4. The rhythm of Latin hymns

‘Het ritme van de Latijnse hymnen’ [The rhythm of Latin hymns], Tijdschrift voor Gregoriaans, 32 (2007), 147-151, 33 (2008), 25-29.

(This article is a revision of the notes at the end of the book Hymnen, Tournai 1967, which was written with J.W. Schulte Nordholt.)

The article in Dutch you can find here as pdf (253KB).


Regarding the rhythm of Latin hymns I have closely followed the detailed viewpoint of Ewald Jammers. From 17 to 19 August, 1965, I visited him in Heidelberg to learn more about his views. I had already been very impressed by the fact that he had managed to identify the oldest hymn melodies of the Milanese and Cistercian tradition, connected with metric texts of Ambrose and probably dating from his time, by comparing their characteristics with those of hymn melodies from a later time, connected with non-metrical, rhythmic texts. (See Ewald Jammers, ‘Rhythmische und tonale Studien zur Musik der Antike und des Mittelalters’ II (Auf dem Wege zum Mittelalter), Archiv für Musikforschung 8 (1943), 27-45.) The oldest melodies are composed from melodic units corresponding to a metrum, a half verse; in those melodies we find disjunct motion mostly between the metra, sometimes also between the feet (the iambs), while melismas are found mostly on the long syllables of the iambic dimeter. In later melodies the whole verse is a melodic unit, disjunct motion is found mostly before accented syllables, and melismas mostly on syllables before an accented syllable. The metrical verse structure leads easily, in the early melodies, to the assumption of an original ternary rhythm, and for the later melodies to the assumption of an original binary rhythm:  the necessity of a ternary rhythm is not present in the rhythmic verse structure. I found a nice proof for a mensural rather than the later customary equalistic performance (performance with more or less equal rhythmic values) of the hymns of Ambrose in the melody of Intende qui regis Israel. When the understanding of the metric character of the text had disappeared, the words to the melody, apparently still sung in ternary rhythm, have been shifted so that the text accents fall on the downbeat.

Although after ca. 800 a binary rhythm will have been normal for hymns, it is not implausible that the old metric hymns kept their original rhythm, and were simply more ornamented. There is an indication that in Milan the Ambrosian hymns were still sung in a ternary rhythm even in the beginning of the 20th century. In the course of the 12th century we must assume for the hymns and sequences an influence of the modal, ternary rhythms that spread from Notre Dame in Paris . In the 14th century the existing hymn melodies had become largely equalistic, i.e. each tone had the same length, independent of its structural or ornamental character. This apparently happened under influence of Gregorian chant, that already around the year 1000 had lost its original rhythm. Only a few melodies, such the popular Conditor alme siderum, seem to have escaped this process.

If we place the variants of one and the same melody from the various manuscripts next to each other, we can observe which notes must be interpreted as ornaments, developed from the manner of singing, and which are more structural. Passing tones and neighboring tones could be added, as well as anticipations and suspensions. The last two types, in which at the end of a syllable the (first) tone of the next syllable is already reached, or in which the beginning of a syllable repeats the (last) tone of the previous syllable, demonstrate a preference for ‘Verschleifungen’, for a portamento manner of moving from one tone to the next. One must be careful not to sing separate tones, but to join them, to interpret them as a unit of movement.