Comments on Gombert's Ave Maria


      by Peter Urquhart

Nicolas Gombert (c.1495-c.1560) was a composer who consistently chose to pit linear against harmonic integrity. In his music, one frequently cannot satisfy all of the so-called "rules of music ficta" at the same time; either one or another of these dimensions must fail. The choice of which element to favor or to sacrifice relates to the most basic decisions about performing this music: what is most important, what is secondary, and how did singers sing from the printed notation in part books.

In 1990, the ensemble Ars Nova, directed by Bo Holten, released a compact disc of works by Gombert, perhaps the earliest recording devoted solely to the composer's music. It was an adventuresome and thought-provoking release, for the director was not about to be tied to the somewhat vague accidental suggestions provided in the Collected Works edition; Holten wrote in his liner notes:

The question of Musica Ficta, the adding of accidentals into the fabric of polyphonic music of the renaissance - done by the singers!, is especially tricky in the music of Gombert. The rules of Ficta seem to be in constant mutual conflict in his dense and complex textures, thus making "correct solutions" impossible. The accidentals added in these recordings are the result of a fairly consistent policy of interpreting the often self-contradicting rules, in combination with opinions of convenience on the part of the singers, developed through many performances (Kontrapunkt 32038, 1990)


One motet on the recording, a five-voice Ave Maria, contains an imitative section that obsessively turns over the following figure for twenty measures, primarily at two pitch levels, either starting on D and G. The ensemble's decisions in this passage well reflect the difficulties of "constant mutual conflict" in applying these rules.

The initial figure used in the imitation is revealed in the cantus at m. 54; the ensemble Ars Nova chose to inflect the second note with a sharp, probably in order to inflect the suspension between the cantus and tenor, thereby making a cadence to D (Ex. 1). The contour of this figure was then reproduced in the following voices at the pitch level G with accidentals on F# and Eb; these inflections work well at first in the counterpoint and produce their own cadential approaches to D in m. 56 and 57.

Ex.1: Ave Maria, mm. 54-58


However, the use of these inflections with the figure cannot be maintained. The figure on D occurs most often with C#, but occasionally Cnatural is needed to avoid problematic harmonies. The altus avoids using C# in the figure on D in m. 56 in favor of allowing the lower voice to make the cadence with Eb; in the next measure, the tenor does the same. The figure on D thus comes in essentially two forms, as shown in Ex. 2; the third measure shows the slightly elaborated form of the figure in the altus and tenor beginning in m. 55.

Ex. 2: Two forms of the imitation figure on D


The figure on G, because of the two pitches that might or might not be inflected, comes in four different forms in Ars Nova's performance (Ex. 3). The initial form is favored because it matches exactly the intervallic shape of the original figure given by the cantus voice in m. 54. But the ensuing inflections are adjusted to suit the harmony, which turns out to be the overriding consideration in Ars Nova's performance throughout the work. All vertical imperfect 8ves that might be caused by the adoption of the initial inflection have been avoided, and all substantial imperfect 5ths have likewise been corrected, at the expense of any consistency in the linear presentation of the figure.

Ex. 3: Four forms of the imitation figure on G


The quintus line (Ex.4) provides a good example of this linear inconsistency, when seen by itself.

Ex. 4: Quintus line, mm. 55-63


The middle phrase here is shorn of the inflections given to the outer phrases; the reason for this change in its melodic contour is the harmonic context, solely. In m. 58 the F of the tenor is made to agree with superius above it; the same agreement is created in m. 59 between the Enaturals, even though the shape of the inflected figure is thereby lost (Ex. 5). If Ars Nova's performance is meant to represent an interpretation of performance practice, the question that comes to mind is, how would the singer have known when to inflect one way or another? These inflections all require score-knowledge, and not a small amount of study.

Ex. 5: Ave Maria, mm. 57-61


It may be suggested that the cantus could be altered in mm. 58-59 to fit the quintus rather than the other way around; after all, the quintus here has the imitative figure, while the cantus material is only connective, not motivic. However, Ars Nova has sensibly avoided this solution, probably because of the violence it might do to the cantus line, which , if the tenor maintained the initial contour, could be forced either into an erratic sounding of the pitches F or E, and Eb to F# motion that is close or even direct (Ex. 6).

Ex. 6: Cantus, mm. 57 to 60


On the other hand, such extreme choices of inflection from a melodic point of view are present in the Ars Nova performance elsewhere, always for the purpose of ensuring the smooth harmonic surface. In m. 60 (Ex. 5 above), the sudden switch to Eb in the tenor is for the purpose of creating a perfect 5th where otherwise a diminished 5th might occur. Similar decisions are made elsewhere in the performance of the piece, indicating that the perfection of harmonic diminished 5ths was valued more than the linear perfection of vocal lines in this performance. Ex. 7 is typical, where the decision to perfect a 5th between bass and tenor forces the cantus into an inflected form unthinkable from a melodic point of view.

Ex. 7: Ave Maria, mm. 81-84


I do not wish to criticize Ars Nova's performance decisions in this motet too harshly, for theirs was the first performance of the work, and a fine example of working out the results of a "consistent policy of interpreting the often self-contradicting rules." Despite the difficulties of line imposed on the singers by the interpretation, the performance is a good one; the question is, is it still Gombert? It is important to point out that their editorial choices well reflect the standard assumptions about "musica ficta" held by most editors over the past 30 years. Ars Nova's performance shows that such assumptions--here, the adjustment of line to suit harmonic goals--threaten to destroy the character of Gombert's work. Not only are these adjustments impractical to expect from singers reading from their own lines, but they also hide the imitative beauty of Gombert's counterpoint, and his very particular harmonic color. This last claim is nearly impossible to investigate in a climate where such changes can be wrought on the composer's lines under the rubric of "musica ficta." The fifth interval pointed out in m. 82 (Ex. 7) is a case in point; if Gombert were interested in creating a diminished 5th in passages such as these--and the cantus line suggests he may well have been--our traditional understanding of the mi-contra-fa rule as a performer's perogative prevents us from appreciating the sonority.

The use of the diminished fifth by Gombert will be addressed more fully in the next chapter; for now, it will suffice for us to reconsider the inflections added to the initial motive of the passage from mm. 54 to 72. A more singerly and less harmonically oriented approach to the passage would accept the initial motive as it lies, without the inflection on the second note. Because the cantus continues melodically past the suspension of m. 54 (Ex. 1), the second note does not demand cadential inflection when seen from the point of view of the cantus: Similarly, there is no linear reason to inflect the 4th note to Eb when the figure begins on G; the reasons mentioned before--to maintain exactly the same melodic contour and solmization, and to make a cadence with the upper voices at mm. 56 and 57--are not linear reasons for inflection, and in my opinion do not necessarily apply. As shown before, the use of the sharp and flat inflections on the second and fourth notes of the figure result in even greater unpredictability of melodic contour because it cannot be maintained. The simpler approach is to sing the figure in just two forms, depending on whether it begins on D or G, either . The cadence inflections are better restricted to just those phrases where cadences are inevitable from the point of view of the line, such as measures 71 and 75 in the altus, and possibly the cantus at m. 66. The entire passage accepts these recommendations readily, and is free of any serious mi-contra-fa clash.

Click here to see the passage from m. 55 to the end, and to compare interpretations in sound: Ex. 8: Ave Maria, end


The final cadence of the motet presents the harshest side of Gombert's harmonic style. There can be little doubt that mm. 89 to 90 present the final cadence, set up between the cantus and tenor in the most traditional way. The quintus line, however, interferes in the most insistent manner, with both the expected cantus F# inflection, and the altus Eb in m. 89. The editor of the Opera Omnia, Joseph Schmidt-Görg, has allowed this interaction to affect inflection of this final cadence; he refrained from adding the F# to the cantus, and he corrected the quintus line's E to Eb. Bo Holten's decision to allow the final F / F# clash is the right one in my opinion, and a very common gesture in music by Gombert, L'Heritier, Clemens, Crecquillon, and other Northern composers of the "post-Josquin" generation.

-Peter Urquhart

For more on the style of Gombert and contemporaries see my article: "Cross-Relations by Franco-Flemish Composers after Josquin."  Tijdschrift van de Vereniging voor Nederlandse Muziekgeschiedenis, 43 (1993): 3-41.
peter.urquhart@unh.edu