R. John Blackley

R. John Blackley, director of the Schola Antiqua, has written various articles on the subject of proportional rhythm chant. Two of these---"On Realizing Gregorian Chant" and "Rhythm and Nuance in Chant"---are given below and may be downloaded and printed. Contact: blackley@embarqmail.com.

On Realizing Gregorian Chant

It is an extraordinary and complex treasure. There are hymns from as early as the 4th century, psalm-tones of ancient age, an immense body of Mass Proper chants in five important 10th-century manuscripts (and in a few valuable 9th-century fragments), the body of Office Proper chants in one 10th-century manuscript, and a collection of Mass Ordinary chants some of which are no later than the 10th century.

What is not so wonderful is that melodic pitches for the hymns were not written down until the 11th century, and then in many versions; we do not have early musical indications of psalm-tones in their simplest form; the musical signs for the 9th- and 10th-century Mass and Office Proper chants indicate melodic configuration and rhythm, but not pitches; and there is no early music for Mass Ordinary chants (except for two examples from the 10th century).

Clearly, the place to start is with the 9th-century fragments and 10th-century manuscripts. They contain musical signs, commonly called neumes, which are graphic representations of the shapes made in air by the conductor's hand; the singers knew the music by heart, needing only the movements of a hand to bring forth song. Fortunately for us, there is a wonderful congruence between the pitchless neumes and the later "diastematic" manuscripts whose notes indicate pitches: we can take the information from the neumatic manuscripts, conflate it with the pitches from the diastematic manuscripts, and come up with a reasonable likeness of how the chant was sung in the 9th and 10th centuries and perhaps earlier.

The big problem in all the early chat is its rhythm. The main controversy had been between those who hold that the melodic notes are basically equal in length and those who claim that a proportional measure of long and short is discernable in the neumes. The conflict has been long and sometimes nasty. But both sides agree on one thing: the rhythm of the chant is free--and it is with this point of agreement that we shall begin.


Western music from the mid-12th century till the present has been in regularly recurrent measure--modal rhythm during the Ars nova, isorhythm during the time of Ockeghem, and metrical rhythms for the most part thereafter. Before the mid-12th century, however, it would appear that freedom, not regular recurrence, was the rhythmic norm in Western music, its matrix. Individual notes related to one another freely, not as parts of any overall regular pattern. There were occasional instances of duple or triple meter, as in the simpler hymns or in polyphonic pieces, but these arose from the matrix of rhythmic freedom and returned to that matrix.

If we examine chant manuscripts from the 11th century on, we find for the most part no indications of length in the neumes and no overall rhythmic regularity: the notes are equalist, i.e., of equal length (though the singer naturally holds notes at the ends of phrases), and they are arranged freely.

If we examine the 9th-and 10th-century neumatic fragments and manuscripts, however, we find two kinds of rhythmic indications, those that are essential and those that are expressive. The essential indications are found in the shapes of the neumes themselves; expressive or interpretative indications are in certain letters that accompany some of the neumes. A bifolio fragment in Laon 266 as well as the Mass Gradualia Chartres 47 and St. Gall 339 show rhythmic indications solely in the shapes of their neumes.

Among these three sources, four hymns are to be found: Benedictus es in firmamento cæli, the Gloria, laus et honor and Crux fidelis/Pange lingua from Holy Week, and Omnipotentem semper adorent. Examination of these and of recitative portions of more sophisticated Mass and Office chants tells us something of the mentality in which early Western chants were written. Each syllable was considered important and was given its own parcel of musical time. That parcel might be divided into two, or it might be augmented: but ordinarily a syllable had no less than a certain basic portion in time during which it could be itself, and within which it could manage properly to related to its fellow-syllables, so that each might in rhetorical phrase and sentence have its own just weight and find its own role. And each sung syllable held its weight freely. Only rarely did it enter into regularly recurring patterns with other sung syllables, and this only when the metrical nature of the spoken text itself urged it to do so.

So we might expect to find simple early hymns sung in regular or nearly regular meter, and expect that later, more sophisticated hymns or versions of hymns would be sung with some hint or sense of regularity. And always with deliberation. We can expect the prose texts of the Psalms to be sung slowly in their simple melodies, each syllable deliberate, occasionally divided or expanded, but normally one note per syllable, measured in free pace, not rushed. In a certain solemnity. A chant sacred in its grounding, not ethereal: feminine of Earth, of Tellus, and not of the air.

And so the fruits of this musical approach include a vocabulary of ordinary notes which might be divided or augmented, a matrix of freedom with the capacity for hints of regularity, and a stateliness of movement. And melody was always sensitive to textual accent, since each syllable had it own parcel of musical time that allowed it to relate properly and with ease to the weight of other syllables.


Let us take for an example the responsorial hymn Omnipotentem semper adorent. A good part of the hymn with its neumes is given on one side of the bifolio fragment in Laon 266, dating from around 880-. The monk Walahfrid Strabo, author of Hortulus, the great poem about herbs and gardening, wrote the text (and quite possibly the chant) between 825 and 850. We can safely trust that the musical signs are uncorrupted and represent the composer's intentions. The hymn is virtually unknown, but it is a very great one. Here is the refrain and the first two verses, with their neumes. For the second verse, only neumes different from those in the first are reproduced.

Beginning with the first line, observe the neume over the syllables mni-, tem, o-, rent, cant, and mne. It looks like the wings of an upside-down bird, and is the shape of the conductor's hand when it is held still for a moment horizontally in the air, slightly bent upward at the middle knuckles, signaling the singers to hold the note. It is the ordinary neume. Its name, tractulus, is the diminutive of the Latin tractus, meaning "to draw out," having reference both to rhetorical lengthening and to a physical spread or "tract."

Example 1

At the very beginning of the hymn, two neumes are found over the syllable O-: the first is a tractulus; the second, which is called a virga ("branch"), is the shape of the hand's movement upward and forward. This particular virga has been turned into a liquescent--a note that divides into two and becomes muted, enabling the difficult mn sound that follows to be sung more easily. Compare these two neumes, tractulus and liquesced virga, with the single neume over the syllable po-. The single sign over po- indicates two notes, one lower and one higher; it is called a pes ("foot"). What is the difference between the two notes over O-, indicated by two neumes, and the two notes over po-, indicated by one neume? In the first syllable, the ordinary note or tractulus has been augmented by the addition of a second note; in the syllable po-, the ordinary note has been not augmented, but rather is divided into two. A similar division occurs over the syllable per, where the notes descend, in a neume called the clivis ("slope"); in a pes over the syllable ad-; in a liquescent pes over the word et; and in a clivis over the syllable ne-.

It would seem sensible to say that the tractulus is a long note, and that the virga is also; that the pes consists of two shorter notes ascending; and that the clivis consists of two shorter notes descending.

When we examine these chant lines, it becomes clear that each syllable ordinarily receives at least the length of a long, which is sometimes augmented by one or more notes and sometimes divided into two. Common sense tells us that, although there is no overall regular measure or meter of any kind in this music, it does have longs and shorts measurable with regard to one another, the short being half the length of the long. For this reason the chants of the neumatic manuscripts have been called proportional, for they contain longs and shorts that are measurable relative to one another, while occurring freely.

Three other signs are used in this line. The point over the syllable sem-, called punctus, is higher than the tractulus that precedes it and is not held like the virga that follows; it appears to be a short. Its effect is that of a rhythmic ornament, for it breaks up a sense of the binary that might result from the divisibility of the longs; it occurs again, with the same effect, over the syllable di-. The clivis in the word omne is liquesced. Finally there is the neume over the word per; it is like a clivis, but with an introductory note, so that its movement is low-high-low. Comparative analysis of this neume shows that the third note is often divided into two, which would indicate that it is a long. The sign has been called a torculus or "winepress," probably because of the movement of the hand going up and pushing down.

In the two verses that follow, only one new neume may be seen: the shape with two curves, at the end of the first syllable, like a backward s lying on its side. It is named oriscus, perhaps from the verb orior, "to arise," and is an ornament whose execution is uncertain. Comparative analysis seems to indicate that it is a long. We might sing it as a note beginning on pitch, subtly bending upward, increasing musical tension before resolution in the note or notes that follow.

As mentioned, there is a remarkable melodic congruence between Mass chants in the neumatic gradualia and in the later diastematic manuscripts in which the pitches may be found. But this responsorial hymn entered late into the liturgy and may have been sung only locally, so we might expect difficulty finding a source we can use for pitches. Bruno Stäblein's Die Hymnen gives only one possible source, in a 12th-century graduale from Narbonne (Paris Bibl. Nat. lat. 780, f. 114v). Here is its version of the melody:

Example 2

Some creative imagination is required to conflate neumes and pitches in this instance, but not too much. In the transcription or realization that follows (at the top of the next page), we use four lines with neumes of square and diamond shapes. Experience has shown us beyond any doubt that it is far easier to read and sing chant in this notation than in modern staff with modern notes, or by using slur marks to show ligatures as in the example above; modern notation is designed for meter, not for free rhythm. Do is located on the top line, as indicated by the sign at the upper left. Black notes are longs, the ordinary notes; white notes are shorts. There is no significance in the difference between square notes and diamond notes; the shapes arose naturally from the use of stub-tipped quill pens, and the variety helps the eye see the notes weaving together. Notes are read from left to right; when one note is directly above another, the bottom one is read first. A small note following a larger indicates that the larger note is liquescent: sing the larger note, then move the voice to the pitch indicated by the smaller note, while subtly muting the sound; this produces a barely discernable space that allows a difficult sound such as m, t, p, or n to be sung easily. For convenience, the early neumes have been written above our neumes. We have added short vertical strokes or ictus below or above notes, not so much to show accents as to serve as focal points for the singers; they are editorial, and not to be found in the manuscript.

Example 3

While the text itself moves in a regular accentual pattern of threes and twos, there is only a vague sense of regularity in the melody: in fact, the melody moves freely among its longs and shorts. The singer must be careful to sing from a free place: the longs and shorts must not be regimented, but time must be taken with them so each syllable has a chance to be itself in terms of the whole musical line. (The chant in my commercial recordings with the Schola Antiqua has been too fast and too strict. In my concern that musical line be made clear to the listener and that the chant in no way be considered lugubrious or boring, we sang too quickly and the line did not have a chance to breathe. The reader is cautioned to learn from my error.)


A bifolio fragment in Laon 121 dating from around 900, and also the 10th-century manuscripts Laon 239, St. Gall 359, and Einsiedeln 121, give rhythmic indications not only in the shapes of their neumes, but also in letters indicating rhythmic nuancing.

It is significant that there is not just one neumatic system among the sources we have specified, but three: Chartres 47 is paleofrankish, the three Laon sources use what is sometimes called "Metz" notation, and 339, 359, Einsiedeln 121, and the Hartker Antiphonale contain "St. Gall" notation. There is essential agreement also with six page from a 10th-century graduale in Nonantolian notation (Monza B. 1, 41; Ambrosian Library, Milan, S. 37 sup.). Thus we have eight early sources dating from across 120 years, from different geographical areas, using four different musical notations--all in remarkable agreement as to melodic configuration and rhythm. Free proportional rhythm was not a local or temporary phenomenon. It was the way chant was sung in the West in the 9th and 10th centuries, and most likely in the 8th as well.

Let us take from our second group of manuscripts a more complex chant: the Introit or entrance-song for the Mass of Pentecost, Spiritus domini, from Laon 239, f. 63rv. Among the congruent 11th- through 14th-century diastematic manuscripts from which melodic pitches may be drawn, we prefer the so-called "German" over the "French": the former, which at climaxes and cadences tends to stretch upwards a minor third to do or fa rather than being content with a major or minor second, seems more ancient. We go mainly to the 12th-century Graz 807, but also to the 11th-century H-159 Montpellier, the 13th-century Verdun 759, and the 14th-century Thomaskirche Graduale. Disagreements among them are few and relatively unimportant. In the example that follows, Laon 239 has bene conflated with the diastematic sources; the early neumes have ben written above. In the opening antiphon we have three completely new neumes. The ornamental neume over the first syllable of terrarum is a stropha ("artifice"), two shorts and a long that are repercussed. We think it should be sung lightly, in most cases, and never sharply like staccato notes. The second sign over the o- of omnia seems to be not a liquescent pes but rather a long preceded by an ornamental light and quick grace-note. The first sign over the lu- of the penultimate alleluia is a porrectus ("tyo reach out"): three notes, high-low-high, the last shown by comparative analysis to be a long. We understand the small letters accompanying some of the neumes to be expressive or interpretative in their intent: t = tene, "to hold"; c = celeriter, "quickly"; n - non, "not [held or hastened]." a = auge, "to lengthen," we understand as a slur or tie: between two notes of the same pitch, it is a tie, and the notes are held as one, without separation; between two notes of different pitch, it is a slur, the notes are conjoined in the voice.

A cantor intones the antiphon through the word domini, and the schola cantorum or group of around four to eight singers continues. The cantor sings all or the first half of the Psalm-verse Exsurgat deus, and the schola repeats the antiphon from the beginning. Note how important it is that the reciting notes in the center of the antiphon and the ordinary notes within the Psalm-tone are longs, so that the text is declaimed. The reader is invited to learn, sing, and savor this piece: to see how integral the text is within the long and noble melodic lines, and with what ecstasy the words "and that which holds all tings in its grasp has knowledge of the voice" are painted. No other music could do that!

Example 4

It remains now for us to take a brief look at some relevant sources of music and musicological thought. Most manuscript sources are given in facsimile in the series Palégraphie musicale (Tournai or Solesmes; reprinted in Berne); not also H 159 Montpellier: Tonary of St Benigne of Dijon, transcription by Finn Hansen (Copenhagen: Fog, 1974) Das Graduale der St. Thomaskirche zu Leipzig (two volumes, reprinted Hildesheim: Olms, 1967); and Verdun Bibliothèque Municipale 759 Missale (Padua: Linea, 1994).

Clear investigation of free proportional rhythm is to be found in Jan W. A. Vollaerts' Rhythmic Proportions in Early Medieval Ecclesiastical Chant (Leiden: Brill, 1958, 1960) and in his Benedictine student Dom Gregory Murray's Gregorian Chant According to the Manuscripts (London: Cary, 1963). John Rayburn's Gregorian Chant: A History of the Controversy Concerning its Rhythm (1964; reprinted Westport, Connecticut, 1975) is very helpful through 1960. Some articles by liturgist Stephen J. P. van Dijk are important as to the development of chant before the 10th century: "Papal Schola versus Charlemagne" (Organicae Voces, Festschrift Joseph Smits van Waesberghe; Amsterdam 1963); "The Old-Roman Rite" (Studia patristica, 5, Texte und Untersuchungen, 80; 1962); and "The Urban and Papal Rites in seventh- and eighth-century Rome" (Sacris erudiri, 12; 1961);

Van Waesberghe, S.J. wrote against Vollaerts, S.J., then recently deceased, and Murray made very thorough response. The Benedictine Dom Eugène Cardine's critique of Vollaerts and Murray in "Is Gregorian Chant measured music?" (Translated by Aldhelm Dean, Solesmes, 1964) was not answered by Murray, who had withdrawn from the controversy, perhaps at the urging of his religious superiors. Cardine's objections depended upon a rigidity in the interpretation of parallel chant phrases or formulæ that is wrong: though we use comparative analysis, we should not presume that a given phrase is always rhythmically consistent, for it is clear that the cantors and conductors delighted in variety. Cardine ended his essay by equating equalist with free rhythm, strongly implying that proportional rhythm is not free, which is incorrect. Cardine's interpretations are described in his Semiologia Gregoriana (Rome: Pontificio Instituto de Musica Sacra, 1968): equal-length, non-divisible notes are nuanced, in an æsthetic not far removed from that of Solesmes' Dom André Mocquereau a hundred years ago. In their approach, sensitivity to textual accent may be achieved only through subtleties of singing that can hardly be communicated or reproduced, a situation most unsuitable for liturgical song.

Argument in favor of the divisible long as the ordinary note is found not only in the shapes of the neumes, but in the writings of several medieval theorists. While Calvin M. Bower in his article "The Grammatical Model of Musical Understanding in the Middle Ages" (In Hermeneutics and Medieval Culture; Albany, 1989) recognizes that medieval writers stress the importance of lengthening notes at the ends of phrases, he ignores the fact that they are also concerned about the lengths of individual notes within those phrases. For example, the Commemoratio brevis, written around 900, states that

all the longs must be equally long, all the shorts of equal brevity; the only exceptions are the distinctions [phrase-endings], which in the chant must likewise be observed with care. Everything of long duration must rhythmically concur with what is not long by legitimate and reciprocal durations… And in accordance with the length durations let there be formed short bets, so that they be neither more nor less, but one always twice as long as the other.

David Hiley's massive study Western Plainchant (Oxford: Clarendon, 1993) stands as the epitome of an approach championed over the last 150 years by the Abbey of Solesmes, beginning with Abbot Prosper Guéranger, Joseph Pothier, and André Mocquereau and ending with Eugène Cardine. Following most university scholars today, however, Hiley is less concerned with the evidence of neumes in the earliest extant manuscripts than with abstract questions regarding oral dissemination of chant and the interaction of Western chant traditions. Richard Crocker in his lengthy article "Gregorian studies in the twenty-first century" (Plainsong and Medieval Music, 4, 1, pp. 33-86; Cambridge, 1995) does little more than continue this new scholasticism.

We need to return to those wonderful manuscripts, to look now mainly at the shapes of the neumes themselves, compare them with one another and with the later diastematic sources, and arrive at intelligent realizations that we can sing. But this time we must examine the neumes without any predispositions that would blind us. For example, the lack of weighted accent in the French language has kept Solesmesian theorists from seeing what an absurd situation it is when, in equalist rhythm, two notes fall on an unaccented syllable and only one note falls on a neighboring accented syllable, since weight is thereby taken away from the accented syllable and placed upon the unaccented. Again, the monks' political need to come up with publications that could be used liturgically throughout the world encouraged them to think that there must have been some single written archetype and that the best manuscripts should be perfectly correlatable. Finally, the prejudice that equates free rhythm with equalist must be abandoned.

In Charlemagne's words, "We also, then, who till now drank from the troubled waters of the stream, must go back to the clarity of the source." For us, this clarity lies in the neumatic 9th-century fragments and 10th-century manuscripts, with their æsthetic of deep respect for poetic words.

R. John Blackley

This article first appeared in the spring 1996 issue of the Japanese quarterly Worship and Music; it is dedicated to Bernard Freiland and John Coggins.


Rhythm and Nuance in Chant

The controversy concerning rhythm in chant is far from over. The semiological interpretations advanced by Dom Eugene Cardine were the late attempts of the Abbey of Solesmes to justify what remains at base an equal-note approach to chant, an approach for which no trace may be found before the 11th century. That a basic rhythm of freely arranged longs and shorts in 2:1 proportion was proper to chant until well into the 10th century has been demonstrated clearly in the writings of Jan Vollaerts, S.J., Dom A. Gregory Murray, and the Schola Antiqua. Their developing arguments, based upon the shapes of the neumes in the 9th- and 10th-century manuscripts, the writings of medieval theorists, and the non-conformity of equalist rhythm to Latin accentual patterns, have not been disproved and in fact stand firm; and it is through these sources that a fairly unified view of mensural chant has evolved.

Followers of the Solesmes school say that their method is not equalist, but nuanced. Yet nuancing is not a rhythm--one can only nuance some rhythm that already exists. A rhythm of nuances is no more possible in real life than the smile without the cat: a nuance graces a given rhythm as a smile graces a creature, both assuring an ease of interpretation.

There are two basic kinds of rhythm in western music, free and regular. Regular rhythm may be syllabic (in early chant hymns), or of recurrent patterns (for example, in modal rhythm and isorhythm), or metrical. Free rhythm is also of three types: syllabic (in psalm-tones), proportional, and equalist. Both a bifolio flyleaf to MS Laon 266 (in Metz notation) and the whole of MS St. Gall 339 (in St. Gall notation) contain fully sophisticated proportional-rhythm neumes unaccompanied by any indications of rhythmic nuance. Of course, the singers would have nuanced rhythmically their singing of the proportional chant, and indications of such indeed appeared in the form of letters in both Metz- and St. Gall-notation sources (t = tene = hold, c = celeriter = quickly, etc.). Judging from their presence in a bifolio flyleaf fragment within MS Laon 121 and in the cantatorium St. Gall 359, both dating around 900, and from their absences in 266, c. 870, and in the mid-10th-century 339, one cannot reasonably place their origin any earlier than the end of the 9th century, significantly later than the introduction of neumes. (In his "Charlemagne's Archetype of Gregorian Chant" [JAMS, spring 1987, pp. 1ff.], Kenneth Levy demonstrates the strong likelihood that chant was notated by the late 8th century.)

It is reasonable, then, both because of the nature of nuance and rhythm and from an overview of the neumatic manuscripts, that the importance of the rhythmical letters is secondary to that of the proportional rhythm in the neumes themselves. The fact that, for any given chant piece, there is remarkable proportional-rhythmic agreement among its neumes in various manuscripts and in different notations, but no such clear agreement regarding the rhythmic nuancings, serves dramatically to illustrate this argument. The rhythmical letters were meant to direct (though not autocratically or exclusively) the expression of the singers, much as the careful rhythmic and dynamic signs in a score by Webern direct modern performers.

Semiology nuances lengths that are, at their base, the same, and so Solesmes' chant remains just as equalist in orientation as it always has been; only the clothing is a little different. But, as has been and can further be demonstrated, the shapes of the neumes themselves describe proportionally measurable longs and shorts. What, then does one nuance? Logically, one follows the clear indication of the 10th-century manuscripts and gives nuance to the proportional longs and shorts. For summary, a classification of rhythm in outline:

A. Free

1. Syllabic (psalm-tones)

2. Proportional

    a. straight (Laon 266, St. Gall 339)

    b. nuanced (Laon 121, Laon 239, St. Gall 359, Einsiedeln 121; approach of the Schola Antiqua

3. Equalist

    a. straight (Vatican Edition)

    b. nuanced (Solesmes editions; approach of the semiologists)

B. Regular

1. Syllabic (early hymns)

2. Recurrent Pattern (modal rhythm, isorhythm)

3. Metrical

The chant of the 9th- and 10th-century neumatic manuscripts was not, as might be assumed from the various Solesmes methods, a music ethereal in style or essence. Such longing for a beauty that is Gothic, redolent of spired churches, moving with grace from assured minds to trusted heavens, stems from habit, and understandably, but not from facts and history. History, along with honest attempts to sing the chants as they appear in the earliest manuscripts, discloses a song more robust in stature: and the long notes that bear its melodies are founded on, nestled in the earth just as soundly and nobly as any Romanesque structure.

R. John Blackley

This article first appeared in Fanfare, the record review.