3. The rhythm of Gregorian Chant

‘Het ritme van het gregoriaans’ [The rhythm of Gregorian Chant], Tijdschrift voor Gregoriaans 30 (2005) 15-22, 49-53, 82-88.

The article in Dutch you can find here as pdf (888 Kb)

An English Translation by Kevin M. Rooney of the complete article can be found in Rhythm, Meter and Tempo in Gregorian Chant (2016), Lancelot Andrewes Press.


The so-called semiological interpretation of Gregorian Chant, as developed by Cardine and his followers, was a great step forward from the previously customary ‘Solesmes’ style of singing. I think however that we must go some steps further if we desire to recover the meaning and the beauty of the original Gregorian Chant, in particular as regards the connection of tones, the ornaments, and the rhythm. During the seven years that I led a Gregorian choir I was able to work out and to test my ideas in practice.

Regarding the tone production, I think that we approach the intention of the neum writers more closely if we listen not only to the tones themselves but in particular to the transitions, the movement from one tone to the next. As we can observe in other vocal monophonic repertoires as well, glissando-like transitions will play a major rôle. It is characteristic of Greek church music, in which we can frequently observe this phenomenon, as well as of Byzantine church music, not to indicate the individual tones of a melody, but only the intervals between the tones.

Regarding the ornaments, the ‘semiological’ interpretation has drawn our attention to the ‘gleitende Tonansatz’ [gliding attack] of the ‘quilisma-pes’, the ‘pes initio debilis’ and the ‘torculus initio debilis’. I have shown that it is likely that besides these neums, which are in fact identical with respectively virga’s and a clivis approached by a short lower auxiliary tone, other neums as well begin with such auxiliaries. Further I have shown that the musical and textual context in which an oriscus often appears makes it likely that this sign indicates a tone that is preceded by a short upper auxiliary. In Greek church music this is one of the most frequent ornaments.

Regarding the rhythm, it is my opinion that the desire among the semiologists to avoid any and every form of ‘mensuralism’ is based on a prejudice. There are good arguments for the assumption that there in fact were originally various categories of note values, and not only a broad range of ‘nuances’. I refer to a) relevant statements by medieval theoreticians, b) the fundamental difference in form of the uncinus and the punctum in the notation of Laon, and the generally disjunct notation of  ‘nicht kurrente’ [non-running] neum elements as opposed to the conjunct notation of  ‘kurrente’ [running] neum elements in the same manuscript, and c) the fact that a ‘mensural’ Gregorian Chant then falls into line with genetically related vocal monophonic repertoires like Byzantine, Greek, Syrian and Coptic church music.

We can use the neum tables of Göschl and Agustoni in their well-known book on the semiological interpretation of Gregorian Chant as the point of departure for the rhythm:  the ‘nicht-kurrente’ notes give the value of one chronos, one unit of time (longae), and the ‘kurrente’ notes give the value of half a chronos, a subdivision value (breves). A number of corrections is however necessary. In the first place it must be observed that not only the short form of the salicus/scandicus, but also the short form of the torculus, porrectus and climacus have the rhythm brevis-brevis-longa. This is manifest from the notation of Nonantola. In Greek church music this is also the normal rhythm of three-tone syllables. If we leave out of consideration the occasional short ‘introductory syllables’ with the value of half a chronos at the beginning of a phrase, then the one-, two- and three-tone syllables have a rhythm that brings their total duration to an integral number of chronoi. This is applicable to a large number of four-tone neums as well. Only syllables that seem to have the rhythm brevis-brevis-brevis-longa would be the exception. Closer analysis however leads to the suspicion that the first tone is then a short grace note:  we can almost always interpret these neums as one of the three-tone neums salicus, torculus, porrectus and climacus, preceded by a short auxiliary. For the porrectus plus short auxiliary this interpretation is in a number of cases demonstrably correct. Five-tone syllables appear to behave similarly. Apart from the short introductory syllables mentioned earlier – which we also encounter in Greek church music – and from the ornaments, there is thus in Gregorian chants that limit themselves to the use of one- to five-tone syllables, a continual movement in longae. We can consider the longa a beat, exactly as in Byzantine and Greek church music.

The rhythmical interpretation of longer melismas is a complicated affair. But it is possible to formulate rules for compound neums, and a satisfactory solution can usually be found. J.W.A. Vollaerts published trailblazing work on this matter in 1960. My article gives a number of examples of what we may expect to encounter in melismatic chants. In the case of longer melismas with many subdivisions the movement in longae frequently seems to be interrupted. Was there at the time of the earliest musical notation of Gregorian chant uncertainty as to the original rhythm?

Although we often encounter more or less extended passages in Gregorian chant that seem to have a basically binary meter, this regular meter is less fundamental than in Byzantine and Greek church music:  see the ‘mensural dashes’ under the staves in the musical examples.