Gregorian Rhythm Wars contains all previous installments of our series.
Please refer to our Chant Glossary for definitions of unfamiliar terms.
S WE APPROACH THE NINE-MONTH MARK IN THIS SERIES, I wish to reiterate the position I have defended all along: the oldest extant manuscripts show the chants with long and short notes in 2:1 proportion, which fit into the framework of a steady beat (not to be confused with a meter or time signature), contrary to what nearly everyone in our era has been taught. In his fourth response to me, Jeff wrote, “Patrick, you correctly noted that 339sanGall|1039 does not match the others, but you give no explanation for this.” Is that so? In my fourth reply, I noted that St. Gall 339 (G) shows exactly the same form of the clivis at observaberis, sustinebit, and propitiatio, mentioning that it “is typically dated as the latest of the five sources”—Jeff’s article dates it 73 years later than Laon—and commenting later that:
In fact, by the eleventh century, we see not “mind-boggling correspondence” regarding the rhythmic indications, but confusion, which soon degenerates into cantus planus without differentiation between long and short notes; indeed, such confusion is already somewhat evident in St. Gall 339.
In my third response, I quoted Aribo, who wrote in ca. 1070 that the idea of composing and singing proportionally had “already been dead for a long time, even buried.” Peter Wagner, citing the Musica enchiriadis, notes the evidence for the slow performance of organum (harmonized chant), requiring a slow rendition of the chant itself (Einführung in die gregorianischen Melodien, vol. 2, 1912, pp. 370–371). Anyone claiming that “the ‘authentic rhythm’ became corrupted or forgotten because scribes ‘could no longer properly write adiastematic neums’” is confused. Does anyone claim such a thing, or is this another straw man? It’s akin to claiming that the rhythm of Old Hundredth (Geneva 134) was corrupted because the publishers forgot how to print mensural notation. The fact is, the rhythm was corrupted because of the ponderous tempo and equalization of note values—the same things that happened to chant five centuries before. Is it plausible that someone today understands the rhythm of the 1565 version better than someone in 1584? Well, compare the 1565 and 1584 versions to see for yourself:
1565:
1584:
Is it plausible that someone 500 years from now might have a better understanding of the operation of a rotary telephone from reading about it than today’s teenager, who is only a few decades removed from when that technology was in common use? I rest my case! I will now get back to the business at hand and analyze a couple of chants according to the oldest extant sources.
Another Introit Analyzed • Jeff has introduced the introit Exaudi Domine . . . adjutor into the discussion, so let’s examine it:
Summary of note values (antiphon only):
red – long in both sources = 45
purple – long in at least one source = 4
brown – long in a source other than L or E = 3
blue – long by interpretation = 4
green – short in both sources = 46
yellow – short by interpretation = 2
total = 104
Laon 239 (L) is generally considered the oldest extant source for this chant. In comparing Einsiedeln 121 (E), we see that the two sources are in unambiguous agreement for 98 of the 104 note values, marked in red, brown, blue, and green. I interpret L’s c at adjutor as applying to both notes, which together correspond to E’s pes rotundus. E’s pes quassus at the end of derelinquas can be interpreted as two long notes without in any way contradicting the literal meaning of the sign. The c at the unison salicus in salutaris can be read as applying to the oriscus rather than the virga. That gives us agreement for 101 note values. Of the remaining three notes, all on the first syllable of neque, the two yellow notes are ambiguous and could also have been colored purple; L and E could be read there as agreeing with each other, without forcing the interpretation. That leaves the first note of neque as the only outright contradiction between the two sources. I think you’ll agree that a correspondence of greater than 99 percent is excellent. E writes eight virgae, but only the one at Deus has an episema. I would hold that note a beat and a half, not on account of the episema alone, but because of context; L and E are in literal agreement here regardless. In this particular chant, L’s cursive clivis always corresponds to E’s clivis with c. Apart from the psalm verse, E writes a plain cursive clivis without c only once, at adjutor. Jeff could hardly have proposed any other chant where the neumes are written with such nearly impeccable consistency. As for the three notes marked in brown, which are not explicitly long in L or E, our first observation should be that the one at the end of Deus is followed immediately by a bar line and dotted in the Solesmes editions; therefore, there seems to be no objection to the long interpretation of it from the partisans of the Solesmes and pure Vatican equalist interpretations. It makes little practical difference to the singer or listener whether a note is lengthened because of a neumatic sign, an episema, the letter t or a, a dot, a bar line, a melismatic mora vocis, or tying two short notes together. The last note of the porrectus at despicias is marked with an episema in St. Gall 374, and the entire torculus at -ta- is long in St. Gall 376—for whatever those relatively late sources are worth. I would be reluctant to rely too heavily on them, but Jeff has already placed them on an equality with L, E, and every other manuscript, despite devoting most of his previous article to demonstrating that various St. Gall manuscripts contradict each other. By failing to differentiate between rhythmic and non-rhythmic manuscripts and by appealing to cursive neumes without an episema from later and less reliable sources to disprove “nuances” from earlier and highly reliable sources, it seems he wants to have his cake and eat it too!
The Parable of the Singer • A singer was given a printed page of song lyrics, without any musical notation. Although she didn’t have it completely memorized, she was already familiar with the song and only needed a few reminders here and there. Just to be safe, she penciled in all the note letter names and then made a handwritten copy of the lyrics with chord symbols for her accompanist, who was also familiar with the song but didn’t know it by heart. The singer was asked to sing the same song a year later and used her little musical cheat sheet again. She never found a published version but eventually had the song totally memorized and gave her copy of the lyrics with note letter names to someone else. Years passed and nobody found a published copy of the song, but the cheat sheet was photocopied many times and circulated widely. Remarkably, nobody ever recorded it until many centuries later. Since the song was only sung one week every year, people started to forget the exact rhythm. Over the course of years, decades, generations, and centuries, people continued to sing the song annually, and it was published a number of times, with quite a few variants. At some point, a photocopy of an entirely handwritten version with chord symbols turned up, where someone else had written in the rest of the melody notes. A millennium after that lady made her two cheat sheets for herself and her accompanist, someone rediscovered a copy of the songbook her grandparents used in grade school, 101 Songs to Sing. The song was in there, with the same notes, but the rhythm was very different from how people had been singing it for a thousand years. Two years later, a new volume was published, The Greatest Songbook of All Time, which purported to include a definitive edition of that song and hundreds more, with tens of thousands of copies sold worldwide, but the melody was notated entirely in eighth notes, except for quarter notes at the ends of phrases and a few other spots where that lady had separated her letters wider apart than elsewhere. A year later, a facsimile of 101 Songs to Sing was published in a musicological journal, but many scholars dismissed it as unimportant not only because most of the songs had been sung another way for many centuries, which they considered to be the “traditional” version, but also because nearly all of the songs in 101 Songs to Sing were thought to have been at least a century old already when it was published, and there were fragments of older editions of some of the songs to prove that theory. Another 70 years later, the publisher of The Greatest Songbook of All Time released a companion edition that included the versions from 101 Songs to Sing and the handwritten cheat sheet alongside the “definitive” version.
Write What You Mean • Although the parallels may be imperfect, I hope my parable serves to illustrate the concurrent change to diastematic notation that Jeff mentions. Many of us have created our own musical cheat sheets at some point, but very few have filled an entire notebook with them. They are sufficient as memory aids for someone who already knows the song but are of limited use to someone who doesn’t know how the song is supposed to go because they lack rhythmic indications; one might as well sing or play it in straight eighth notes throughout or make up the rhythm according to one’s own tastes. If the rhythm has already been evened out, a non-rhythmic notation that is melodically precise would seem like an improvement over a notation that is rhythmically precise but melodically equivocal. This has nothing to do with scribes no longer understanding how to write the older notation, but it doesn’t make much sense to continue differentiating long and short notes on the page once they’ve all become equal in performance. As I have already said at least three times in this series, the later manuscripts have nothing to add to the rhythmic indications of the neumes copied in the triplex editions, which are taken from the most ancient relatively complete sources, and I don’t know how to be any clearer about this. For the record, I stated all the way back in December that “I wish to withdraw from further analysis of those sources [from the second half of the eleventh century and later] and allow the other contributors to have their say.” The later manuscripts generally cannot be considered a commentary upon or correction of the oldest sources. As Jeff has acknowledged, many manuscripts don’t differentiate between long and short forms of the clivis at all. What about the torculus?
Being Sensible • Four notes have been marked in blue as long by interpretation, each one of them at the end of a cursive torculus. Now that the evidence has been examined, I would like to appeal to reason to complete the big picture. Dear reader, I ask you to ignore the paleography for a moment as well as anything you’ve read from Pothier, Mocquereau, Gajard, Vollaerts, Murray, Cardine, Blackley, Van Biezen, Ostrowski, Weaver, me, or anyone else. Does it make more sense to interpret those four notes in blue as long, resulting in 80 binary beats (not counting the double-long notes and rests at bar lines), or short, resulting in 72 binary beats and four ternary beats? Perhaps common sense will be sufficient to convince you of something nobody can prove from the evidence of this particular chant alone, namely that the ordinary rhythm of the cursive torculus is short-short-long. The binary nature of the remainder of the chant is displayed in color before your very eyes. Now all you have to do is connect the dots for yourself.
Another Example • I have also analyzed the offertory Exaltabo in like manner and uploaded a recording to YouTube (updated 8/8/23). Despite the length of this chant, the rhythm is remarkably straightforward:
red – long in both sources
purple – long in at least one source
brown – long in a source other than L or E
blue – long by interpretation
green – short in both sources
black – lacking in one source
This chant has a total of 461 notes. According to my interpretation as presented in Thirteen Offertory Chants, which normally gives preference to L, it has 361 binary beats, not including the repetition, and also not including the double-long notes and rests at bar lines. If the six notes in blue (five of which are at the end of plain cursive torculi) are interpreted as short, it has 349 binary beats and six ternary beats. Which interpretation of those six notes makes more sense musically?
Nuances and Clarifications • Jeff wrote, “How can we explain the discrepancies? I have suggested that many of the so-called ‘rhythmic’ indications were probably nuances. Therefore, when scribes ignore, jumble, or modify them, it’s no big deal.” Jeff, I challenge you, just as I challenged Charles: Where does any medieval theorist write of rhythmic nuances for the episema or ordinary long note (tractulus, uncinus, or virga) that are somewhere between single and double in duration? Can a single shred of evidence be produced, or is it just a made-up theory? The ad hominem attack you made was not that my claims are untrue because proportional rhythm is not widely used, but rather that you considered that it might be improper to continue discussion with me because I lack the courage of my convictions, since I allegedly don’t use proportional rhythm at all with my own choirs—which is not only fallacious, but patently and demonstrably untrue. I don’t demand an apology (this is a war, after all!), but know that I will continue to call you out on these kinds of things. Charity toward our readers demands accuracy and honesty. If your cursive clivis comparison chart was made in response to my question, “Why not take the time to compare a few sources carefully, note for note?” it should have been obvious from context that I was referring to correlation of pitches, but I appreciate the effort, even though you omitted what I and others believe to be not only the oldest but also the most reliable sources from your comparative table of manuscripts spanning approximately 169 years. One must question the exclusion of Laon 239 and Einsiedeln 121 from a chart including sources dated to as late as ca. 1074. In this series, I have already answered your concluding question, “What evidence is there that the stroke known as an episema denotes a longer note?” (slightly paraphrased), but I will be happy to restate my answer yet again after you answer a couple of the questions I asked previously, namely: 1. Is it “miraculous” that Old Hundredth (Geneva 134) is sung with the same melody today as in 1551? 2. Do you believe that there was a conspiracy among printers to suppress the authentic rhythm of the chorale melodies, or that the Protestant congregations suddenly forgot the original rhythm? And one new question: 3. What does an analysis and comparison of the 1791 and 1854 versions of Old Hundredth reveal about the 1565 version?
1791:
1854:
1565:
A Personal Note • In my previous reply to Jeff, I wrote, “it appears quite possible that . . . we are reaching the point of merely repeating ourselves ad nauseam.” Let us, however, recall the Latin axiom Repetitio est mater studiorum (“Repetition is the mother of learning.”). Perhaps Jeff is right that we shouldn’t call things off just yet. With that said, my contributions to this series have become very time consuming in recent weeks, and I will not be able to continue churning out articles at the same rate. The liturgy doesn’t take a summer break, and neither do my adult choirs. My workload doesn’t diminish considerably over the summer, which means that I have to sacrifice some of the time I ought to be working on my Cantatorium editions in order to research and write articles. As we move forward, I will contribute to this series no more than once a week, posting only on Sunday or Monday. This will also help to avoid monotony for our readers, lest it appear that Corpus Christi Watershed has been taken over by the discussion of Gregorian rhythm. I will post blog articles on non-Chant Rhythm Wars topics whenever I have one ready. With discussion of Office antiphons and ongoing examination of second-millennium manuscripts, we are drifting from my area of expertise, which is the Proper of the Mass according to the oldest extant sources. In my opening contribution to this series, I provided a modern notation transcription in quarter and eighth notes. At the beginning of my follow-up article the next day, I said as clearly as possible that I mean that notation to be taken literally, with notes in 2:1 proportion. I’ve been surprised at the inability or difficulty so many people have in grasping this simple, straightforward concept of proportional rhythm. People who read music and know the Solesmes style of chant have said they don’t understand most of what I’m talking about. Others have commented that proportional rhythm actually seems more nuanced to them than the Solesmes method (!). Others have said that they think what I propose would be more difficult than the Solesmes method for an ensemble to sing together. Still others have commented on how interesting it is that I “added” rhythm to the chant. They don’t have a clue. I feel the same frustration Vollaerts and especially Murray must have felt; even when examples are provided in perfectly unambiguous modern notation, along with the graphical evidence of the adiastematic neumes, people still don’t “get it.” Maybe I’m just a lousy teacher, as Jeff suggests (another ad hominem?). I have followed the formula he mentioned of saying what I’m going to say then saying it. Now I would like to close by saying what I said, not in this article alone but in the entire series. IT HAS A STEADY BEAT, AND THE LONG AND SHORT NOTES ARE IN 2:1 PROPORTION. Sorry for yelling! St. Chrodegang of Metz, pray for us.
Cantatorium.com edition, Fifth Sunday after Pentecost:
