Hymn of praise, sung in the Latin Mass directly after the Kyrie on festal occasions. Counted as part of the Ordinary of the Mass, the Gloria was provided with over 50 chant settings during the Middle Ages. The text is considered one of the great prose hymns of Christian literature, and the chant melodies are among the more important of medieval chant. The Liber usualis contains 15 of these chants in the Ordinary cycles plus four more among the ad libitum chants. (Throughout this article melodies are referred to by their Vatican number followed by their number in the Bosse catalogue, e.g. Gloria I/12.)
2. Early melodies: Gloria A/39.
5. ‘Doxa en ipsistis’, and Gloria XIV/11.
RICHARD L. CROCKER/DAVID HILEY
The text begins with the angelic hymn from the account of the Nativity in Luke ii.14, and continues with a series of disparate elements that includes reiterated praises (‘Laudamus te …’), acclamatory invocations (‘Domine Deus …’), petitions (‘… miserere nobis’) and a concluding doxology (‘Quoniam …’). The whole text is usually construed in three sections: first, praise to God the Father; second, a Christological section; third, the concluding Trinitarian clause. The nature of the text, however, makes several such constructions possible, and the various stages of development of the text up to the 9th century, as well as the varying structure of the chants, show that differing interpretations were made.
A shorter Greek version was used in the East as a hymn at morning and evening prayer, and some comparable version was used in the West (in Gaul) in the same way in the 6th century. The first extant Latin version, different in important particulars from the received version, appears in the Bangor Antiphoner (c690); the received version is first found in Frankish sources of the 9th century. The Gloria is placed in its familiar liturgical position after the Kyrie in the Ordines romani, in documents of the 8th century that presumably report practice of the 7th century. While the text itself suggests a close relationship to Christmas, liturgical practice (whereby the Gloria could be used unrestrictedly at Easter but was limited to the bishop at other times) suggests a closer relationship to Easter. In any case the use is seasonal, being omitted in Advent and from Septuagesima to Easter.
Documents containing melodies date from shortly after those containing the received text: that is, from the 10th century. If any melody is to be dated before that time it must be on the basis of conjecture. Evidence suggests that at first the Gloria (after the intonation) was sung by the clergy and people together, and from this it is usually concluded that the chant settings must have been simple ones suitable for congregational participation. The point at which the Gloria was presumably taken over by the Schola Cantorum (after 800) coincides more or less with the appearance of melodies far too elaborate for congregational singing. Gloria XV/43 has often been taken to be the oldest of the Gloria chants, and thought to be in fact an early congregational melody (in spite of the fact that it does not appear in the very earliest series of documents). Gloria ad libitum IV is another simple melody of a different type, taken from a 12th-century Ambrosian source. It can be said that such melodies are so simple as to be artistically neutral, basically uninformative about the more elevated liturgical music of whatever period they may have come from.
In contrast to these simple melodies, the chant that has the best early representation in the manuscripts is very elaborate, in fact in some ways the most elaborate of all medieval Gloria settings. This chant, called ‘Gloria A’ (A/39, sometimes ‘Gloria primus’) since it was not included in the Liber usualis, was the one most frequently troped in the 9th and 10th centuries, from which it has been concluded that it was the favourite festal chant at that time (transcriptions in Rönnau; Evans; Falconer, 1993, MGG2; and D. Hiley: Western Plainchant, Oxford, 1993, p.228; no critical edition is yet available). One can go on to conclude that it was the first such chant, and for a period the only one, and that it is much older than the others; but all these conclusions are less secure. In any case it is not Gregorian (as the weight of opinion now seems to agree); whether anything is gained by calling it ‘Gallican’ seems doubtful. It is clearly distinct from chants of the Gregorian corpus in the purposefulness of its motivic arrangement, and closely allied to Frankish chants of the 9th and 10th centuries by the same feature. The relatively florid style of its figuration, however, which might superficially suggest Gregorian models, can be more seriously taken to suggest some other kind of connection before or outside the Frankish 9th century – possibly to a Byzantine prototype (see §5). Boe (1982) has pointed out that not only is Gloria A/39 present in the Old Roman sources, but what appears to be a simpler version of it is also found there, raising the possibility of Roman origin.
Gloria
A/39 is neumatic throughout, with three important melismas marking off three
paragraphs: ‘Glorificamus te’, ‘Jesu Christe’, and ‘Amen’ (not counting the
presumably interpolated melisma on the versicle Regnum tuum solidum,
after ‘altissimus’). The first two paragraphs cadence on a–b
–a, the last on g–f.
A single formula is repeated for the laudes in the first paragraph, with
cadence on g–a. More complex formulae, more freely handled, are
used for the acclamations in the second paragraph and for the petitions in the
third, with cadences on g–a–g. Motivic relationships,
sometimes subtle but often obvious, run through the whole piece.
The
overall pitch set (not to speak of the mode) is difficult to determine and,
perhaps because of manuscript variants, indeterminate, especially in the
intonation. The intent seems to be, however, to base the piece on f,
using mainly the pitches up to d', with both b and b
, and internal cadences on g and a.
By way of exception, the melody descends to d and c, and passes
through e or e;
it ascends at the end to e' (but if the whole chant were imagined on g
instead of f, the top pitch might be f' – that is, e'). There is a strong
emphasis on b
in the third paragraph, as opposed to the more usual a or c'.
Compared with Gloria A/39, Gloria IV/56 is much more regular in its construction, and simpler in style, lacking melismas; it is not however, purely syllabic – indeed, none of the elaborate Glorias are. A single melodic shape, made up of two or three phrases, is used over and over again through the body of the chant. This shape, most easily seen at ‘Gratias agimus … gloriam tuam’, or at ‘Domine Deus … omnipotens’, has an intonation f, d, then finds its way to a mediant cadence on e(approached from below, c–d–e) for the end of the first phrase. The second phrase ends with the neume first heard on ‘Glorificamus te’ distinguished by the descent from a to e (a–g–f–g–e), the fall from g to e being either filled in or left open. An alternative ending for the second phrase is found on ‘gloriam tuam’, with a similar fall to e.
This compound melodic curve is used for every period except the first and last, with great flexibility of detail. The technique could not be compared to simple psalmody or even psalmody at the introit; only the verses of the Matins responsories show a comparable freedom in adapting a formula to a particular text. And the handling of two-note neumes, their obvious decoration of a simple underlying line, also resembles the responsory verses.
The beginning (‘Gloria … benedicimus te’) uses the same motivic material as the rest of the chant, but more freely. On the one hand there is the relatively long construction of the angelic proclamation to be set, on the other hand the manifold short acclamations; it is plain that the composer was concerned to find appropriate solutions for each of these elements. Similarly at the end, the motivic material comes in a different order, to suit the several short syntactic units that make up the closing period beginning ‘Quoniam’. The terminal cadence comes three times in succession on ‘Christe’, ‘Spiritus’ and ‘Amen’, which is thus an integral part of the melody, since ‘Dei Patris’ ends inconclusively on d.
The melody as a whole moves within the range c to a, with the exception of four occurrences of c', distributed throughout the piece. The framework c–e–g is prominent, relieved (again, four times) by a momentary stress on f. These details, at first glance mere random deviations, seem actually to be carefully placed in a manner in keeping with the prose nature of the text – artistic though irregular. Through such detail the potential monotony of a repeated formula is elevated to a higher level; the melody gives the impression of variety and larger form even though structure by paragraph or section is lacking.
Gloria I/12, closely related to Gloria IV/56 in certain idioms (especially at ‘magnam gloriam tuam’), is different in construction. A single melodic shape is repeated, but the shape is so much longer and more complex as to produce an entirely different effect. After the opening period, which as before is more free than the rest, there are four presentations of the basic shape:
(1)
‘Laudamus … omnipotens’
(2) ‘Domine … filius Patris’
(3) ‘Qui tollis … miserere nobis’
(4) ‘Quoniam … Amen’
The shape moves through a series of sub-phrases centred on g, b, and d, to a mediant cadence on b; then it rises through the motive a–e'–f' to its highest point, from which it descends in groups of threes – f'–e'–d', e'–d'–c', d'–c'–b – disguised in various ways but always present as the underlying line. This line eventually descends to g, and may stop there (as at ‘miserere nobis’ in section 3), or may add a concluding cadence on b, as in sections 1, 2 and the Amen. The syntactic division resulting from this melodic plan preserves a clear Christological section (2) distinct from the litanies (3).
The
tonal range is identical with that of Gloria IV/56, making allowance for the
different location on the scale (g–e' instead of c–a'), except
for the high note, here a semitone above the top of the range (f'). The
recurrence of this range of a major 6th, here and elsewhere in Gloria melodies,
suggests that it represented a common ground, a matrix in which such melodies
were conceived; pitches lying outside – particularly above – the 6th might then
be considered variable: the high note in either of these two chants could be a
semitone, a tone or a minor 3rd without changing the essential structure of
either melody. The same 6th can be used to clarify Gloria A/39, providing a
framework much easier to understand than a modal analysis. Furthermore,
internal cadences in Gloria I/12 fall on g, a, b, in ways that show
careful planning. The flexibility of such cadence points, the combination of a
very clear sense of locus with the unstable deuterus ending (e or b), seem to be the result
of composition based on this 6th.
A melody was circulated in 9th- and 10th-century manuscripts over the text Doxa en ipsistis, a Greek version of Gloria (Huglo). What this version represents is problematic: it might be a survival in the West of an old chant; it might be an importation in the 8th or 9th century of a Byzantine version; or it might be a 9th-century Western construction, using materials of Western tradition or invention and cast in the guise of a Greek version for reasons that can only be surmised. (There were a number of other such items circulating in Western sources in the same period.) In any case, the melody is most comparable in style to Gloria A/39, having the same neumatic style. Certain phrases, however, have a more individual character and move more actively through a wider range; and it is precisely these phrases that appear in Gloria XIV/11, a remarkable melody circulated in 10th-century sources.
Some of the material of Gloria XIV/11 that is definitely not derived from the Doxa en ipsistis is close to Gloria I/12; and Gloria XIV/11 begins as if it were to be located within the 6th g–e'; but instead of the high e', the low e is introduced, almost as an afterthought, in the falling-3rd cadence on ‘voluntatis’. This low e then assumes increasing importance throughout the chant until it serves as the final, while the b above g, which might have been taken as a final (as in Gloria I/12), becomes a mediant cadence. The low e also comes to function as the beginning of a phrase, although in that role it remains more clearly outside the central range (as at ‘Rex caelestis’ etc.).
The shift in tonal locus is intimately associated with the intricate phrase structure. There are more periods than in Gloria IV/56, but they are much less stable than the phrase groups of Gloria I/12. There are nine, as follows:
(1)
‘Gloria …’
(2) ‘Laudamus …’
(3) ‘Gratias …’
(4) ‘Domine Deus rex …’
(5) ‘Domine fili …’
(6) ‘Domine Deus, agnus …’
(7) ‘Qui tollis, suscipe ’
(8) ‘Quoniam …’
(9) ‘Cum sancto …
… Amen.’
Because of the way motifs are gradually phased in and out, or transmuted, no clear paragraph structure emerges, even though higher-level relationships are suggested (as ‘Domine Deus rex … Domine Deus, agnus’). The result is a continually unfolding form. The motifs derived from the Doxa en ipsistis play important roles in the development of the form. (Many of the occurrences of c' would be b in a reconstruction of the 10th-century state of the melody.)
Glorias IV/56, I/12 and XIV/11, together with Gloria A/39 and the Doxa en ipsistis, can be taken as representative of the first stage, or stages, of Frankish chant provided for the Gloria in the 9th century. Other melodies, too, can be presumed to go back that far (although the chronological order of the repertory has yet to be worked out in detail), and other pitch structures and modes were represented, especially protus plagal (d final) with Gloria XI/51, and tetrardus plagal (g final) with Gloria VI/30. This early stage reveals a wide spread of technique varying from simple repetition of a melodic formula to a flexible, varied motivic development. There is a wide spread, too, in its complexity. Syntactic structure is different in almost every case.
The most striking of the subsequent stages of development involves a substantial increase in the range within a given melody. Often this increase is apparent within a single phrase or phrase group, giving a bravura aspect to the melody; it may also be associated with a long, clearly perceptible ascent towards the top of the range, which gives the melody as a whole a direction and élan. This use of range seems dependent upon the strength of the tonal set as found in the early melodies, and also upon their techniques of motivic control.
Expansion of range can be studied in the several melodies in tetrardus which appear in the Liber usualis: Glorias VI/30, III/20, V/25, IX/23 and ad libitum I/24. Gloria V/25 moves regularly through a range of a 7th in individual phrases. Gloria IX/23 has an overall range of an 11th; the melodic motion is arranged to show an insistent progress towards the top of the range at ‘tu solus altissimus’. Gloria ad libitum I/24 is even more spectacular, having much more elaborate motion within single phrases (‘Glorificamus te’), and more extreme progressions within phrase groups (‘Qui tollis peccata mundi, miserere nobis’). As with Gloria IX/23, however, the active nature of the line has no effect upon the solidity of the tonal locus, except possibly to enhance the strong returns to g or d after the arcs have swung wide above and below. Indeed, in some ways the basic tonal motion here is less than in earlier melodies, for there is no real move away from the g and the d, which act almost as pedals throughout. Other grandiose melodies have been reported, especially from German sources (Stäblein).
Gloria chants are important in the medieval repertory not so much for their number, which is relatively small, nor for the structure of the text, which is different from the most popular medieval categories. Perhaps the most important aspect of Gloria melodies – an aspect not found in all or even most, but nonetheless distinctive when it does appear – is the construction of a piece out of a changing, developing, but highly interrelated series of motifs. This construction, which is but poorly represented by a tabular analysis of the motivic material, gives to a piece a unique shape whose process and continuity deny any clear sectional plan, but with no loss of clear, forceful design.
Bosse's
catalogue (suppl. by Hiley, 1986) includes 56 Gloria melodies from sources from
the 11th century to the 18th. Since the great majority of Bosse’s 341 sources
date from the 13th–15th centuries, his statistics on the distribution of modes
throughout different countries and centuries have little or no bearing on the
development of the earlier melodies. His demonstration of the late popularity
of the f-final (with b) does, however, seem significant.
Gloria melodies appear in the earliest 10th-century sources in conjunction with their tropes (as can be studied in Rönnau's catalogue). Tropes were provided most frequently for Gloria A/39, less frequently for Glorias IV/56, VI/30, I/12 and XIV/11. Of great importance in their own right, Gloria tropes need to be studied for their musical relationship to the Gloria melodies and for the effect that their interpolation has upon these melodies, for the more massive tropes can virtually double the length of the Gloria. Subtle differences in style between Gloria melodies and tropes, even when of the same period, may perhaps be perceived.
One Gloria may have been conceived with trope verses from the beginning: Gloria IX/23 with the Marian trope beginning Spiritus et alme orphanorum (see Schmid), which appears to have been composed in northern France at the beginning of the 12th century.
Polyphonic settings of Gloria trope verses are already present among the Winchester organa (GB-Ccc 473, mid-11th century). The manuscript W1 (D-W 628 Helmst.) from St Andrews, dating from about 1240, has a two-part setting of a Marian trope, Per precem piissimam (similar in form and sentiment to Spiritus et alme), but here the complete Gloria is set as well (ed. M. Lütolf, Die mehrstimmigen Ordinarium Missae-Sätze vom ausgehenden 11. bis zur Wende des 13. zum 14. Jahrhundert, Berne, 1970). Polyphonic Gloria settings both troped and untroped are common from the 14th century onwards.
MGG1 (B. Stäblein)
MGG2 (K. Falconer)
O. Marxer: Zur spätmittelalterlichen Choralgeschichte St. Gallens: der Codex 546 der St. Galler Stiftsbibliothek (St Gallen, 1908) [with edn of many late medieval Gloria melodies]
M. Sigl: Zur Geschichte des Ordinarium Missae in der deutschen Choralüberlieferung (Regensburg, 1911) [with edn of many late medieval Gloria melodies]
M. Huglo: ‘La mélodie grecque du “Gloria in excelsis” et son utilisation dans le Gloria XIV’, Revue grégorienne, xxix (1950), 30–40
D. Bosse: Untersuchung einstimmiger mittelalterlicher Melodien zum ‘Gloria in excelsis Deo’ (Regensburg, 1955)
K. Rönnau: Die Tropen zum ‘Gloria in excelsis Deo’ (Wiesbaden, 1967)
P. Evans: The Early Trope Repertory of Saint Martial de Limoges (Princeton, NJ, 1970)
C.M. Atkinson: ‘Zur Enstehung und Überlieferung der “Missa graeca”’, AMw, xxxix (1982), 113–45
J. Boe: ‘Gloria A and the Roman Easter Vigil’, MD, xxxvi (1982), 5–37
D. Hiley: ‘Ordinary of Mass Chants in English, North French and Sicilian Manuscripts’, JPMMS, ix (1986), 1–128
B. Schmid: Der Gloria-Tropus Spiritus et alme bis zur Mitte des 15. Jahrhunderts (Tutzing,1988) [study and edn, pubd separately]
C.M. Atkinson: ‘The Doxa, the Pisteuo, and the Ellenici fratres: some Anomalies in the Transmission of the Chants of the “Missa graeca”’, JM, vii (1989), 81–106
J. Boe, ed.: Beneventanum troporum corpus, ii: Ordinary Chants and Tropes for the Mass from Southern Italy, A.D. 1000–1250, pt 2: Gloria in excelsis, RRMMR, xxii, xxiii–xxiv (1990)
K.A. Falconer: Some Early Tropes to the Gloria (Modena, 1993)
For further bibliography see Plainchant.