The Neume Notation Project


•  The Transmission of Oral Culture
•  Evidence that Chants Were Memorized
•  Memory, Analogy, and Imagination
•  Decline of Oral Transmission in the Latin West
•  Relationship of Orality and Early Neume Notation

The Transmission of Oral Culture 
By comparison to the norms of modern musical practice (where detailed writing of music has made the learning and performance of music possible without relying on memory), it might seem highly improbable that people in the Middle Ages could have memorized the enormous repertoire of chants that make up the cycle of religious observances throughout the year. This seems particularly amazing with regard to the Divine Office, which has different chants for each of the Hours (or, Offices) during the day and night. Nevertheless, examples exist today of bards who recite epic poems from memory, which are so long as to take days to complete. Ancient historians also say the Odyssey and other epic poems were recited from memory in ancient times.

Reading from modern musical notation.
Typically today, music is performed by reading musical notation.
Some music, however, (e.g., rock, jazz, and folk music) is improvised from patterns that are learned by oral transmission. This may be reinforced by partial notation (e.g., a 'lead sheet' of just a melody and chord letters).
Image: Celtic Woman, "A Christmas Celebration" (Dublin, 18 July 2007); on PBS (U.S.A.).
By one calculation, singers of the Western repertoire of chant had to memorize texts and melodies that would amount to eighty hours of singing (this estimate is attributed to Kenneth Levy by Treitler, p. 65). By Wright's estimation, singers had to memorize as many as 135 folios of text and music; he remarks, "The capacity to retain this enormous store of music is almost incomprehensible today, but the requirement was very much in harmony with the medieval approach to learning and disseminating of all kinds of information" (Wright, pp. 327-328).

One can more easily believe that liturgical chants were memorized if one considers that boys were entrusted to monasteries from a young age for the purpose of training in chant, an apprenticeship that took many years.

Perhaps surprisingly, it was typically the families of some wealth and social standing that dedicated their sons to a life of celibacy, work and prayer at a medieval monastery. (Of course, the concepts of life and the world for the medieval mind were quite different to what most people experience today.) A fascinating glimpse into the connections one family had to a monastery in southwestern France is provided by Grier. He has spend much of his career researching the life and work of one monk, Adémar, who was prolific both as a neumation scribe and as a 'composer' of chants.

[Adémar's] choice of [the monastery of] Saint Martial for these purposes was hardly providential: an ancestor on his father's side of the family, Aimo, had been abbot there in the first half of the tenth century (while his brother Turpio simultaneously held the office of bishop of Limoges), and his father's two older brothers, Adalbertus and Roger, were monks at the abbey. Adalbertus, the oldest, became deacon, while Roger, the middle brother, filled the post of cantor and tutored his nephew Adémar during his advanced studies. ... Adémar was also drawn to Saint Martial because the abbey enjoyed considerable prestige, and, perhaps most important for a scholar of wide-ranging interests like Adémar, it possessed an outstanding library. (Grier, p. 1)

Grier retells that King Louis XV purchased the holdings of the library at the Abbey of St Martial in 1730. Had this not happened, the manuscripts of Adémar's compositions would have been destroyed and his music lost forever when the monastery was forcibly dissolved in the French Revolution.


Evidence that Chants Were Memorized 
There is little doubt that Western liturgical music was transmitted orally during the Middle Ages prior to the invention of musical notation, and likely for some centuries thereafter. The earliest-known artifacts containing musical notation of chant (while neume notation was still an emergent system of writing) show an already-existing, highly-developed repertoire of chants. Also, prior to the earliest specimens of neumation, text artifacts refer to chants that appear in musical notation only centuries later (for example, lists of chant incipits with modal assignments; cf., Treitler, p. 238).

Wright gives a highly convincing argument that, at least in the case of Notre Dame Cathedral (Paris), the chants of the Divine Office were committed to memory. He studied the Chapter records of Notre Dame Cathedral (which document the Cathedral's expenditures) covering a period of several years, and he tallied the number of candles that the Cathedral acquired. Of course, religious observance required chanting of some Offices during the night throughout the year. Candles could have provided the only light during these hours. Wright calculated the total amount of time that the candles the Cathedral had on-hand could have provided light, such as for sight-singing of written music from manuscripts. His conclusion is that the Cathedral did not possess anywhere near enough candles to permit sight-singing of the night Offices, even if light were provided just for the schola cantorum (which is the small group of singers who lead the congregation in chant). Therefore, at least for the night Offices at Notre Dame Cathedral, the chants had to be memorized (cf., Wright).

The theory that Latin chant was orally transmitted in the early and middle periods of the Middle Ages, until at least the 10th or 11th century, is supported by evidence and argument in the musicological literature. Treitler goes beyond this, however, arguing that the performance of chant during time period may have been improvised to some extent, that is, by using melodic formulae and/or by the cantor directing the chant by cheironomic hand signs (Treitler, pp. 24-25, 32, 43, 53-56, passim). Treitler (along with Crocker, Palisca, and others) questions whether 'memorization' of melodies involved note-by-note recall, or rather the recall of a melodic pattern from a 'family' of similar melodies. Levy suggests that part of the intent of standardization of Latin chants under Emperors Pepin and Charlemagne was to suppress improvisatory variation: "the melodies of the Gregorian mass propers were crystallised under Charlemagne in an authoritative recension that left no substantial license for oral-improvisational manoeuvre" (Levy 1987, p. 61).

Levy also argues that improvisation according to a chant's mode may have been practiced in Byzantium, for example: "our emergent view [is] of the central Trisagion tradition as not a specific melodic configuration but rather a multiplex collection of modally-related options" (Levy 1972, p. 762). In the same article, he continues as follows.

Melodic details like these are too incidental to the conception of these chants to figure prominently and tenaciously in their transmission. What the comparison tends to show, simply, is one melodic option that an eighth-century Eastern version might have had. Once we accept the view of the Trisagion tradition as an aggregate of modally-related options, this is scarcely less valuable than a direct melodic correspondence itself would be.

Other variants and elaborations of the basic Trisagion materials ... circulate among Byzantine manuscripts. ... There is nothing among these that alters the view of the Trisagion tradition as, at bottom, not melody but acclamatory recitative—as musically conventionalized rhetoric, amplified within the boundaries of a low G-mode. Like most Ordinary chants, then, the Trisagion had no single melodic original. Its mainstream lay in its mode; and only as a secondary matter did it develop and transmit specific melodic formulations.

Levy 1972, pp. 762-763)

Memory, Analogy, and Imagination 
Memorization of a chant and representation of a chant in writing by neumation are both functions of the human mind, and they may be related to one another (although the logical and physical constraints on them are different). Perhaps due to this relation, we can see significant parallels between memory and the writing of neume notation. Considering symbolic representations generally, representation of neume notation as digital data also has analogy to memorization. Treitler gives an analysis of the process involved in remembering a chant or an epic poem, in which he touches on some important features of this analogy.
Modern psychology tells us that this [i.e., the storing of a 'photographic image' of the poem or chant] is an unrealistic view of the process of remembering. In a classic of the field Frederic C. Bartlett [Remembering: A Study in Experimental and Social Psychology, (Cambridge, U.K.: 1932)] has written that remembering is a process not of reproduction but of reconstruction. I shall try to describe the important points of Bartlett's theory, especially as they might illuminate the problem about oral transmission.
  1. The theory of remembering depends on the theory of perception, for the way we recall experiences depends on how we grasp them in the first place.
  2. Perceiving is not passive reception, it is active organising. We strive to assimilate newly presented material into the setting of patterns and schemata left from the encounter with past experience. But that always results in the reorganisation of those patterns. Thus perceiving, and indeed conscious life in general, is a continual process of adjusting our own records of our past.
  3. In perceiving we draw out certain salient features of the matter presented that are for us especially prominent. These serve as signposts for the process of assimilating and reorganising.
  4. Those signposts play a central role in remembering. As perceiving is not simply a matter of the reception of stimuli, so remembering is not simply the storage of stimuli strung together and their later reproduction. Rather it is an active process of grouping appropriate details about such salient features. It is a process of construction, not reproduction.
  5. In remembering, therefore, we activate and reorganise the patterns of past experience. This has two important corollaries: (a) each recall is based, not on some fixed model outside ourselves, but on our own assimilated version of the matter recalled - not on the 'original' but on our most recent rendering; and (b) recall must be in conformity with the existing schemata in which our mind is organised. What does not conform will tend to be corrected or eliminated.
  6. The latter tendency, together with the role of persistent detail in remembering, leads to stereotyped forms. This is especially important in the conventionalization of the forms of cultural expression.
  7. In the recall of narratives, beginnings and ends especially provide those stand-out, persistent features that serve as the focal points of reconstruction. Consequently it is beginnings and ends that tend to become most stereotyped in repeated recall.
  8. Form, as well as salient detail, is persistent and is therefore an important factor in what makes remembering possible.
  9. A salient detail may be common to two or more themes or streams of interest, and it may serve as a crossing point between them. In that way the theme originally presented may be left and another entered.
  10. Remembering and imaginative construction are on a single continuum. They differ from one another in degree, but not in kind.
(Treitler, pp. 159-160)

Similarly, Michalski says that experimental results of clinical psychology suggest that the human mind has a "memory organization that is primarily oriented toward storing analogies and generalizations" (Michalski, p. 125).

In connection with the theory that early neumation denotes subtle nuances of vocal expression, we can speculate that memory may be intimately interconnected with emotion. For example, a person seems to remember emotionally-charged experiences more vividly than is the case with other memories. Studies about memory loss in the elderly also suggest that memories of emotionally-significant episodes are not lost as early as are other types of memories in disorders such as Alzheimer's disease. Major progress in the sciences of neurology and psychology must, however, be made before we shall understand the connections between memory and emotion.

Pending further progress in the natural sciences, it may turn out that memory is the key (as-yet poorly understood) to understanding what the neumation scribes intended to denote by neumes, particularly during the period when chant was transmitted orally.

Decline of Oral Transmission in the Latin West
Due to external forces of history, the chain of oral transmission was broken. In (Latin-speaking) Western Europe, this would have occurred by the 15th or 16th century, although the process likely was faster in some places than in others. For instance, Rankin remarks that "Roman cantors ... sustained an entirely oral musical tradition until much later than in the north" (Rankin, p. 288).

An underlying cause of decline in oral transmission of chant in western Europe was the Black Death (ca 1346-1361). Cartwright says that, "The Black Death ... was probably the greatest European catastrophe in history" (Cartwright, p. 38). Somewhere on the order of 24 million people died, perhaps one out of every four people! For monasteries and other religious houses, the Plague caused declines in membership and loss of rents due to shortage of labor in the fields. Wright documents this in the case of Notre Dame of Paris.

In material matters the Church suffered badly from the Black Death. A great loss of manpower and impoverishment through inability to cultivate her vast tracts of land rendered her a less dominant power in 1350 than in 1346. But greater harm resulted from her helplessness in this time of disaster, a large loss of priests and monks, and her failure to control their successors. Parish priests, the best-loved of church workers, died by the hundred and according to William Langland their benefices were all too often hurriedly filled by 'numbers of youths, that had only devoted themselves for clerks by being shaven'. If Langland is to be believed—and there is no reason to disbelieve him—the friars, who had previously been renowned for holiness and charity, gave themselves up to 'gayness and gluttony' ... [that is to say, the new friars did this]. (Wright, p. 48)

Throughout the Late Middle Ages the statutes of Notre Dame [of Paris] decreed that there be sixteen unbeneficed clerks of Matins in the choir of the cathedral. ... In 1347, for example, eighteen clerks of Matins were employed but by 1428, because the Plague and the Hundred Years War had caused the alienation or destruction of much revenue-producing land, the number had been reduced to five. To maintain the dignity of the liturgical service, the chapter of Paris took extraordinary measures. Early in 1426 it sold the older organ of the church for the value of the metal in its pipes, the nearly 800 pounds of tin bringing seventy-two francs into the coffers of the church. Then on 20 April 1436 the canons decided to part with a valuable icon, one given to them on 10 September 1422 by the regent of France, John, Duke of Bedford. The decision of the chapter, as recorded in the capitular acts, is witness of the penury the church experienced at that time .... (Wright, p. 24)

In the same era, the power of kings began to ascend over the Church. This affected the Roman Papacy, bishoprics, cathedrals, and monasteries, all of which had enjoyed a large amount of political financial and independence up to that time. The trend is clear when Francis I of France (reigned 1515-1547) imposed on the Pope (in the Concordat of Bologna; 1516) to permit Francis to make Church appointments of archbishops, beneficed choir positions, and so on. This authority he abused increasingly for payment for political favors. At Notre Dame Cathedral, for instance, "The chapter of Paris became less an independent college of powerful ecclesiastics and more a collection of royal minions eager to do the king's bidding." (Wright, p. 299)

Political appointees lacked the long training in chant traditions that had kept the chain of oral transmission intact. In order to fulfil the religious obligations of their posts, these 'new clerics' needed to sight-sing from manuscripts, and they needed musical notation that was easy to learn. The advent of printed music (ca 1476) gave further impetus to standardization and simplicity of notation. We have to infer that, as the notation of chant was simplified, and as singers relied increasingly upon written notation for their knowledge of chant, the chant itself likely evolved into something simpler and more predictable than it had been in prior centuries.

By the fourteenth century (about the time of Chaucer), composers in France, Italy, and across Western Europe, were writing much more secular music than sacred music as the Church's importance greatly diminished in society (Grout, p. 140). Secular music had existed throughout the Middle Ages (even it wasn't written down), but it gained far more serious attention as 'art' during what was called the 'Ars Nova' movement. Seay describes this momentous change as follows.

Musicians were no longer recruited primarily from within the Church, and even when serving the Church, took little account of the functions their music was to serve. The interest already found in music as a technical toy became, by the end of the century in France, the overriding concern; we can only react, in all too many cases, with a certain amazement at the amount of sterile complexity and meaningless intricacy therein.
(Seay, p. 127)

These events account for the break in the chain of oral transmission in the West. Due to this break, today's oral interpretations cannot be relied upon for giving us an understanding of early chant. The Gregorian Chant that is published today in the Liber Usualis is a historical reconstruction; it is largely speculative with regard to the earliest chants and notations. Our knowledge of early chant and neume notation is further hampered by the fact that the Frankish Emperors Pepin and Charlemagne (8th century) used the power of their government to suppress local customs of religious practice.

Due to the high cost of parchment (or, sheepskin) an unknown quantity of neumed manuscripts were scraped off with a knife in order to re-use the parchment for writing some new text. Some were cut in strips to make book bindings. In modern times, large quantities of liturgical manuscripts were burned during the anti-clerical French Revolution. Then during World War II, Allied bombing destroyed more ancient manuscripts, such as at Chartres Cathedral. It is impossible to know how many of the destroyed manuscripts contained ancient neume notation, or how many of these were the last remaining record of a long-forgotten oral tradition.

Relationship of Orality and Early Neume Notation 
When the enormous repertoire of chant was committed to memory by many thousands of religious, this might explain why the early neumation scribes apparently thought it more important to preserve in writing the subtle nuances of vocal expression, rather than to record exactly the pitches and durations of notes. That is to say, they may have been more concerned with writing down how a chant should be sung, rather than what was sung. In particular, the cantor (who was the person most likely to have consulted these manuscripts) might need reminding of: (a) what were the chants appointed for the particular day; and (b) what was the venerated, traditional manner of vocal interpretation for that day's chants. The purpose may have been for chant to elicit as well as to express feelings appropriate to the texts and the religious occasion—to make the experience of prayer emotionally vivid and also pleasing to God. Until relatively late in the Middle Ages, such aids to memory given by manuscripts likely would not need to include a precise 'musical score' for well-known melodies.

A further consideration is that parchment was very expensive, and writing or copying a manuscript was highly labor-intensive. In many cases, early neumation just was added to an existing manuscript of chant texts. An example is in the manuscript from the St Gall monastery that is excerpted below. Obviously, neume notation was added in the left margin of a chant text that had been written many years earlier.
Unheighted neume notation written in the margins of a manuscript from St Gall.
Example of a chant manuscript in which neume notation was added in the margins at a later date.
Even if neume notation on a staff had been invented by this time, not enough space would have been available for writing a staff above the existing chant text.
Image: 'Codices Electronici Sangallenses' (; used by permission.
Source: "Summi triumphum," St Gallen, Kantonsbibliothek,Vadiana 317, ff. 13v; excerpt.

Using alphabetical pitch letters (A, B, C, D, ...) for writing melodies had been done at least since ancient Greece (cf., image of Greek music on stone fragment). The modes and the degrees of a scale were well known to chant scribes in the early Middle Ages. It seems, therefore, that if their intention had been to record note pitches, they would have done so. Some early types of neume notation have more than one way of writing a melodic figure with exactly the same 'notes'. A scribe apparently would choose one way rather than another for conveying information that cannot be understood just in terms of pitches and durations of notes.

Despite the cultural bias of thinking that modern musical notation 'makes more sense' or that it is 'more accurate' than early neumation was, nevertheless it is impossible to represent in modern musical notation many aspects of chant that early scribes represented in neumes. With regard to the human voice (as opposed to musical instruments with fixed pitches and mechanical actions, such as the pipe organ), early species of neume notations represented subtle nuances of vocal expression for which no equivalent methods exist in modern musical notation.

Wright [in a personal communication] says that the composers and notators of chant in the Middle Ages were much more sophisticated than people today generally imagine they were. They seem to have had more skill, subtlety, and insight about melody and text for the human voice than modern composers have had. Because the chain of oral transmission was broken, however, we shall never know with certainty how chant sounded in the Middle Ages. Some written artifacts are all that survives for us to study.

References Cited

[1] Cartwright, Frederick F., Disease and History, (New York: Dorset, 1972).

[2] Grier, James, The Musical World Of A Medieval Monk; Adémar de Chabannes in Eleventh-Century Aquitaine, (Cambridge, UK: Cambridge University, 2006).

[3] Grout, Donald J., and Claude V. Palisca, A History of Western Music, 4th ed., (New York: W. W. Norton, 1988).

[4] Levy, Kenneth, "On the Origin of Neumes," Early Music History, vol. 7, (1987), pp. 59-90.

[5] ——––, "The Trisagion in Byzantium and the West," Report of the Eleventh Congress of the International Musicological Society, (Copenhagen: 1972), pp. 761-765.

[6] Liber Usualis, The Benedictines of Solesmes (eds), (Tournai, Belgium: Desclée, 1934); also various other dates of publication.

[7] Michalski, Ryszard S., "Two-tiered concept meaning, inferential matching, and conceptual cohesiveness," in Stella Vosniadou and Andrew Ortony, (eds.), Similarity and Analogical Reasoning, (Cambridge, UK: Cambridge University, 1989), pp. 122-143.

[8] Rankin, Susan, "Carolingian Music," in Rosamond McKitterick (ed.), Carolingian Culture: Emulation and Innovation, (Cambridge, U.K.: Cambridge University, 1993).

[9] Seay, Albert, Music in the Medieval World, (Englewood Cliffs, NJ: Prentice-Hall, 1965).

[10] Treitler, Leo, With Voice And Pen; Coming to Know Medieval Song and How It Was Made, (Oxford: Oxford University, 2003).

[11] Wright, Craig M., Music and Ceremony at Notre Dame of Paris, 500-1550, (Cambridge, UK: Cambridge University, 1989).

The Neume Notation Project homepageThe Neume Notation Project homepage

The NEUMES Project homepageThe NEUMES Project homepage

This revision: 24 June 2009.
Copyright © 1995-2009, Louis W. G. Barton.