøyen Collection as MS 5105. The range of estimated dates from various sources is 2000-1600 BCE.

The Schøyen Collection (the source of the image at the right) describes it as the oldest musical notation, written for a long-necked Babylonian four-string lute. The tablet has columns of numbers with the headings “intonation” and “incantation”. The commentary from the Schøyen Collection says that “It further attests that frets were used, and that their values, tonal and semitonal, were purposely calculated. Most significantly, the discovery of this text attests of a music syllabus in educational institutions about 4000 years ago.

This description is based on the analysis of Richard Dumbrill, who studied the text prior to it's acquisition by the Schøyen Collection. His conclusions, published in [Dumbrill 2005], pages 96–110, are that:

  • the text was devised for a lute rather than a harp,
  • the lute was fitted with four strings,
  • the strings were tuned in ascending fifths: C-G-D-A,
  • frets were used, and their values, tonal and semitonal, were purposely calculated,
  • the system of music used was intrinsically restricted to 14 degrees (tones).

Dumbrill goes as far as to propose a transcription in modern music notation:

Proposed transcription of MS 5105

One significant point is that this music was notated in terms of ascending numbers. This is a very different method than was used for harp/lyre notation, which used intervals and was based on a descending scale notation.

Dumbrill also notes that “most significantly, the discovery of this text attests of a music syllabus in educational institutions some 4000 years ago.”

See also the significance of this tablet for Seven-tone Scales below.

UET VII 126 - The String Naming Text

UET VII 126 - The String Naming Text Larger image

UET VII 126 - The String Naming Text

This tablet was inscribed in about 626-539 BCE (CDLI chronology). However, it is a copy of the 32nd tablet of the series nabnītu, based on the discovery to a duplicate fragment from Nippur, N 4782, dated to 1800-1500 BCE ([Franklin 2002a] Terpander: The Invention of Music in the Orientalizing Period, page 138).

The information this tablet contains is the oldest information that we have regarding the names of strings and pitches. The tablet is bilingual with the left column in Sumerian and a translation into Akkadian in the right column ([Dumbrill 2008d]). It has recently been sent to the Iraq Museum in Baghdad. A detailed analysis was done by Richard Dumbrill in [Dumbrill 2005], pages 27–36. The drawing at the right is from [Dumbrill 2005], a copy as originally shown in [Gurney-OR 1974].

This is a critical text, because it provides the names of nine strings of a musical instrument (presumed to be the lyre) and is a key for understanding the other musical texts ([Michalowski 2010], page 210). Most string names simply describe their position and are rather mundane. Most authors assign them string numbers for ease of discussion:

String Names from UET VII 126
Number Meaning
1 Foremost or front string
2 Next string
3 Third, thin string
4 Fourth, small string (in Sumerian); String created by Ea (in Akkadian)
5 Fifth string
6 Fourth string of the rear
7 Third string of the rear
8 Second string of the rear
9 Rear string, or First string of the rear

However, there was a great deal of confusion trying to figure out whether string 1 was the the highest or lowest tuned string.

CBS 10996 - The Interval Text

CBS 10996 - The Interval Text Larger image

CBS 10996 - The Interval Text

This text is known as “the interval text”. It was found at Nippur and believed to be from the early first millenium BCE, but possibly a copy of a far older text on the basis that the terminology is known from another tablet, UET VII 74, dated circa 1800 BCE ([Dumbrill 2005], page 37).

A detailed analysis was done by Richard Dumbrill in [Dumbrill 2005], pages 37–45. The image at the right is from the CDLI database.

The tablet names intervals in terms of the string names from UET VII 126.

Here are the intervals, with information from [West 1994] (W) and [Dumbrill 2005] (D):

Musical Intervals Named on CBS 10996
Line String numbers Musical
Akkadian Name Translation
11 1 → 5 Fifth nīš tuḫri1
nīš gabrîm
raising of the counterpart (W);
rise of the equivalent (D)
12 7 → 5 Third šērum song
13 2 → 6 Fifth išartum straight / in proper condition (feminine) (W);
normal, erect (D)
14 1 → 6 Sixth šalšatum third (feminine)
15 3 → 7 Fifth embūbum reed-pipe
16 2 → 7 Sixth rebûtum fourth (feminine)
17 4 → 1 Fourth nīd qablim casting down the middle (W);
fall of the middle (D)
18 1 → 3 Third isqum lot / portion
19 5 → 2 Fourth qablītum middle
20 2 → 4 Third titur qablītum bridge of the middle
21 6 → 3 Fourth kitmum covering / closing (W);
closing (D)
22 3 → 5 Third titur išartum bridge of the išartum
23 7 → 4 Fourth pītum opening
24 4 → 6 Third serdûm lament (D)

Notice several things about the table above:

  • The odd numbered lines use musical intervals of fourths and fifths - intervals that can be readily brought into tune by ear if you listen for the dissonant beats when the intervals are slightly out of tune.
  • There is no mention of strings 8 and 9 (this will be cleared up below).
  • The name for the perfect fifth interval for strings 3 → 7 on line 15 is embūbum.
UET VII 74 - The Tuning Text

UET VII 74 - The Tuning Text Larger image

UET VII 74 - The Tuning Text

This text is known as “the tuning text” because it provided the decisive clue to the understanding of the Babylonian musical system and its terminology. It was unearthed by Sir Leonard Woolley at Ur in the 1920s and dates to about 1800 BCE ([Dumbrill 2005], page 47). The tablet was sent to the National Museum of Iraq in Baghdad in the 1970s (see the ICOBASE entry). A cast of the tablet is in the British Museum as BM UET VII 74, shown in IC200710:15.

A detailed analysis was done by Richard Dumbrill in [Dumbrill 2005], pages 47–69. The image at the right is of a cast of the tablet at the British Museum, taken by Richard Dumbrill and appearing on the ICOBASE system.

In the early work with the text ([Gurney-OR 1968], [Kilmer 1971], [Wulstan 1971], and [Duchesne-Guillemin 1984]) there was a basic assumption that the higher pitched strings were closest to the player, as is typical on harp. This assumption was identified as incorrect by Raoul Vitale ([Vitale 1982]) and further analyzed in [West 1994]. This has provided us an interpretation of the arrangement of strings and their tuning procedure that has met with general consensus, but not complete agreement, among Assyrianologists and Archeomusicologists.

The tablet describes the tuning of the sammû instrument, using a set of instructions for tightening or loosening various strings. Each instruction has the form ([Gurney-OR 1994], page 102):

If the sammû is (tuned as) X and the (interval) Y is not clear, you tighten/loosen the string N and then Y will be clear.

A complete translation of this tablet is provided in [Dumbrill 2005], pages 49–50. It provided a breakthrough in understanding of Babylonian musical scales:

  • The tuning systems uses only intervals from the odd-numbered lines of CBS 10996, so the only intervals that needed to be tuned by ear were musical fourths and fifths.
  • UET VII 74 provides for the tuning of seven complete, related tunings, each with the name corresponding to the first interval tuned. These seven tunings appear in UET VII 74 and in other documents in a standard order: išartum, kitmum, embūbum, pītum, nīd qablim, nīš tuḫri1, and qablītum.
  • The tuning system, based on perfect fourths and fifths, is analogous to and substantially predates the Pythagorean tuning system of ancient Greece.
  • Each of the seven tunings names an “impure” interval, which corresponds to the tritone interval in the tuning and produced a dissonant chord. The tuning instructions are also designed to re-tune that impure interval using a perfect fourth or fifth to obtain another interval in the series of seven tunings.

Perhaps most significantly, each of the seven tunings has a pattern of tones and semitones (whole steps and half steps) that closely matches the seven modes of the present-day diatonic scale. One significant difference is that the Babylonian tuning procedure uses intervals of fourths and fifth that produce a Pythagorean tuning, where the Western music tradition has, for the last two centures, has “tempered” that scale so as to distribute the dissonance of the tritone throughout the tuning (i.e. “equal temperament”).

VAT 10101 - Tabulation of Songs

VAT 10101 - Tabulation of Songs Larger image

VAT 10101 - Song Catalog

This text provides, among a great deal of song-related information, a census of a collection of Akkadian love songs, classified by their tunings. The tablet is dated from about 1400-1000 BCE, as shown in the the CDLI database. It is curated by the Vorderasiatisches Museum (Near East Museum), Berlin, Germany.

This “Liederkatalog” (“song catalog”), provides this table of a count of songs in various tunings, of which four of the counts have survived. This translation is from [West 1994], page 170:

45:   23 love songs in išartu
46:   17 love songs in kitmu
47:   24 love songs in embūbu
48:   4 love songs in pītu
49:   […] love songs in nīd qabli
50:   […] love songs in nīš tuḫri1
51:   […] love songs in qablītu

A detailed analysis was done by Richard Dumbrill in [Dumbrill 2005], pages 77–82. See also [Groneberg 2003].

UET VII 74 - The Tuning Text
UET VII 74 - The Tuning Text

Hurrian Hymn Six - Front and Back - Tablets
RS 15.30, RS 15.49, and RS 17.387 from [Kilmer 1976] Larger image Larger image

The Hurrian Hymns

Beginning in 1928, a set of tablets were excavated at the site of the ancient city of Ugarit, present-day Ras Shamra, on the Mediterranean Sea near Latakia, Syria. See [Jackson-W 2004] for a background on the excavations.

Among them were 29 tablets written in the Hurrian language that were originally thought to be mathematical texts, but provided to be music notation and words for a set of Hurrian hymns. The tablets date from about 1400 BCE. Most of the notations for the hymns are incomplete, but one tablet, for what is known as Hurrian Hymn Number Six, is largely complete. This represents the oldest existing complete, notated song.

The detailed analysis of this song and how it might have sounded has been the project of intense work by Assyrianologists and Archeomusicologists for the last 50 years. I took a look at one of the proposed transcriptions in 2002 and posted a version transcribed for Native American flute at that time. However, I am now revisiting this song, and Hurrian Hymn Six will be the subject of another project that will be posted on Flutopedia in the near future.

The images at the right are from the front and back covers of [Kilmer 1976] showing the obverse (front) and reverse (back) of three tablets arranged in their original configuration: RS 15.30, RS 15.49, and RS 17.387. They are currently curated by the National Museum in Damascus.

The Octave

The absence of strings 8 and 9 in CBS 10996 is also cleared up in UET VII 74: Whenever string 1 is tuned, string 8 tuned to an octave away. Likewise, string 9 is tuned an octave away from string 2. This is described in the booklet of [Kilmer 1976], page 10:

The preserved portion of the tuning text U.7/80 — fortunately — includes evidence that supports the identification of the 8th string with the 1st. In line 10, the directions refer to the 2nd string and the 9th (“behind”) together; and in line 15 the same reference is made again in the ascending series of tunings. This reference demonstrates clearly that strings 8 and 9 are the equivalent of strings 1 and 2, and lends additional support to the belief that this equivalence involves the interval of an octave.

No word in Sumerian or Akkadian has yet been identified for “octave” ([Kilmer 2000], page 114, cited in [Crickmore 2008], page 12, last ¶). As Leon Crickmore proposes ([Crickmore 2008a], page 13):

The octave may not have been thought of as a unit in its own right, but rather by analogy like the first day of a new seven-day week.

Seven-Tone Scales

Another issue that is discussed and debated deals with whether these early musical cultures used five-tone (pentatonic) or seven-tone (heptatonic) scales. The earliest use of seven-tone scales (hepatonism, the heptatonic model, and heptachords) has been pushed back in time from the Greek traditions fostered by Pythagoras back into Babylonian times.

The tablet MS 5105 is decribed in [Dumbrill 2008c], page 14:

The text lists numbers up to 14. This is crucially important because it attests of the doubling of the heptatonic model as [1-2-3-4-5-6-7]-[8-9-10-11-12-13-14], clearly 7 + 7, excluding the octave, otherwise there would have been a 15th number, obviously, because had 1 equated to 8 then 8 would have equated to 15, absent in the present series, significantly excluding the concept of the octave. Therefore, the series of numbers expresses the double heptachord and not the double octave as many would be tempted to say. The second series of numbers: 8-9-10-11-12-13-14, would be the replication of 1-2-3-4-5-6-7, at a higher or lower register, and not as an octavial continuation of the series, as we would understand it nowadays. This is the third text attesting of heptatonism.

The Tunings

At this point, we have almost enough information to decipher how Babylonian lyres were tuned. Several issues remain:

First, we have no definitive information on whether string 1, the “foremost or front string”, had the highest pitch or the lowest pitch. A great deal depends on this, although we will see an amazing coincidence with respect to the embūbum tuning.

Second, the tuning instructions produce a relative tuning on the instrument. Nothing specifies how the tuning of the first string in a sequence is set, and there is no mention of absolute pitch. Some Archeomusicologists propose that the first string was tuned to a fixed-pitch flute or whistle ([Kilmer 2002]).

Finally, we would like to express the tunings using present-day pitches in our now-standard Equal Temperament tuning system, which is different from the Babylonian tuning system that resulted in what we know call a Pythagorean tuning system. Leon Crickmore notes this in [Crickmore 2008], page 18:

Although a modern piano is tuned to equal temperament - that is, all twelve semitones are equal in size, and their mathematical expression involves irrational numbers and a logarithmic spiral which would have been beyond the capacity of ancient Mesopotamian mathematicians - nevertheless, for the purpose of practical explanation, the use of the white keys, though approximate, is quite adequate.

So, for this discussion, we use “natural scales” as examples. A natural scale is a sequence of white keys on the piano without sharps or flats. Each tuning is demonstrated by picking an initial natural note so that the resulting tuning has only natural notes. This makes the relative pitches correct within a given tuning, but there is no correspondence (when we use natural scales) between the tunings. In practice, all the tunings on the Babylonian lyre were very close in pitch to each other, since the tuning instructions involved moving a given string by only one semitone.

The use of natural scales is sometimes known as “dynamic notation”. The use of a notation that is rooted on the note E is sometimes called “Thetic notation” ([Kilmer 1971], page 141).

Here is a table that shows the seven tunings of the nine-string Babylonian lyre using notes from natural scales, based on the musical texts. In this chart:

  • Both “string 1 lowest” ([Gurney-OR 1968], [Kilmer 1971], [Wulstan 1971], and [Duchesne-Guillemin 1984]) and “string 1 highest” ([Vitale 1982], [West 1994], and all subsequent references) versions are included, although “string 1 highest” is now accepted as being the most likely.
  • The notation C′ is an octave higher than C.
  • Scale steps show the number of semitones between each of the notes in a scale.
  • The Intervals columns show the interval relationships from the lowest string (what would be considered the root of the scale in Western musical traditions). This shows which degrees of the scale are flatted (b, also called “diminished” or “minor”), natural (called “perfect” for the fourth and fifth, called “major” for the second, third, sixth and seventh), or augmented (#).
Babylonian Tunings
Meaning2 String 1 Lowest String 1 Highest
Scale Steps Intervals Natural
Scale Steps Intervals
išartum straight / in proper condition (feminine) (W);
normal, erect (D)
E→F′ 1-2-2-2-1-2-2-1 1 b2 b3 4 5 b6 b7 8 b9 B→C′ 1-2-2-1-2-2-2-1 1 b2 b3 4 b5 b6 b7 8 b9
kitmum covering / closing (W);
closing (D)
A→B′ 2-1-2-2-1-2-2-2 1 2 b3 4 5 b6 b7 8 9 F→G′ 2-2-2-1-2-2-1-2 1 2 3 #4 5 6 7 8 9
embūbum reed-pipe D→E′ 2-1-2-2-2-1-2-2 1 2 b3 4 5 6 b7 8 9 C→D′ 2-2-1-2-2-2-1-2 1 2 3 4 5 6 7 8 9
pītum opening G→A′ 2-2-1-2-2-1-2-2 1 2 3 4 5 6 b7 8 9 G→A′ 2-2-1-2-2-1-2-2 1 2 3 4 5 6 b7 8 9
nīd qablim casting down the middle (W);
fall of the middle (D)
C→D′ 2-2-1-2-2-2-1-2 1 2 3 4 5 6 7 8 9 D→E′ 2-1-2-2-2-1-2-2 1 2 b3 4 5 6 b7 8 9
nīš tuḫri1
nīš gabrîm
raising of the counterpart (W);
rise of the equivalent (D)
F→G′ 2-2-2-1-2-2-1-2 1 2 3 #4 5 6 7 8 9 A→B′ 2-1-2-2-1-2-2-2 1 2 b3 4 5 b6 b7 8 9
qablītum middle B→C′ 1-2-2-1-2-2-2-1 1 b2 b3 4 b5 b6 b7 8 b9 E→F′ 1-2-2-2-1-2-2-1 1 b2 b3 4 5 b6 b7 8 b9

1Transliterated from cuneiform, from [West 1994], page 168.

2Translation from [West 1994] (W) and [Dumbrill 2005] (D).

Note that nīd qablim is the declared mode for Hurrian Hymn 6 ([Dumbrill 2005], page 121).

Also note that the pītum tuning is the same nine-string tuning regardless of whether string 1 is the highest or lowest tuned string.

It is tempting to equate these nine-tone (enneatonic) tunings to the seven-tone scales of the Greek modes and the Western Classical music traditions. This might be reasonable from a tuning perspective, since strings 8 and nine were tuned as shadows of strings 1 and two, an octave apart. However, once the lyre was tuned, we have no idea what was the tonal center of the music that was played. Even if you assume a seven tone scale, there are three choices of where in the nine-string lyre to play those seven tones (strings 1-7, 2-8, or 3-9).

The Embūbum Scale and Flutes

The central enigma of relating the Babylonian tuning system to flutes is the naming of the scale embūbum. Given the names of the other scales that seem to be based on the mechanics of the tuning process on the Babyloian lyre, the use of the word embūbum seems to argue against some arbitrary or fanciful naming of the scale, but indicates a real relationship with the flute.

There are two possibilities:

  1. A flute was used to provide absolute pitch reference for the beginning of the tuning process.
  2. The embūbum scale was tuned in the manner of embūbum flute instruments.

Possibility 1 was suggested in [Kilmer 2002]: “The embūbu ‘reed pipe’ was probably used for the tuning process”.

Possibility 2 was suggested in [West 1994].

I propose that there is a strong case for the idea that flutes were tuned to the embūbum scale:

  • The list of Babylonian tunings appeard in all the musical texts in the same order, shown in the table above. The embūbum scale is third in the list, and it would be unlikely to use an external reference that late in the tuning process.
  • The proposal by Galpin that the original Sumerian ti-gi (Akkadian tigu, tegu) or ka-gi flutes has three finger holes lends itself to a scale that exactly matches the embūbum scale.

This second point is demonstrated by a present-day class of flutes known as overtone flutes. We have three-hole versions of overtone flutes from a wide range of present-day ethnic cultures, the Slovakian Fujara and the Tabor-Pipe being prime examples. These instruments are designed to play predominantly in the upper registers.

On the Tabor pipe, the fingerings (all from [Kelischek 2011]) Papago three hole finger diagram closed closed closed Papago three hole finger diagram closed closed open Papago three hole finger diagram closed open open Papago three hole finger diagram open open open produce the notes C5, D5, E5, and F5 in the first register of the instrument. However, because the second register starts an octave higher, there are no playable notes between F5 and C6, making melodies difficult to play. So, the instrument is normally played beginning in the second register, with the natural scale of C6, D6, E6, F6, G6, A6, B6 played by the fingerings Papago three hole finger diagram closed closed closed Papago three hole finger diagram closed closed open Papago three hole finger diagram closed open open Papago three hole finger diagram open open open Papago three hole finger diagram closed closed closed Papago three hole finger diagram closed closed open Papago three hole finger diagram closed open open in the second and third registers. In the fourth register there are two more notes, C7 and D7, with the fingerings Papago three hole finger diagram closed closed closed and Papago three hole finger diagram closed closed open. So, the tabor pipe scale in second, third and fourth registers is C6, D6, E6, F6, G6, A6, B6, C7, D7, which has scale steps of 2-2-1-2-2-2-1-2 that exactly match the embūbum scale (assuming string 1 is the highest string).

In situation is similar for the Slovakian Fujara, except that it is substantially larger and has a higher length-to-bore diameter ratio that the tabor pipe. Most of the traditional melodies played on the fujara use the Papago three hole finger diagram closed closed closed fingering in the third register as their root. The fundamental tone on a typial fujara is G2, with the second register playing G3, A3, B3, C4.

Traditionally, tabor pipes are played with single musician playing the flute with one hand and a small drum with the other. There is no visual art showing this setup in Ancient Mesopotamia, but there is a lot of visual art in Babylonian and Egyptian cultures showing a single musician playing two long woodwind instruments. Ideally, these would have three or four finger holes for the best ergonomics of the musician.


The development so far raise some interesting questions and possibilities:

Did the scale of the flutes dictate that there were nine strings on the lyre?

The natural scale of C→D′ for the embūbum scale fits neatly into the second, third and fourth registers of an overtone flute. There is a natural bias today against the idea that flutes may have dictated how lyres were designed. We have much more visual art and texts about the lyre, probably because it was considered a more royal and proper instrument, and the flute was more commonplace. However, flutes may well have predominated in the actual culture of the time and determined how musical tunings were designed.

Is the seven-tone major diatonic scale prevelant in Western music traditions today a lineal descendent of overtone flutes from Ancient Mesopotamia?

The first seven notes of the embūbum nine-tone tuning are the un-tempered versions of the notes of diatonic major scale. And it is now apparant that the work of Pythagoras in developing the Greek modes and the Pythagorean tunings were based directly on his exposure to the Babylonian music traditions.

Of the seven Babylonian tunings, did embūbum become the most popular because it allowed lyres and flutes to play in ensemble?

Note that VAT 10101 shows that more love songs were tuned in embūbum than the other tunings, although we have song counts for only four of the seven tunings.

Did we learn to hear intervals of fourths and fifths from overtone flutes?



1. The older transliterations of nīš gabrîm and nīš GABA.RI have recently been rendered as nīš tuḫri. See [Crickmore 2008], footnote to figure 1.

Previous PageNext Page