Notes on Graduals in Modes 2 & 8

Tracks 124: mode 2
Tracks 2527: mode 8


The melodies of Graduals can be found in the “Vatican Graduale,” Graduale sacrosanctae romanae ecclesiae de tempore et de sanctis SS. D. N. Pii X. pontificis maximi jussu restitutum et editum (Romae: Typis vaticanis 1908); and in the Graduale triplex, seu Graduale romanum Pauli PP.VI cura recognitum & rhythmicis signis a Solesmensibus monachis ornatum neumis Laudunensibus (cod. 239) et Sangallensibus (codicum San Gallensis 359 et Einsidelensis 121) nunc auctum (Solesmes 1979).

The “signs of nuance” can be studied in the facsimiles of the chant books “Laon 239” (“L”), published in Paléographie musicale X (1909); “St. Gall 359” (“C”) in Paléographie musicale IIe serie, II (1924); and most conveniently in the Graduale triplex, where the notation from L and C (and also “E,” Einsiedeln Cod. 121) has been entered by hand by Marie-Claire Billecocq and Rupert Fischer. Here, references are given to the Graduale triplex (“GT”).

Liturgical assignments are as in René-Jean Hesbert, Antiphonale missarum sextuplex (Brussels 1935); reference is to the number of the Mass formulary, preceded by “M.”

The standard work on the Gregorian chant repertory, including bibliography, is by David Hiley, Western Plainchant, A Handbook (Oxford: Clarendon Press 1993). The Graduals are discussed on p. 81.

See also The Early Middle Ages to 1300, in The New Oxford History of Music, Vol. II (new edition, Oxford 1990), ed. R. Crocker and D. Hiley, p. 203.

The nuance notation is discussed in Eugene Cardine, Gregorian Semiology (Solesmes 1982), trans. by R. Fowels from “Sémiologie grégorienne,” Etudes grégoriennes 11 (1970) pp. 1–158.

Modal finals and terminal descents are discussed by Jean Claire in “Les Répertoires liturgiques latins avant l’octoechos. L’Office ferial Romano-Franc,” Etudes grégoriennes 15 (1975), pp. 5–192.

For a representative detailed study of the model melody, see  Bernard Ribay, “Les Graduels en IIA,” in Etudes Grégoriennes 22 (1988), pp. 43–107.

In James McKinnon, The Advent ProjectThe later seventh-century creation of the Roman Mass Proper (University of California Press 2000), the second-mode graduals are discussed inter alia in chapter 9, and passim.

For further resources, see the About the Archive.

Graduals in the Second and Eighth Modes 

Among the 100-plus Graduals of the Carolingian archetype there are twenty-four classed as second mode; these appear here as tracks 1–24. The archetype contains only three Graduals classed in eighth mode; these appear as tracks 25–27.

Of the twenty-four traditionally classed as second mode (Nos. 1–24), all but one use the same melody, with more or less modification for the sake of the individual wording of each item. The anomalous one, Adiutor meus (no. 24), clearly different from the other twenty-three, is more appropriately classed as first mode. It seems to have been classed mistakenly in the second mode because of its low incipit.

Two of the eighth-mode Graduals, (nos. 25 and 26), use one and the same melody (and they share one phrase with the second mode melody).The third eighth-mode Gradual, no. 27, has its own melody.

The second-mode model melody

The melody of the twenty-three second-mode Graduals (nos. 1–23) is universally regarded as one of the most beautiful of the Gregorian repertory, and rightly so. Without idiosyncrasy, it flows through a series of moderately melismatic phrases, harmoniously proportioned, intricately detailed yet sweeping in their larger contour. The melody can be read as serene, or as intensely expressive.

In the putting together of the Carolingian archetype this melody was obviously well regarded, perhaps because of its beauty, perhaps because it proved suitable for a variety of slots to be filled in the calendar of Mass Propers.

The sources of Urban-Roman chant that date from the eleventh century have a similar melody for the same sets of words assigned to the same liturgical occasions. It lies close at hand to assume that in Rome, around AD 750, these sets of words were sung to a melody from which both documented versions—the Gregorian and the Urban-Roman—could be derived. It is not my purpose here to imagine how such a melody might have gone.

The fact that this melody is used for twenty-three different sets of words makes it a “melody-type” or, better, a “model melody.” The ways in which the melody has been accommodated to different sets of words has elicited a number of detailed studies which attempt to show the basic structure of the melody and the various techniques of adaptation. Some of the studies attempt further to show how the melody came to be, either by identifying one of the existing settings as the original one, or—in a more analytic mode— attempting to reconstruct what the original may have been before adaptation to any of the existing sets of words or to any words.

Some observers might want to include more than the twenty-three given here, for instance the Requiem aeternam for the Mass for the Dead, or possibly some other item. The strongest reason for such inclusion is that these and still other Graduals in the second mode share the same melody. The only question is whether such other Graduals should be included in our reckoning of the Carolingian archetype, and that is purely a matter of documentation in liturgical sources. I see no way of telling, from the details of the adaptation of melody to words, at what time the adaptation was made—in the time of the archetype or later. Another set of Latin words can be adapted to the melody as recently as yesterday, leaving no clue as to date and provenance. What the twenty-three Graduals of the archetype give us is enough versions to present a very clear, confirmed idea of the melody, datable to circa AD 850.

The form of this model melody can be easily understood in terms of the plan of words normal for Graduals in general. The plan consists of two equal parts, a respond and a verse. Each of these parts contains a sentence, complete and independent. Either sentence may contain one, two, or infrequently more clauses; the abstract norm is two clauses, separated by a colon and embodying psalmic parallelism.

            Respond: first clause: second clause.

            Verse: first clause: second clause.

This norm is not so often found: the first in the list, Haec dies, shows the plan, but has peculiarities that result in an individual adaptation of the melody to these words. The Gradual Ab occultis is a better manifestation of the norm.

The melody for the verse provides two long musical phrases, one for the first clause, the other for the second. This consistent assignment is what makes this melody an instance of a model melody, rather than a repertory of phrases that can be used in various configurations.

Each musical phrase allows various small changes in its succession of pitches to accommodate varying number and kind of syllables in its verbal clause; these are more easily seen in examples than described. Some of the detailed studies have categorized these small changes, and have arranged the versions of the melody in tabular form showing how and where the changes have been made. In general, the melodic shape of each phrase is maintained to a remarkable degree; or to put it another way, the melody is really the same from version to version. This will become very apparent in listening to the melodies of this group one after another.

There are, however, numerous changes from one version to the next in the configuration of phrases, so that the normal plan is often not followed exactly. (The respond uses some of the same phrases, according to the same principle, but with less regularity.) It is the variation in the verbal plan that elicits the variation in musical phrase plan. If there is only one clause in a sentence instead of two, or if a clause is unusually short or long, or if there are more than two clauses, then a musical phrase may be omitted, or another phrase added.

These melodic materials are applied ad hoc to individual cases, and while various studies have made heroic efforts to account for all cases in a systematic way, still, the most useful way of understanding the melody is following any one version in performance.

The form of the melody that has seemed most normal appears in the four Graduals for Ember Saturday in Advent. Each of the four Ember Saturdays in the year has a special order of Mass that includes five extra lessons before the Epistle; four of these are each followed by a Gradual. Ember Saturday in Advent has the most complete and specific set of assignments of chants for these lessons, the other Saturdays having less specific assignments. And for Advent the four Graduals all use the second-mode melody. All this instance really tells us, however, is that the assignment of the same melody gives statistical weight to this particular form of the melody; and perhaps that is the main reason this form seems to us normal.

It is clear that this second-mode melody is more than just a collection of musical phrases: it is in some sense an integrated melody. We can understand it, if we wish, in purely musical terms, without reference to any set of words. That does not mean, however, that it was ever sung without any words; nor does there seem any way, on intrinsic grounds, to determine which set of words was sung first to this melody. By the same token, the logic of succession in the use of the melody for various sets of words is indeterminate: hence, the history of the liturgical use of this melody cannot be written, save only in fantasy.This normal melodic plan, with respond and verse each divided in half, is often obscured in its effect by the subdivisions, which seem to articulate a plan of four phrases rather than two. Then, too, the melismatic style so prolongs the syllables that the bihemistich plan of the words, even when present in a clear, simple form, is so extended as to be not perceptible, with the result that the effective phrasing is that of the melody rather than the words.

I believe that during one hearing of all twenty-three versions of the melody, it will emerge in a clearly articulated outline—yet not as a completely comprehensible succession of details. This was the first Gregorian melody I came to know well, more than a half-century ago; during the intensive involvement with it for the present recording (done without interruption for other recording projects) I of course came to know it as well as I have known any piece of music; none the less, the melody has kept something of itself hidden from me: its every performance is for me still something of an adventure.

Throughout the Notes, I use the medieval chant scale of pitch letters:
            A B C D E F G a b c [“middle C”] d e f g
Hence, the four finals of the modes are
            D (modes 1, 2), E (3, 4), F (5, 6) G (7, 8).

Adiutor meus

The melody for Adiutor meus is not the second-mode model melody just described, nor, in fact, is it in second mode, in spite of its traditional classification. The respond does make unusually consistent use of its lower register, corresponding to second mode; and the verse, for its first half, continues in a range that could be understood as second mode. The latter half of the verse, however, moves decisively upward into a           first-mode register, through the fifth D-a ascending as well as descending. The dwelling on the above, supported by a and F, is also distinctive of the first mode.

Three Graduals in the eighth mode

Deus vitam and Deus exaudi are contrafacts one of the other; they seem to have been composed to fill out the Lenten roster (whose ferial Masses required a large number of Graduals as well as chants of other kinds). The similar incipits of these two sets of words presumably elicited the contrafacture. Both sets belong to the program of Passion music that put Psalm quotations into the mouth of Jesus suffering on the cross.

The melody (it is the same for the two Graduals) has as central tonal space the fourth G a b c. It ascends above that for intense expression, and below it for a half-cadence on D (used in the same way in the Passion Graduals in third mode).There is another internal cadence on E, not often used anywhere else.

The third eighth-mode Gradual, Dilexisti, is unlike the other two. Standing slightly to one side of the Gregorian repertory, it seems to me to have a latent similarity to the Offertory Elegerunt apostoli, known to be from the Gallican repertory.

Notation and performance

It is well-known that medieval scribes and notators, in the course of putting together a chant book, had in mind that the book would be used in liturgical calendric order: in singing the second appearance of this model melody the singer would have read it previously in its first appearance. The terminal melismas, for instance, of the model melody are regularly given only by cue in most appearances after the first (which is Tollite). In this practice case, of course, the singer knows the melody by heart, and the first notation, complete, is for the sake of “archiving.” That is, subsequent appearances of this model melody have the function of showing how the melody is adapted to differing sets of words. Manuscripts L and C, however (and also some others), have the additional purpose of giving the signs of nuance, and that invites a close comparison of these signs in successive appearances of the melody throughout the book. There are discrepancies; and while paying close attention to such discrepancies, I make no attempt to normalize the details of either the melodic notation or the rhythmic nuance; rather, I look for some special sense to the variant notation.


Of the twenty-three Gradual responsories in the second mode in the archetype, one of them, Haec dies, appears six times, each time with a different verse, assigned to Easter Sunday and the following weekdays. Twelve others are for the Proper of the Time (Temporale), five are for saints' days (Sanctorale).

Within the Temporale the melody is used for various seasons throughout the year. Six settings are used in Advent, two for Christmas, six for Lent, and one for Easter and the five days of the following week.

If we observe the type of occasion involved, a specific but paradoxical picture emerges. Out of the eighteen Graduals for the Temporale (nos. 1–18), twelve are for ferias (in Advent or Lent). Dominum refugium, (no. 9) and Ab occultis (no. 10) are also assigned to ferias, but are used as well for Sundays in the Pentecost season; the number varies with the source in Sextuplex.

Of the remaining four, Hodie (no. 18) is for the Vigil of Christmas, meaning the day before Christmas,           24 December.

Tecum principium (no. 7) is for the First Mass of Christmas. This Mass is on the day itself, 25 December; it is traditionally called “Midnight Mass” because the day starts at midnight. So, even though sometimes called a “vigil,” the First Mass of Christmas has a status higher than ferial, but still lower than the Third Mass, the Mass of the Day, Puer natus.

Angelis suis (no. 8), for the First Sunday of Lent, is “dominical,” there being a dominical Mass for every Sunday (with the ever-present exceptions in all liturgical matters). In passing, we can note that Angelis suis was apparently used to replace Tenuisti, which is found in what seems to be the oldest layer represented in the Sextuplex, in the Blandiniensis.

Hence, only the one second-mode gradual of the Temporale, Haec dies (no. 1), is for a feast of the first rank, Easter—and it has always been a puzzle why this should be so. It has also been noticed, with equal puzzlement for some, that Haec dies is the one truly individualized melody of the set.

The five Graduals for saints' days show a similar distribution.

Exultabunt (no. 20) serves for two different pairs of martyrs (both Roman).
Nimis honoratus (no. 23) is used for a pair of Apostles, but also for a Vigil Mass.
Dispersit (no. 19) and In omnem terram (no. 21) are used for Vigil Masses.
Iustus ut palma (no. 22) was used for the First Mass of St. John Apostle.

So, of the five, only Exultabunt is for a feast of first rank.

Adiutor meus (no. 24), the non-conforming melody, is for a Lenten feria. Two of the Graduals in eighth mode (nos. 25 & 26) are also for Lenten ferias.