World’s Oldest Musical Notation Deciphered on Cuneiform Tablet
Excavations at ancient Ugarit in modern Syria began in 1929. To date, thousands of cuneiform tablets have been unearthed revealing a Canaanite civilization which, in many respects, is linguistically and culturally closer to the civilization of the Old Testament than any civilization ever uncovered. Thus, these cuneiform tablets have been the subject of numerous scholarly studies.
By now, scholars know a great deal about cuneiform writing. It was used to write many different ancient languages, like Sumerian, Akkadian, Hurrian, and Hittite. In some cuneiform sign writing systems, the signs represent primarily whole words—logograms; in others, they represent mostly syllabic sounds. Often both are used together. Still other signs are non-phonetic—that is, they tell the reader something about the word, although the sign itself is not pronounced. The Ugaritic system is unique among cuneiform sign systems because it is alphabetic—no doubt adapted from the Semitic alphabet that had developed in Canaan a few hundred years earlier.
Between 1950 and 1955, during the 15th and 17th archaeological campaigns at Ugarit, archaeologists dug up many fragments of Hurrian cult songs. Among them were three fragments of a single tablet in different states of preservation. Miraculously, these pieces fit together. As a result, we now have an almost complete text known as the “Song Tablet.” After the tablet was put together, it measured about 7.5 inches long and about 3 inches high. It is inscribed on both sides and even on the edges in wedge-shaped cuneiform characters running from left to right horizontally across the tablet.
The left half of the tablet, as can be seen easily from the picture, is in excellent condition. It consists of a large fragment and a small fragment found together during the 15th campaign. This half of the tablet is light cream-colored and each cuneiform sign is distinct and clear. The right half of the tablet, from the later 17th campaign, is, unfortunately, in very poor condition. It is a dark brownish-gray and very difficult to read. Parts are illegible. Apparently, the tablet was broken before it was buried and the two halves lay for over 3000 years under different conditions—the left half in a dry area and the right in a wet one. The right half is so friable, or breakable, that bits of the surface have been lost even since excavation.
The writing on the tablet consists of three parts. First, there are four lines of text that run over, on the front (or obverse, as scholars call it) and continue on the back (or reverse side) covering even the right edge of the tablet. Below this four-line text on the front of the tablet are two finely drawn parallel lines. Between the parallel lines at each end, two angle wedges have been inscribed. Below the two parallel lines is the second part, consisting of six lines. This does not, however, continue on the reverse, although a few signs run over on the right edge of the tablet.
A third part is at the bottom of the tablet’s back (reverse) side. To read it, the tablet must be turned upside down. Without even knowing cuneiform, one might guess that this third text is a label which describes the contents of the tablet and which, perhaps, identifies the author or scribe.
This label or colophon on the back of the song tablet is written in Akkadian, one of the best known ancient languages. Translated it reads as follows:
“This is a song in ni
µ d qibli [tuning], a hymn (?) of the gods, from [the collection of] Mr. Urhiya; copied by Mr. Ammurapi.”
In other words, the label tells us that the music and words recorded on this tablet are those of a hymn of the gods.
The four-line text, on the obverse, contains the song’s lyrics. Unfortunately, the text above the parallel lines on the obverse is written in the Hurrian language. Hurrian is still imperfectly understood, and, although we do not yet have a proper translation of this cult-song we can, for the most part, read the text as it was pronounced in Hurrian. One phrase is quite clear: “Thou (the goddess) lovest them in (thy) heart.” The closing phrase of the hymn probably means “Born of thee.” The song appears to be a hymn in praise of the moon goddess and contains an assurance that good things derive from her. For her worshippers, she has love in her heart.
The second part of the text—below the parallel lines—proved to be the most difficult but most exciting. This text consists of Akkadian terms written in a Hurrianized manner. Initially, the text was undecipherable.
If the song tablet were all we had, we might have been able to guess that the text below the parallel lines was the song’s musical notation. Without additional sources, however, it would have been impossible to decipher that notational system.
Fortunately the song tablet came to the attention of scholars who were working on the problems of ancient music notational systems. In the end, four cuneiform tablets from different parts of Mesopotamia provided the key to unlocking the musical notation system of the song tablet from Ugarit.
The four other tablets are referred to as the “theory tablets.” Two of them concern the names of lyre strings and the names of intervals between the strings. The other two theory tablets relate to instructions for tuning the lyre and the names of the seven different tunings.
The most important information the theory tablets provided came from part of a mathematical text which was excavated at Nippur (in Iraq) and is housed in the University of Pennsylvania Museum in Philadelphia. The two theory tablets which describe how to tune a lyre were excavated by Sir Leonard Woolley at ancient Ur (in southern Iraq) and are in the collection of the British Museum in London. The fourth tablet came from Assur (in northern Iraq) and is owned by the Pergomon Museum in Berlin. Together these four tablets provided the clues needed to decipher the cuneiform note system.
The dates of these four tablets vary widely—from about 1800 B.C. to about 500 B.C.—but they all come from the Tigris-Euphrates basin and were all written in the Sumero-Babylonian cuneiform system.
Just as the theory tablets were widely scattered, so were the scholars who worked on them. Significant contributions to the step-by-step solution to the musicological problem were made by scholars at the University of Chicago, at Oxford University, at Liege, Belgium, at London, at the University of Tubingen in Germany, and at the University of California at Berkeley.
When the notational system was finally deciphered, it was discovered that the peoples of the Near East had a musical scale consisting of seven notes which closely resembles the do-re-mi scale we were taught as children. The 3500-year-old notational system was capable of recording melody and harmony together. Indeed, that is what is recorded on the song tablet.
Several other song tablets were found at Ugarit, but these are very fragmentary. In each case where the name of the tuning survives, however, the song is written in ni
This Hurrian music does not indicate rhythm, tempo, or musical ornamentation. But it does indicate both melody and harmony. The upper note of the interval represents the melody and the lower note contains the harmony. In other words, we have both melody and accompaniment written simultaneously, an advancement previously thought to have been accomplished no earlier than the European Middle Ages.
To play the Hurrian hymn, a modern replica of an ancient instrument was needed. The British Museum staff provided Professor Anne Kilmer, an Assyriologist at the University of California at Berkeley who first published the Hurrian text and music, with scale drawings of the best preserved actual ancient lyre—the so-called Silver Lyre from Ur (c. 2600 B.C.). Although the original wood of the lyre had completely decomposed, excavator Professor Leonard Woolley found that the silver which had encased the entire instrument was still intact. The silver from the original lyre is now mounted on a wooden museum model which can be seen in the British Museum today.
Professor Robert R. Brown, also of the University of California at Berkeley, built a working model of the Silver Lyre out of wood. Professor Kilmer collaborated with Professor Richard L. Crocker another Berkeley Assyriologist. Together they learned how to string, tune, and play this instrument which has a rich and sonorous sound.
Since all third millennium lyres have a clear bovine shape to their sounding boxes, they are often called “Bull Lyres.” The deep “voice” of these lyres may have reminded the ancient Sumerians and their neighbors of a bellowing bull or cow. The Sumerian bovine lyres are perfect accompaniment for a male voice.
Professors Kilmer, Crocker and Brown cooperated in producing a sound recording, called Sounds from Silence (Bit Enki Publications, Berkeley), in which Professor Crocker sings the Hurrian cult song and accompanies himself on the replica of the Silver Lyre. On the same recording you can also hear Professor Kilmer sing the song, accompanied by another replica lyre that is closer in time and place to the Song from Ugarit. Professor Brown based the reconstruction of this second lyre on a second millennium drawing on a piece of ivory from ancient Megiddo. The “Megiddo Lyre” is smaller than the bovine lyres, no longer has a bovine-shaped sound box, and its sound is in a higher register. It is better suited to a woman’s voice.
Although the performance recorded in Sounds from Silence is admittedly experimental, the song is interpreted as closely to the text as possible. Some listeners feel that the melody and style are too Western; others find it reminiscent of simple early church hymns. Some critics believe that the ancient music must have sounded more like today’s Near Eastern music than the interpretation presented.
Although we have recovered the scales and the basic notes of this Hurrian song, we cannot know what it really sounded like. We do not know whether the chords represented in the cuneiform notation were simple indicators that could be embellished (like some modern guitar chord notation), whether vocal performers stayed on pitch, strayed from it, or exercised melismatic patterns as they sang. All we have is a voiceless artifact of musical notation from the distant past.
(We express our appreciation to Professor Anne Kilmer who reviewed the manuscript and captions and offered many helpful suggestions.)
Sounds from Silence: A Record of the Past
Professors Anne Draffkorn Kilmer, Robert R. Brown and Richard L. Crock have collaborated to produce Sounds from Silence, a recording of the text and music of the Hurrian song tablet discussed in this article. The 33 1/3 rpm long playing record includes a recording of Professor Brown singing the Hurrian cult song accompanied by a replica of a 3rd millennium B.C. “Silver Lyre” from Ur, which Professor Brown himself constructed. Professor Kilmer then sings the same cult song accompanied by another replica Professor Brown constructed—this one of a 2nd millennium B.C. lyre from Megiddo. On the record’s other side, the professors explain both how to tune a lyre and how to play the seven scales on the instrument.
Also with the record is a large and beautifully illustrated booklet detailing how Assyriologists and musicologists from all over the world were able to decipher the Hurrian music tablet. The booklet contains drawings of several examples of ancient instruments. The text also details how Professor Brown reconstructed the lyres used in the recording.
To order the record send $21 ($24 it outside the USA) to BIT ENKI Publications and Records, P. O. Box 9068, Berkeley, California 94709. California residents please add 6% sales tax. Prices include surface shipping and insurance.
Step-by-Step Decoding of Musical Notation
The critical text for understanding the musical notation system is the so-called “mathematical text” which comes from Nippur and dates to 500 B.C.
The “mathematical text” includes in each line a pair of numbers and the name of the pair of numbers. Each line in the list begins or contains the Sumerian sign SA. In Sumerian, SA can mean sinew or gut and is also one of the signs in the word for cat. SA can also mean string of a musical instrument. It seems that gut was used to string musical instruments thousands of years ago. Each line of the tablet that begins with SA names two strings. The names of strings are known from the second of the theory texts, a lexical text from Ur. In the “mathematical text,” the word SA is followed by a string name and then SA is repeated with another string name. In the same line two numbers are given, and SA is again repeated with a special word following. The two numbers in the line, in conjunction with the SA, must stand for the distance between the two strings identified by the first two SA’s in the line; that is, the two numbers identify the musical interval between the two strings. This—the name of the interval—seemed the only logical meaning for the last appearance of SA in the line. This meaning was later confirmed by another tablet which we have not yet discussed.
Thus, in the “mathematical text,” the SA lines take the following form:
SA [meaning “string”] fore [the name of a string] and SA [string] fifth, 1, 5, SA [interval] ni
µ sû gabarî or (“rise of the duplicate”) [the name of the interval]
or
SA [string] next and SA [string] fourth-behind, 2, 6 SA [interval] ishartu (“normal”)
The names of the strings, as we have said, were also found in the lexical text from Ur. In addition, the “mathematical text” contains a section in which the lines contain only two numbers and the name of the interval preceded by SA (meaning “interval”). These lines take the following form:
1, 5 SA ni
µ sû gabarî
Thus, this confirms that SA in conjunction with the special word following it identifies the interval and its name. For example the numbers 1, 5 indicate an interval of a fourth, say from C to F. The numbers 1, 3 could indicate an interval of a third, say C to E.
The “lexical text” names nine different strings but the “mathematical text” refers to only the first seven strings listed in the “lexical text.” This suggests that the musical intervals listed in the “mathematical text” (and consisting of two notes or strings) are to be played on a seven-note scale, with the eighth note being the same as the first. This reasoning was later confirmed by the so-called “tuning text,” which is the third theory text.
Thus from the mathematical and lexical texts, we learned the names of the strings, the names of 14 musical intervals (7 fourths and fifths, 7 thirds and sixths) and the fact that the scale consists of seven notes before repeating the octave.
Additional information comes from the last of the theory texts, which is a catalog of cult songs. A portion of this text simply lists love songs in particular tunings.
What was meant by the different tunings is revealed in the already mentioned “tuning text.”
The “tuning text” contains instructions for tuning—or more precisely, for changing the tuning of—a lyre.
The “tuning text” takes the following form:
If the lyre was in X tuning, but the Y interval was unclear, you alter two particular strings and then the lyre is in Z tuning.
The key to understanding these instructions is to understand what is meant by “unclear.” Oxford musicologist, David Wulstan, first suggested that “unclear” referred to a particular interval (known to musicians as a “tritone,” an augmented fourth or diminished fifth) which sounds a bit discordant, unstable and, if you will, unclear. By changing one of two strings on the lyre by a half-step, this interval would be “clear,” but the instrument would then be in another tuning (or “key”) and another interval in the new scale would then be “unclear.” This interval could then be made “clear” by the same retuning process; the lyre would then be in yet another tuning.
Although this sounds somewhat technical, it can be understood by anyone who can play a scale on a piano.
As we have seen, the ancient scale had seven notes, just as we do: for example, C, D, E, F, G, A, B. Then the octave C is repeated. Unlike a piano, however, the ancient lyre had no “black” notes. Imagine a piano on which you cannot play the black notes. It would be possible to play seven different scales on such a piano, one beginning on each of the white notes. Each of these scales would be different because, without black notes, the half steps would come in different places within the scale. (The half steps on a piano are between E and F, and B and C; there are no black notes between these intervals.) With black notes we can play a half step (or half tone) above or below any other note. But without the black notes, we can’t do this. Using only white notes, we have seven different scales on the piano, one beginning on each of the notes. These are the seven tunings identified in the “mathematical text.”a
We can play each of these seven different scales on the piano without retuning the piano simply by beginning on the next higher white note each time. But this takes 14 successive notes. On the lyre, it would take 14 different strings. To obtain these seven different scales on a lyre of only 9 or 11 strings, it was necessary to re-tune the instrument so that the half steps would occur in the right places depending on which of the seven scales we wanted.
In a C-major scale, a half step occurs between E and F, and between B and C. Otherwise we proceed by whole steps or whole tones (that is, with a black note between). If we begin a scale on D without using the black notes, the half steps will occur at a different place within the scale (between the second and third, and sixth and seventh notes) and this is why a white-note scale beginning on C is different from a scale beginning on D.
If we understand how stringed instruments are tuned, we will be able to understand better what is meant by “unclear” in the tuning text. People who tune stringed instruments (including pianos) commonly do so by tuning a cycle of fifths, that is by starting on a particular note and tuning a fifth interval above it. Thus if we start on C, we could tune G above it. This interval is easily accessible to the ear, it is reliable, and is repeatable; in short, it is “clear.” After tuning G in relation to C, the tuner can then proceed to tune D in relation to just-tuned G. To do this, he descends from the G to a fourth below it; this interval is also clear and easily recognizable because, although the interval is a fourth, if inverted, it is a fifth; that is, D above G is a fifth rather than a fourth. So the tuner begins on the particular strings and tunes by ascending fifths alternating with descending fourths. When he has tuned all the strings in this relationship, the instrument is “in tune.”
Without the black notes, all scales will include a tritone (an augmented fourth or—the same thing—a diminished fifth). In a C-scale, this tritone or interval occurs between F and B. This interval is discordant or “unclear.” To “clear” it one string must be raised or lowered to change it into a pure fourth or fifth. This will alter the place where the half-step occurs within the sequence, and will change the “all-white note” scale.
At one point in the tuning tablet, the instructions are to lower the bottom string of the augmented fourth, thus producing a pure fifth.b If this were done seven times to each augmented fourth, the entire instrument would be tuned one-half step lower, and if we followed this procedure seven times, we would be, in terms of note relationships, precisely where we were at the beginning. That is why, in the tuning text, there is the instruction “no more” after the instructions for seven tunings.
This understanding of making an “unclear” interval clear and thereby creating a different scale or tuning, is consistent with everything else known about texts and no other hypothesis explains all the data.
At this point, the scholars working on the tablets knew from the mathematical and lexical texts the names of the tunings (or scales) and the names of the intervals. The names of the seven tunings were identical to the names of seven intervals. But which tuning applied to which interval? Professor Hans Kummel of the University of Hamburg put forward the technical but ingenious hypothesis that each tuning was named for the interval of a fourth or fifth with which the tuner began to tune the instrument. The tuning text instructed the tuner to “clear” the “unclear” interval by lowering a particular augmented fourth (or diminished fifth) one half step, thereby creating either a “clear” fourth or a “clear” fifth. The tuning then proceeded by tuning ascending fifths and descending fourths. The interval from which the tuning began (the “clear” interval created by lowering one of the strings of the “unclear” interval) became the name of the particular tuning.
In this way, scholars were able to identify, the seven scales that were named after the seven intervals of fourths and fifths. The remaining intervals—thirds, sixths, and the tritone—could then be identified exactly by their place in any particular scale.
Now it is time to return to the song text tablet from Ugarit which contains the words to our Hurrian hymn. Below two parallel lines on this tablet the text seems to contain the musical notation to the song.
The notation text takes the form “word + number, word + number, word + number, etc.” The words were easily identifiable as Hurrianized forms of the Akkadian intervals we had already deciphered and understood. These intervals could be played! (We didn’t and still don’t know what note they start on—i.e. what pitch—but we do know the interval relationship and where the interval occurs within the scale.)
Another way of expressing the form of the musical notation written below the parallel lines is:
interval name + number, interval name + number, interval name + number, etc.
or
interval name + 3, interval name 1, interval name 3, etc.
or
(using the actual Akkadian name of the interval)
qablite 3, irbute 1, qablite 3
or
(using the Hurrian form of the interval name, which is actually what is used in the notations below the parallel lines)
qabitu 3, rebutu 1, qablih 3
What the interval name meant was known but what was the significance of the number following it? Several possibilities were tried but the only one that “worked” (in the sense of matching the lyrics) was taking the number to indicate the number of times the interval, like a chord, was repeated. Thus, qablite 3 meant to repeat the qablite interval three times. This hypothesis fit well with the total number of syllables in the song. Certain minor philological assumptions had to be made about the song text, but with these assumptions (relating to the number of syllables), the text fit the music quite closely.
The resulting music is reprinted in the article. The particular tuning is indicated in the label or colophon on the reverse of the song tablet. This song is to be played in nid-qabli tuning.
In this particular tuning, the half tones fall in the same places they do in our modern major scale. Thus our song was written in the equivalent of a modern major scale.