The Neume Notation Project

urely, music has existed for as long as Homo sapiens, for (like the birds) man is a singing creature. The desire to record music in writing, too, is very ancient but (by definition) only since historical times. In the ancient world the motivation for notating music was likely to preserve a good melody of a song, as insurance against its loss by failure of memory.

In this role, music notation served as an aide-mémoire for someone who (with a bit of prompting) could remember how the melody ought to sound. Music notation was not -- until the Renaissance -- expected to be scientifically accurate, i.e., that the melody could be reconstructed solely from the notation. Actually it is questionable whether a writer of music expected that the song should be performed in exactly the same way each time it was sung. The idea that one could correctly sing a melody just by sight-singing the notation (and never having heard the melody before) seems to be a relatively recent idea.

I would add that music notation also had a mystical function. Ancient Egyptian tombs and burial chambers of many other cultures too, included implements, food, pictures, and writing that would in some way help the deceased person in the next life. Religious inscriptions on the walls of a tomb could bring good fortune to the spirit of the tomb's inhabitant. Likewise for the living, documents containing sacred texts, pictures, and music can be embued with mystical qualities. Sacred books can be thought of as embodiments of prayer, belief, and spiritual doctrine. As such, they are what the Catholic Church calls sacramentals (sacramentals are physical objects where the spirit realm touches the physical domain). One of my tenets is that books of chant served a mystical function in the Middle Ages: apart from any practical use they might have had, the books themselves were considered to be holy objects.

A church or monastery that possessed such books owned objects of spiritual power. (Compare the cave paintings at Lascaux, France, shown at left, ca. 13,000 b.c.e., suggestive of the mystical power that graphic representation has for humans.) The presence of these books could increase the holiness of a place and confer spiritual merit upon those who were near them, touched them, or looked upon them. I would add that, in some cultures, a person's knowledge of writing might have been considered to be a mystical art -- a type of shamanism, -- not just a credential of education.

To actualize the desire to write down music, a technique needed to be invented. Several methods were tried over the millennia. Today we take for granted the naturalness and efficacy of the modern music-notational system (called common practice notation), where sound is divided (or quantized) into discrete pitches that sound for various lengths along a timeline of regular beats. Common practice notation is now in use throughout the world. This method of writing music (a sort of Cartesian view of sound) was not, however, an obvious concept for the ancient mind.

Medieval chant notation–the subject of our study–differs from modern notation in significant ways. These dissimilarities comprise both the obvious visual differences (that is, graphological differences) and more subtle, conceptual differences (or, semantic differences). Despite the inexactness of medieval notation compared to modern notation as a method of quantization, it is important to bear in mind that both systems have their strengths and weaknesses. Common practice notation is not necessarily better; we shall see that medieval notation is able to capture subtlties of semantics that are practically impossible to capture in modern notation.

The oldest examples of music notation from Western Europe were written as annotations to sung texts. (Typically the chant text was written on the manuscript sheet first, then the music notation was added, perhaps by a different scribe.) Scholars who specialize in early medieval manuscripts date the earliest, surviving Western European examples to the 9th century of the current era (c.e.). The idea of music writing as text annotation remained the dominant form of music notation until fairly late in the Middle Ages.

The concept of using vertical placement of musical marks to indicate pitch was a novelty of medieval music writing in the West. We take this concept for granted in common practice notation, because it seems so natural and useful, but it was a significant breakthrough at the time it was invented. The notion of using distinct symbols to indicate the duration of notes, too, was also novel; as the Middle Ages progressed into the Renaissance, this idea of measured time took firmer hold.

In some ways, the semantics of medieval notation were more expressive than what we have available to us today; it was possible to notate subtleties of vocal rendering that were richer than what is possible to represent in modern notation. Knowledge of the correct interpretation of many of these subtleties of medieval notation was, however, lost with the decline of the oral tradition in Western chant. Today, reconstructing this lost meaning is an active area of research among medieval musicologists.

Top of document

We present here what we believe to be the cultural history that led up to the development of medieval chant notation. The history of music writing in the ancient world is quite interesting in itself, and it is a subject of much research and debate among musicologists. Probably the oldest record of any form of music notation in the Western World is from ancient Egypt. Tomb and temple paintings strongly suggest that the Egyptians used a system of hand signs for music, called cheironomy. By these signs a music director would indicate to musicians, or singers, the melody to be played or sung (click image at left). Some scholars argue that hieroglyphs associated with some song texts on paintings, or reliefs, indicate the melody to be sung [1]. Thus far, the scholarly evidence that ancient Egyptians had a written notation for music is not completely convincing.

Cheironomy is still in use today in the liturgical practice of some Middle Eastern sects. In particular, cheironomy is in use by some traditional Jewish synagogues to direct the singing of liturgical songs, and it is used also in several other Western religious traditions of ancient origin. Because we know that Egyptian cheironomy was in use by (at the latest) the twenty-fifth century b.c.e., and that the Hebrew people spent a substantial period under bondage in Egypt after that time, one plausible hypothesis is that the Hebrews adopted (or adapted) the Egyptian cheironomic system to their own use. An ancillary argument is that Egypt was a wealthy, highly centralized society in which art and abstract learning could flourish, whereas the Hebrews were then a nomadic people without permanent buildings nor the other accoutrements of a leisure society.

In more modern times, Hebrew cheironomy came to be written down in the form of mnemonic signs on Hebrew scrolls of religious texts. (Mnemonic signs are symbols to aid the memory.) These signs guided the cantor in the proper rendering of the text to be sung, basically by indicating the shape of the melody. These signs are written at the beginning of a verse, or immediately above the text to be sung.

Top of document

Many centuries after our earliest evidence of Egyptian cheironomy, but before the advent of Christianity, the ancient Greeks experimented with notation of music using a system of letters to indicate pitch. These letter names for pitches followed the Greek system of modes. The Greek modes are mentioned, but poorly explained, by surviving Greek manuscripts. Most of what scholars know about the Greek modes is from the sixth-century c.e., Latin writer Boethius, who had access to Greek documents that are no longer extant. There do remain several stone tablets (see, example at right) and papyri from ancient Greece that include song texts and accompanying pitch letters. The Greek notational system indicated the pitch of notes but not their durations; consequently, we can only guess at note durations. Presumably the indication of note durations was not necessary for someone already familiar with the melody. Several informed guesses have been made about how these ancient songs might have sounded [2].

Some scholars believe that the ancient Greeks were aware of the Egyptian system for writing down cheironomic signs. We know for certain that both Pythagoras and Plato were educated in Egypt, which had been the center of learning in the Western World for many centuries. Thus far, however, there is no indication that the Greeks adopted Egyptian cheironomy. Furthermore, no strong argument has been forwarded to suggest that scholars in medieval Europe were aware of the Greek system of writing notes using letters. The system of Church modes that arose during the Middle Ages was related to the Greek system of modes in name only; there appears to be no other connection.

Top of document

Patristic writings, the New Testament Epistles, and comparison of chant melodies indicate that the early Christian Church borrowed heavily from Hebrew worship for its musical liturgy. In particular, it seems highly likely that the Psalms were sung by the early Church in the traditional manner; early Patristic documents attest to a strong tendency by the early Church to defer to the authority of historical precedents, despite the inevitability of local, folk practices. James McKinnon argues that the musical heritage borrowed by the early Church was that of the Hebrew Temple, rather than the synagogue [3]. Musicological analysis indicates close similarities between Gregorian psalmody and their modern equivalents in Jewish liturgy; it is our view that these, at least, share a common ancestry. As with many other facets of early music study, however, the origins of early Church music are vigorously debated.

No evidence has surfaced to indicate that the early Christians attempted to write down the music of their liturgy. The earliest evidence of Christian music notation comes from the Eastern Christian Church in Byzantium, dating from perhaps the fourth century c.e. Indeed, judging from the poor quality of their written Latin, it seems safe to assume that chroniclers of the early Church did not possess sufficient education for this. Furthermore, before the year 1000 c.e., the Church's faith in immanent return of their Savior gave little incentive for committing such things to writing.

Top of document

During the time that Rome and its Empire were overrun by barbarians, the light of Western culture was largely extinguished in Europe. This is the period properly called the 'Dark Ages' (which some people mistakenly equate with the Middle Ages). Nevertheless, Western culture continued to thrive in the Eastern Empire centered at Byzantium (Constantinople), which maintained unbroken links to the cultures of the ancient world. There, Christianity continued to evolve in a peaceful, prosperous, and educated environment. The Byzantine Church developed a system of ecphonic signs by which to notate music above the sung liturgical texts (click example at left). Like the Hebrew mnemonic signs, Byzantine ecphonic signs represent the shape of a melody and its inflection, rather than specific pitches or durations of notes. Eventually, Byzantium fell to the conquest of Arabs, yet the system of ecphonic signs survives today in liturgical texts of the Eastern Orthodox Churches.

During the Dark Ages, literacy in Western Europe was sustained almost exclusively in the Catholic monasteries. They were insular and economically self-sufficient. It was not until much later that universities would be established or that monarchs would take any interest in reading, writing, and the transmission of culture. For hundreds of years virtually the only material that was committed to writing was liturgical texts, administrative documents of monasteries, and a few ancient texts that were copied principally for pedagogical purposes. Writing was done on parchment (sheepskin). Most documents were seen to be of only practical and temporary value. We see many examples of parchments where the ink of earlier writing was scraped off so that they could be reused for later scribal needs. Still others were cut into strips for use in binding of later books. Virtually no secular texts were written down and Europe relied heavily on oral tradition for the transmission of songs, histories, and epic poems. Well into the Middle Ages it would continue to be the case that monasteries and major churches would continue to be the only centers of education, and that only documents of ecclesiastical importance were committed to writing.

Top of document

The first evidence of music writing from the early Middle Ages is of uncertain date. The Benedictine monks of Solesmes Abbey (who have done much of the pioneering work in the study of medieval chant) estimate that earliest manuscripts come from the Saint-Armand monastery prior to a.d. 870 [4]. One of the goals of the present project is to help bring the power of computers to bear on the analysis of manuscripts for assigning dates trace the evolution of musical notation in Western Europe. The early history of music writing in Western Europe is exclusively the history of liturgical chant, known popularly as Gregorian Chant. It would not be until fairly late in the Middle Ages that popular music appears to have been written down.

Pepin

A watershed event occurred around a.d. 800. The Frankish king Pepin succeeded in bringing much of Europe under his control through military conquest. His son Charlemagne extended and consolidated his father's kingdom. Charlemagne was interested in the written word both for cultural reasons and as a powerful means of administering a far-flung kingdom. Latin was spoken throughout Europe but in local dialects that were mutually incomprehensible. Charlemagne saw written Latin as a means of standardizing communication and facilitating his political administration. He undertook to bring to his court some of the most educated clerics of his day, devise an efficient and standardized form of writing, and undertake a massive program to educate his subjects and promote the copying of manuscripts in great numbers.

Charlemagne's throne at Aachen

One can argue plausibly that it was during Charlemagne's reign, or shortly thereafter, that the various early forms of music notation using neumes were invented. After Charlemagne's death his vast kingdom fell apart, but the cultural trends that he had set in motion would have a deep and lasting effect upon Europe.

Both Charlemagne and his father before him were deeply religious. One can argue too (perhaps cynically) that they saw, in narrowing the gap between Church and State, that they could gain a more firm control over the minds and behavior of their subjects. For either or both of these reasons, Pepin and Charlemagne wished to standardize the liturgy of the Catholic Church. In many, isolated regions a form of liturgical song had evolved; at the major churches and in the monastic communities their members clung fiercely to their traditions of song and liturgy. Pepin undertook to suppress these local traditions. Charlemagne continued his father's policy; he brought from Rome representatives of the Papal chapel to instruct his clerics in the 'authentic' Catholic liturgy. A program of propaganda spread the idea that the chant used in Rome came directly from Saint Gregory the Great (died a.d. 604), who was universally venerated. Pictures were made to depict the dove of the Holy Spirit perched on Gregory's shoulder, singing into his ear God's authentic form of chant for the liturgy (click illustration at right). For many years a struggle between local autonomy and central authority raged on. Eventually most of the local traditions were suppressed. This is significant to musicology because, in most cases, the only record that remains of these local traditions of music is the manuscripts that have survived. Ironically, some of these traditions may have been older and more 'authentic' than what was received from Rome.

The standardized form of Gregorian Chant is largely the music that is still in use today; it is the official music of the Roman Catholic Church. It is compiled in a large book called the Liber Usualis using a modern, stylized form of square-neume notation.

Top of document

In the evolution of Gregorian chant, there was an important shift from oral tradition to written tradition that occurred gradually in the 16th and 17th centuries. This shift can be simplistically, but fairly accurately, understood in relation to the reign of Francis I of France [reigned 1515-47].

An important consequence of the plague, or Black Death (1346-61), was the loss both of personnel and income by Church entities. An estimated 25 million people died in Europe, or about a quarter of the population. I argue elsewhere (Silent Music, in progress) that this was the main factor in bringing on an end to the Middle Ages. Because of the huge decline in population, fewer men were joining the priesthood and religious orders. The vast land holdings of Church entities were producing far less income, because there were not enough laborers to bring in the harvests. The power of the Church had traditionally offset the power of kings, but in the generations following the Black Death, the Church's power waned. By 1516, Francis I was so powerful compared to the Church that he forced an agreement, whereby he would have the authority to appoint bishops and heads of monasteries. The ages-old tradition of Church leaders rising through the ranks was replaced by, essentially, political appointments. These political appointees had not been raised in the oral traditions of the Church; when they had to sing in choir, they needed books from which to sing; unlike their predecessors, they had not memorized the Church's complex body of chants. This, in my view, was the central factor that caused a shift away from oral tradition to written tradition.

Top of document

Our understanding of music from past periods stems from three classes of sources: (a) surviving music notation; (b) written commentaries about music; and (c) musical artifacts such as instruments. Our understanding of music after the invention of movable-type printing, in 1455, takes a quantum leap in both the level of detail and in quantity [7]. The invention of music printing made possible the dissemination of musical scores far beyond what had been possible in earlier periods; the resulting quantity of exemplars, if nothing else, improved the chances that some would survive to our day. Admittedly, printing from engraved blocks had been done in China as early as the eighth century c.e., and also in Europe by the fourteenth century, but it is inarguable that movable type vastly increased the facility with which documents could be printed. (See, also my discussion of fonts.)

Top of document

There are grave questions surrounding the whole matter of 'performance' of early- and mid-medieval chant. First, there is virtually no disagreement that ecclesial chant was an aural tradition for a very long time before and during the development of a system of music writing in Western Christendom [5].

Despite the great care taken–particularly in the St. Gall style of notation–to accurately record the nuances of melody, early manuscripts were probably not thought of as "scores" in the modern sense, i.e., they were not used as things from which to sight-sing. Craig Wright [6] argues convincingly that singing from manuscripts in choir would have been impossible: there was not enough candle light to permit reading during the night offices, the manuscripts were too small for group reading, and the manuscripts were produced in too small quantities for each choir member to have a copy. For this and other reasons, it is anachronistic to think of the early manuscripts as being associated with a musical "performance." The considered opinion of musicologists today is that for the first several centuries notation served primarily as a guide to the complex liturgical calendar and prescribed chants for each liturgical occasion, that is to say, as a reminder to the cantor about chants which he and the schola cantorum knew intimately by heart. A small number of the early manuscripts apparently were used for pedagogical training.

In addition to the purpose of musical manuscripts as aides mémoires, I argue elsewhere (Silent Music, in progress) that early chant manuscripts had a "book fetish" function. The cost in labor and materials of producing a chant book was very high, on the order of the cost of a new house today. The books were ordinarily kept locked in a cache box within a locked scriptorium. Their sometimes bejeweled covers and meticulous illuminations, sometimes in gold leaf, attest to a relic function rather than as books to be sung from. A central doctrine of the Church has been the sacredness of the Word, and by extension, the written Word of the Scriptures. Particularly after Charlemagne [reigned 768-814], the musical liturgy was raised to about the same level of sacredness, and so we can fairly infer that chant books were sealed away as precious and mystical sacramentals, much as were the bones and other artifacts of the saints. Some early examples of music notation were in books that were used for musical instruction.

In any case, it would be several centuries before music notation would be evolved to the point where it could accurately represent the important musical features of a "performance." As an oral tradition, the learned methods of chant were arguably obscure to us today. Despite the efforts of the Benedictines at Solesmes, starting at the end of the 19th century, and of others to "restore" Gregorian chant, we cannot know with any assurance what medieval chant sounded like.

For these and other reasons, it is our position that there exists, and probably cannot exist, any reliable "performance" tradition of the early manuscripts. The urge to provide an automated playing back of neume chants is quite alluring, but intellectually without good foundation. We therefore dispense with design criteria that would allow manuscripts to be translated into MIDI code, for instance.

On the other hand, there is a wealth of valuable information to be gleaned from the written manuscripts themselves regarding, for instance, the provenance and transmission of individual chant melodies, the dating of manuscripts, the culture of monastic calligraphy, and many other, related matters. In our representation we focus entirely on the logical and graphical aspects of these manuscripts. The logical aspects concern modelling the manuscripts in a paradigm that is at the same time useful to the musicologist and practical for computer implementation. The graphical aspects, in summary, concern preservation of as much detail from the manuscripts as possible (a so-called "lossless" digital representation) to permit analysis of their content as data tokens, and to maximize the latitude of known and anticipated use of these data.

Top of document