Acclamation sung in the Latin Mass directly after the introit. The basic text, which is Greek, consists of ‘Kyrie eleison’ (three times), ‘Christe eleison’ (three times), ‘Kyrie eleison’ (three times): ‘Lord, have mercy … Christ, have mercy … Lord, have mercy …’. The expression ‘kyrie eleison’ was widely used as an acclamation in pagan civic and religious ceremonies under the Roman Empire, and continued in Christian usage, becoming fixed in various Christian liturgies from the 6th century onwards. Numerous musical settings are documented from the 10th century on. Since (in more recent practice at least) the text did not change from day to day, the Kyrie is counted a part of the Ordinary of the Mass.
Kyrie chants survive in manuscripts of the 10th century onwards; the earliest are from France and Germany, while those of the 11th and 12th centuries include large numbers from other European countries. The catalogue by Landwehr-Melnicki includes 226 items; even so, it is not complete. The whole repertory tends to divide itself between items that have large concordances, hence were widely known and used, and items whose isolated concordances show them to be purely local products. The earliest and best-known melodies – the repertory of the 10th and 11th centuries – are very well represented among the Ordinary chants included in the Liber usualis; Kyries are normally identified by their numbering in this source.
The Liber usualis also identifies Kyrie chants by the incipit of the text formerly used with a given melody (e.g. Lux et origo for Kyrie I). Such texts are commonly called ‘tropes’ by modern scholars, and are not commonly discussed as integral parts of the Kyrie, on the assumption that being ‘tropes’ these texts are later additions to originally melismatic chants. The terminology has now been corrected on two points. First, text underlay is not troping, so that the Latin words set to Kyrie melodies are not tropes. Second, quite apart from the Kyrie melodies themselves with their Latin words, there are a few compositions that qualify as Kyrie tropes in the strict sense. They are in quite a different style, consisting of short verbal phrases with their own melodies in antiphon style, and they are used as insertions before phrases of the Kyrie melodies (Bjork). Only these short insertions, not the Latin words referred to by the Liber usualis titles, are to be called Kyrie tropes; it is convenient to refer to the Latin words laid under the melodies by the term ‘Kyrie verses’ (even though most are in prose, only a few being hexameters and only some later ones using syllable-counted verse). These Latin words conform to the basic form of the nine-fold Kyrie, falling into nine lines or verses, each usually ending ‘eleison’. The Latin words appear together with the melodies in the earliest sources (10th century). As with sequences, Kyries are notated in these sources in two forms, once in melismatic notation, with only the Greek words for identification; then again with the Latin words with syllabic notation. The early manuscripts give no indication that the melismatic notation represents an earlier form of composition – nor indeed any firm indication that the melismatic notation represents a melismatic performance. At best, the melismatic notation can be taken as an option in performance, and research has suggested several different ways in which the melismatic and syllabic notations can be used in combination and alternation.
For a number of early Kyries the manuscripts provide alternative sets of words; in some cases an alternative set may match the syllable count exactly; in other cases the alternative set requires modification of the Kyrie melody.
The liturgical use of the expression ‘Kyrie eleison’, especially in the Greek rite, went far beyond the one occurrence at Mass with which we are most familiar; ‘Kyrie eleison’ tended to be used as a ubiquitous response to other liturgical items. In western Europe it became associated with the various litanies, and this is in many ways its most characteristic use. The invocations in litanies are typically sung by the deacon, the people responding with some such expression as ‘eleison’. Exactly how ‘kyrie eleison’ came to be an independent item at the beginning of Mass – more precisely, after the entrance cortège reached the altar – has been the subject of extensive research, finally clarified by De Clerck and Baldovin.
So compact is ‘Kyrie/eleison’ that it invites elaboration in the form of more prolix invocations, comparable to those preserved in numerous Latin litanies from the 8th century on. Precisely such elaborations were mentioned by St Gregory in his famous letter: ‘In daily masses, moreover, we do not say the other things usually said, but only ‘Kyrie eleison’ and ‘Christe eleison’, in order that we may concern ourselves with these supplications at greater length’. Gregory further said that in the Roman rite ‘Christe’ was said as many times as ‘Kyrie’; but at that time, and for a century or two thereafter, the total number of petitions was not yet fixed. Ordo romanus I (?7th century) leaves the number of invocations up to the pontiff, while Ordo IV (‘St Amand’, early ?9th century) says that nine petitions shall be sung, at which time the pontiff shall give the signal to stop – an obviously redundant and now vestigial signal.
Latin texts tend to be provided for festal occasions, in some cases bordering on the function of a ‘proper’ text. Simple Kyries were intended in the 11th century for simple occasions, when, as Gregory said in the 6th century, the ‘other things usually said’ were omitted. Another possible application of Gregory’s comment is to assume that a more elaborate Kyrie could be sung on feast days with its Latin text, on lesser occasions without it. Still another is to imagine that each of the nine petitions was sung twice, first with, then without its Latin text, as a reminiscence of the 18-fold antiphonal practice prescribed for festal occasions in Ordo romanus XV.
Documents of the 10th century provide composed Kyries with Latin texts for major feasts in the ninefold shape, so that by that time the number of petitions was definitely fixed. In general both texts and melodies of this early repertory seem to be the product of Frankish monasteries. Many early Kyries are of a musical complexity that suggest performance by a schola, not by the whole congregation. There are, admittedly, simple Kyries, and one or the other of these might represent an earlier congregational use (as shown by John Boe in his studies of Kyries in Italian – Beneventan – sources).
More likely, the 10th- and 11th-century repertory includes reminiscences of earlier (pre-Carolingian) practice rather than integral melodies. Such stylized reminiscences of earlier functions can most easily be imagined in the ‘eleison’. In some Kyries, for example Kyrie IV (Cunctipotens genitor), the syllable ‘e-(leison)’ has a neume of several notes, as opposed to the mostly syllabic relationship in the preceding invocation with its Latin text. Even assuming the Latin text to be underlay, it would still be striking that the whole ‘melisma’ was not texted, but only the part preceding ‘eleison’. In this and other ways the ‘eleison’ is often set off from what precedes, as if it were functionally a distinct part, reflecting the earlier division into the solo invocation and the people’s response. Sometimes, as in Kyrie Rex genitor (Liber usualis VI), each of the nine ‘eleison’ settings is identical, again suggesting a response, although such ‘homeoteleuton’ could be explained on other grounds (it occurs in sequences for purely musical reasons). Usually the ‘eleison’ is varied, but variation occurs in other litanies too. The falling-3rd cadence on ‘eleison’ in Kyrie Fons bonitatis (Kyrie II) and elsewhere might be considered a reminiscence of the kind of recitation tone that was perhaps associated with a litany.
Ordo romanus I tells us that the Kyrie was sung until the pontiff signalled the prior of the schola to stop. In the Kyrie from the Mass for the Dead we might see a way in which the prior would indicate to the rest of the schola that the Kyries should end: all phrases save the last begin alike; the last begins abruptly with a different motif, as if the prior, singing the invocations, broke in at that point with a different melodic motif to signal that this was the last time around.
Many Kyries, early and late, have a phrase repetition within the invocation of the last Kyrie. This kind of repetition could have been at one time another means of signalling the choir (although here again the same repeated phrase structure occurs in sequences within certain couplets for the purely musical purpose of marking the climax of the melody). None of this is to be taken as argument that individual Kyries go back to the 7th or 8th centuries, only that Kyries composed in the 9th and 10th centuries might well recall, deliberately and in stylized manner, an earlier practice.
The establishment of the ninefold shape seems to have been the work of the Franks and their preoccupation with order; from that time on, the large tripartite division remained a basic feature of Kyrie construction (most visible in Kyries with the form AAABBBCCC), and has been used as such by modern scholars in analysing Kyrie chants. Yet the litany ingredient remained strong in the Kyrie, engendering other modes of construction that cut across the tripartite division while being no less characteristic of Frankish craft. A series of litanies has no inherent sectional structure: it goes on endlessly, perhaps rising in intensity, but not falling naturally into larger groupings unless these are imposed from outside. Many of the early Kyrie melodies show an overall ascent in pitch from start to finish; often ‘Christe’, and even more often the following ‘Kyrie’, attack a higher pitch. The Kyrie ad libitum VI, favoured for high feasts in the 10th century, shows the ascent most dramatically.
Here as elsewhere, however, the overall ascent does not proceed directly, but rather through a series of alternations: successive petitions are higher or lower, or the melodic material is arranged in some such pattern as ABACBCDCD (Conditor kyrie omnium, ad libitum V). In some respects this pattern shows an enthusiasm for interlocking musical construction comparable with northern interlace patterns in graphic arts, and as such would not be due directly to a litany model; but such interlocking design always tends to obliterate a sectional structure, thus creating the continuity characteristic of litanies (in a very artful way), and might even imitate some kind of antiphonal performance.
The most important segment of the repertory – the Kyries in widespread use during the 10th and 12th centuries – show a remarkable wealth of melodic invention and organization. There are occasional borrowings of phrases from one piece to another, and (as in other medieval categories in the process of creation and development) much reworking of individual pieces through a long series of variants. Nonetheless, the individual work is clearly perceptible in its artistic integrity, exhibiting a carefully wrought plan and detail. Most characteristic, perhaps, is the development of motivic material (sometimes of great expressiveness) to bind the ninefold shape together.
J.A. Jungmann: Missarum sollemnia (Vienna, 1948, 5/1962; Eng. trans., 1951–5/R, as The Mass of the Roman Rite)
M. Landwehr-Melnicki: Das einstimmige Kyrie des lateinischen Mittelalters (Regensburg, 1955/R)
P. de Clerck: La prière universelle dans les liturgies latines anciennes: témoignages patristiques et textes liturgiques (Münster, 1977)
D.A. Bjork: ‘The Kyrie Trope’, JAMS, xxxiii (1980), 1–41
D.A. Bjork: ‘Early Settings of the Kyrie Eleison and Problems of Genre Definition’, Journal of the Plainsong and Medieval Music Society, iii (1980), 40–48
J.F. Baldovin: The Urban Character of Christian Worship: the Origins, Development and Meaning of Stational Liturgy (Rome, 1987)
J. Boe, ed.: Beneventanum troporum corpus, ii: Ordinary Chants and Tropes for the Mass from Southern Italy, AD 1000–1250,pt 1: Kyrie eleison, RRMMA, xix–xxi (1989)
J. Boe: ‘Italian and Roman Verses for Kyrie leyson in the MSS Cologny–Genève Biblioteca Bodmeriana 74 and Vaticanus latinus 5319’, La tradizione dei tropi liturgici [Paris 1985 and Perugia 1987], ed. C. Leonardi and E. Menestò (Spoleto, 1990), 337–84
For further bibliography see Plainchant.
RICHARD L. CROCKER