Antiphon.

In Latin Christian chant generally, a liturgical chant with a prose text, sung in association with a psalm. In Gregorian psalmody, for example, psalms and canticles are usually preceded and followed by a single antiphon, and the psalm tone used for the recitation of the psalm itself is often musically incomplete without the antiphon. Antiphons of this kind may be regarded as typical and are represented above all by the Gregorian antiphons to the psalms of Matins, Lauds and Vespers in the Divine Office. There are also other categories of antiphon, some of which may lack psalmody or have versified texts.

The antiphon and responsory are the two musical genres with Latin prose texts that occur in all the Western liturgies and are the most abundant within the chant repertory. A given medieval Office source might have as many as 1500 antiphons, and even up to 2000; such a large number could occur in a monastic usage, whereas the secular (or canons') repertory would generally have fewer (see §4 below).

The term ‘antiphon’ in early texts is frequently to be distinguished from ‘antiphony’ (i.e. antiphonal psalmody). Though etymologically related, these terms designate different practices: ‘antiphon’ referring to a musical composition does not imply an alternating or ‘antiphonal’ style of performance; and ‘antiphony’ referring to the alternation of performing groups does not necessarily suggest the use of antiphons.

1. Early history and terminology.

2. Origins and composition of texts.

3. Origins and composition of melodies.

4. The Gregorian antiphons to the psalms at Matins, Lauds and Vespers.

5. Other antiphons in the Gregorian repertory.

BIBLIOGRAPHY

MICHEL HUGLO, JOAN HALMO

Antiphon

1. Early history and terminology.

The Greek word antiphōna, derived from the classical Greek antiphōnos (‘resonating with’), was adopted in Latin without translation. For Pseudo-Aristotle (Problems, xix.39) the neuter adjective antiphōnon signified the interval of an octave. Like other ancient Greek theorists, he scarcely distinguished antiphōnia, the octave, from the unison (Problems, xix.17). Even in the writings of the 14th-century Byzantine theorist Manuel Bryennius the term retained this meaning.

In the writings of the Jewish philosopher Philo of Alexandria (d c54 ce), the term ‘antiphon’ refers to a practice of the Therapeutae sect. It has been asserted that in this context the word indicates the singing in alternation of male and female choirs (De vita contemplativa, xi), a usage that may reflect some element of the term's original meaning, since women's voices are pitched an octave above those of men. According to another interpretation, all the ways in which Jewish communities sang the psalms during this period involved a soloist, and thus the kind of alternation referred to in Philo's description would have been responsorial rather than antiphonal (Bailey, 1994).

The writings of the historians Socrates and Sozomen testify that sung pieces called antiphons were used in Antioch in the 4th century. These may have been texted with theological tenets connected to the Arians, or they may simply have been doxological in nature. It has been suggested that, at this early time, an antiphon was a piece characterized by its non-biblical words and intended for popular singing outdoors (Hucke, 1953). Such music seems to have involved some manner of performance by one or more soloists and an assembly.

In the West the term antiphona first appeared in the late 4th-century Itinerarium of the Gaulish pilgrim Egeria. Describing in detail the Jerusalem liturgy as she witnessed it, Egeria referred to the antiphon about 30 times, usually associating it with the psalms (‘psalmi et antiphonae’). Here the term apparently means a piece sung with the psalms rather than the practice of psalmody in antiphonal style. Although antiphonal psalmody had been in use in Jerusalem, this ancient form did not survive. In the Byzantine rite it was replaced by troparia and kanōnes (see Troparion and Kanōn). Vestiges of it are found in the West, however: the Mozarabic chant Introeunte te is one of several antiphons that are literal translations of Byzantine stichēra derived from the old antiphonal psalmody of Jerusalem.

The term antiphona (taken as a neuter plural) appeared at an early date in Provence. It is found in the writings of John Cassian (d 435), who had lived among monks of the East for a time and who later founded the abbey of St Victor at Marseilles (‘cum stantes antiphona tria concinuerint’, Institutiones, iii, 8; cf ii, 2), and the Rule of Aurelian of Arles (c547), modelled on the usage at Lérins (‘antiphona tria parvula’).

Communal monastic life as established by St Benedict (d c547) offered a permanent context in which the antiphon would flourish in the Western liturgy. In the Rule of St Benedict, dating from about the mid-6th century, chapters viii to xviii contain a developed Office, which includes the obligation to recite the enitre Psalter on a weekly basis. The rule uses the term ‘antiphon’ for a separate chant sung with a psalm, although it does not specify the number of repetitions of this chant. The Rule of the Master, now thought to be from the first quarter of the 6th century, mentions the term ‘antiphon’ only twice.

The manner of performance and psalms referred to in the monastic Rules may have involved some kind of alternation among the singers, but performing practice did not necessarily remain consistent or unchanged from the early centuries onwards. In a study that thoroughly reconsiders past assumptions, Joseph Dyer concluded that in monastic practice from the Egyptian monks to Carolingian times, the psalms in the communal Office were sung by individuals rather than by the assembly as a whole (Dyer, 1989). The leadership of a soloist highly skilled in psalmody would explain why the number of psalm terminations is much greater in early tonaries than in later ones, when choral psalmody had become the customary method of performance. It has been proposed that in the Old Roman repertory at least, and possibly much more widely, the antiphon itself may originally have been performed with solo intonation of the first half and with more soloists – or with any of the community who were able – completing the second half (Nowacki, 1995).

Isidore of Seville (d 636) seems to have been responsible for the long-standing credit given to Ambrose of Milan (d 397) rather than to monastic circles for the introduction and spread of antiphons in the West. Isidore's comments accord with those of Ambrose's biographer, Paulinus, who reported that antiphons and hymns were sung by the bishop Ambrose and his flock during a siege in 386 (PL, xiv, 31). What were called antiphons in the report of Paulinus appear not to have been the type of antiphons intricately linked to psalmody as known later; and in any case, the psalmody practised in the time of Ambrose must have been responsorial. Thus, the attribution to Ambrose of a historic role in the spread of antiphons of the kind later universally sung in the West cannot be upheld (Bailey, 1994; Nowacki, MGG2).

Isidore, and similarly Aurelian of Réôme in his writings of the mid-9th century, used the term ‘antiphon’ in reference to a chant with two choirs in alternation (‘antiphona … vox reciproca duobus scilicet choris alternatis psallentibus’, Etymologies, vi, 19.7). In Spanish and Greogrian antiphoners and in manuscript fragments of the late 8th century, the abbreviations, ‘A’, ‘AN’, or ‘ANA’ (i.e. ‘antiphona’) simply indicated a chant to be performed with a psalm. Also in the late 8th century and onwards, psalmody sung in alternation by two choirs was described in Frankish sources as if it were the customary kind. The earliest unequivocal description of refrains in alternating choral psalmody – that is, the first account of regular medieval psalmody with antiphons – was given in the first quarter of the 9th century by Amalarius of Metz.

Most of the antiphons for the Psalter have brief texts, not always grammatically self-contained, consisting of a simple invocation or acclamation of a few words; they are quite similar to the responds (responsae) used in ancient responsorial psalmody (Huglo, 1982). In the psalter of St Germain-des-Prés (F-Pn lat.11947, written in 6th-century uncials), which may have originated in Italy, the responsae are indicated by the letter ‘R’ in gold. These responds are very brief; they include ‘Praeceptum Domini lucidum’ (f.30v) for Psalm xviii, ‘Beatus qui intelligit’ (f.77v) for Psalm xl, and so on. Even the responds for the vesper psalms are equally terse, such as ‘Juravit Dominus nec penitebit eum’ (f.225v) for Psalm cix. For Psalms cxlviii–cl, sung in most liturgies at Lauds (Ainoi), the respond is ‘Alleluia’ (written in gold at f.288v); this corresponds to the alleluiaticum of the old Gallican liturgy (see Gallican chant, §10). These ancient responds, like many antiphons, are very short – between three and six words as a rule – and often consist of the first words of the psalm. The following examples of antiphons from the Roman Psalter make this clear: ‘Diligam te Domine virtus mea’, antiphon to Psalm xvii, borrowed from the beginning of the psalm; ‘Benedictus Dominus in aeternum’, antiphon to Psalm lxxxviii; and ‘Benedicite gentes Deum nostrum’, antiphon to Psalm lxv. Of the 95 antiphons to the Psalter, 50 consist of the opening words of the psalm. This evidence could suggest that the antiphons to the Psalter were derived directly from the older responds, but such a conclusion is by no means definitive.

In antiphoners of the late 8th century the number of repetitions of the antiphon in performance is never indicated. Descriptions of contemporary practice may, however, be found elsewhere. For the antiphons of the Mass (e.g. introits), the usage is described in the medieval liturgical documents Ordines romani (see Introit (i), §1), For the antiphons of the Office, according to Amalarius of Metz, in the 9th century the antiphon was repeated after each verse of the psalm (‘antiphonis, quas vicissim chori per singulos versus repetunt’, PL, cv, 1251). This practice was maintained in the Invitatory psalm, the Venite (Psalm xciv, sung daily at the beginning of Matins except during the final days of Holy Week): the antiphon is sung in full (e.g. for Tuesday, ‘Jubilemus Deo salutari nostro’) at the end of all the odd-numbered verses as well as the doxology; besides this, the antiphon is sung in full, as normal, before and after the invitatory psalm. It was impossible for musical reasons to separate the antiphon from the psalm and sing it merely at the beginning and the end, as was done elsewhere. The invitatory has eight special melodies, one in each mode except for the 1st and 8th (none), and 4th (three), although in many instances the number varies from one source to another. The potentially numerous and time-consuming repetitions of the antiphon (especially at Matins, with nine psalms in the secular usage and 12 in the monastic) caused these particular settings to be rapidly abandoned in subsequent centuries for other psalms (see Stäblein, MGG1).

In the various ancient Latin rites – Gregorian, Ambrosian, Gallican and Mozarabic – the term antiphona generally precedes specific short chants with texts averaging between ten and 25 words and with simple melodies. These antiphons appear before psalms and are also occasionally copied before the final melodic formulae of the psalms. Such indications were essential to the performance of the antiphons, which were conceived as integral parts of the psalm in the succession antiphon – psalm – antiphon. This form has several consequences. The antiphon ends with a clear cadence on a final note, which then determines the choice of the reciting note for the psalm (in Gregorian psalmody five reciting notes are possible; in Ambrosian chant, seven). The psalm tone ending (differentia, diffinitio, or varietas) can be chosen with the first notes of the antiphon in mind to ensure a smooth progression between the psalm and the recurring antiphon. While the latter explanation became the conventional one put forth by many theorists, there is often no perceptibly close relationship between the psalm tone ending and the opening of the antiphon.

In Offices written at the end of the 9th century and beginning of the 10th, both by East Frankish composers (as in the Office of St Otmar composed at St Gallen) and West Frankish (as in the Office In plateis by Hucbald and the Offices by Stephen of Liège), the numerical position of a chant in an Office determines the choice of its psalm tone. Thus the antiphon corresponds to a psalm sung to the 1st psalm tone, and so on until the eighth antiphon; the ninth antiphon again corresponds to a psalm sung to the 1st psalm tone, and the pattern continues. Psalm tones continued to be chosen according to their numerical order in the Office until the end of the Middle Ages, especially in the 13th-century Versified Office.

Several categories of antiphon developed without any link to psalmody: they include the great processional antiphons in the Gregorian repertory (see §5(vi) below); and the Gallican and Ambrosian antiphons with verses, which resemble responsories in that the antiphons are repeated after the verse (see §5(ii) below).

Antiphon

2. Origins and composition of texts.

The antiphons of the various Latin rites normally have biblical texts. Those of the regular Sunday and weekday Offices are drawn from the Psalter (see §5(i) below), the earliest source of Latin antiphon texts. Some Mozarabic antiphons retained textual variants originating in the African Psalter; this seems to indicate that during the revision attributed to Isidore of Seville of biblical texts and the liturgy, no-one dared touch the oldest antiphons in the repertory, perhaps for the sake of their melodies. Similarly, certain Ambrosian antiphons display textual variants older than the Milanese Psalter.

The antiphons of the Proper of the Time are biblical, often taken from texts read during the Office or the Mass of the day; the same biblical extract may be used in a variety of ways and liturgical positions in the different Western liturgies. The antiphons of the Proper of the Saints have texts based on the Acta martyrum or biographies of famous saints. The Gregorian antiphoner also contains about 100 antiphon texts drawn from Christian poets or basilical inscriptions; these are in metrical or rhymed verse. Many of the prose texts, too, are assonanced and balanced quantitatively in the manner of verse. Finally, a few antiphon texts in Latin are known in Greek as well, although resemblances between their respective melodies are not always evident.

Deciphering the origins of antiphon texts had already fascinated scribes associated with the Mozarabic Léon Antiphoner and with the Gregorian Zwiefalten Antiphoner (D-KA Aug.LX). Some years after the latter had been completed, an obviously experienced scribe went through the manuscript repeatedly, indicating in the margins the biblical sources of many antiphons and marking with the term ‘cantor’ those texts of ecclesiastical origin for which he had been unable to discover the source. During the 20th century, research on texts and textual sources of antiphons became available in a number of publications (Marbach, 1907; Lipphardt, 1965; Hesbert, in CAO, iii, 1968; Bailey, 1994; and Halmo, 1995).

Throughout the various Western liturgies, a literary comparison of antiphon texts with their sources illustrates the independence of the respective authors, who aimed to create a balanced verbal structure by means of additions (such as the word ‘Dominus’ before ‘Jesus’, or one or two ‘alleluias’ at the end of a chant), or omissions (e.g. of the enclitic ‘autem’, common in the Gospels). In such ways the author generally produced two subdivisions, each further divided into two clauses. Other schemes are also found, some with tripartite divisions, and some with even greater complexity, especially in the antiphons from Offices composed in the 10th and 11th centuries. Clearly the authors of antiphon texts felt free to modify the original biblical or ecclesiastical literature according to the requirements of musical composition.

The creation of new antiphons seems to have come to an end fairly early in the Ambrosian and Mozarabic liturgies, but it continued in other places, notably in ecclesiastical centres of the Carolingian Empire, even after the promulgation of the ‘antiphoner of St Gregory’ (see Old Roman chant, and Gregory the Great). The composition of Proper chants for the Office, both antiphons and responsories, never really ceased, particularly for the patron saints of dioceses and great abbeys. By contrast, very few new Proper chants for the Mass came into the repertory apart from alleluias. In the Proper of the Time for the Office, for example, the antiphon O virgo virginum was added to the series of seven Advent ‘O’ antiphons after the diffusion of the original Gregorian antiphoner (see O Antiphons). The same is true of the antiphon Rex pacificus and others for First Vespers of Christmas, composed in about 800 ce (see Hucke, KJb, 1953); the Veterem hominem antiphons for the octave of Epiphany, with texts translated from the Greek and melodies related to the surviving Byzantine melodies (see Strunk, 1964); and, according to Amalarius of Metz, the antiphons for the three Matins psalms of Easter and Pentecost.

Several observations can be made about the texts of antiphons with regard to their liturgical functions, especially during the Proper of the Time and the Proper of the Saints. Endeavouring to make the Office increasingly specific for feasts and seasons, medieval liturgists or musicians gradually replaced antiphons of the psalmic or ferial series with new ones. In many instances – and this has been demonstrated in detail for the Ambrosian liturgy (Bailey, 1994) – the newer substitute antiphons were attached to psalms with which they had some kind of textual link, perhaps only a single key word, so that very often there exists a nuanced and sensitive alignment of antiphon text and psalm text.

In the celebration of a given liturgical day, antiphons thus came to serve not only a musical function but also a liturgical one, that is, to convey concisely through their texts the spirit of a particular feast. The theological and spiritual aspects of feasts and seasons as reflected within the psalms and canticles are highlighted by the imagery of the antiphons. Further, when juxtaposed with the regular psalmic cursus sung week by week, various antiphons may summon quite different resonances from the same psalms on different occasions, while themselves acquiring an increasing wealth of meaning. During the liturgical year as a whole, therefore, these richly multifaceted antiphon texts, along with other Propers and psalms of a given day, often allude to words and phrases linked to other festivities and seasons. Repeated in changing contexts over time, they form a mosaic of connected liturgical themes and associations. In this contemplative interweaving of theological thought and spiritual expression, which is certainly one raison d'être of the Office, antiphons play a very significant role.

Antiphon

3. Origins and composition of melodies.

(i) Overview.

(ii) Adaptation.

(iii) Centonization.

Antiphon, §3: Origins and composition of melodies

(i) Overview.

There is no evidence that the writers of antiphon texts, even where known (e.g. St Ildefonsus of Toledo, c606–67, author of an alleluia and other chants for St Leocadia), also created the melodies; experienced singers are just as likely to have been the composers. The musical evidence – the manner in which melodies are adapted to texts, and the relationships between text and music – suggests that the composers knew Latin at least as thoroughly as Carolingian scholars such as Alcuin, Hrabanus Maurus and Loup de Ferrières. It suggests also a remarkable knowledge of music theory, in particular of the rules governing the relationship between words and music.

These composers did not, however, always place the textual accents in accordance with the practice of classical Latin (the accents normally correspond to a note or melisma in the chant that is higher in pitch than the pitches for unaccented neighbouring syllables). Thus ‘mulierem’, originally accented on the antepenultimate syllable, was turned into a paroxytone (i.e. a word accented on the penultimate syllable) in the antiphons Inter natos mulierum and Mulieres sedentes. Latin, which was spoken by the educated throughout the Middle Ages, had continued to evolve. Thus the composers of Gregorian chant in the second half of the 8th century adopted the current rather than Ciceronian usage in pronunciation and accentuation. In later medieval antiphons, the Latin reflects further changes. The mid-16th century gave rise to a movement whose intention was to correct the accumulated ‘errors’ in chant, that is, deviations from the Ciceronian standard. In this attempt at revision, the melodic accentuation of words was usually ‘corrected’ by the placing of a single semibreve (lozenge-shaped note) on unstressed syllables, as, for example, in the normal cadence in the communion antiphon for Epiphany (ex.1). These melodic adjustments were retained by the composers of Neo-Gallican chant in the 17th century. The study of word accentuation has been revolutionized since the late 19th century, however, and chant melodies have accordingly been reconstructed in careful editions such as those of Solesmes (although, most recently, the search for historical authenticity has prompted scholars and performers alike to favour the use, whenever possible, of individual manuscript sources rather than even these very reliable composite editions).

Although the origins of individual antiphon melodies cannot be determined, close analysis suggests that their roots can often be identified. For this, Hucke (1951) proposed three categories of song, recitiation-like melody, and free-form ‘Strophen’, each one giving rise to certain characteristics: the song is musically cohesive and resistant to significant alteration; a melody based on recitative gravitates towards one pitch or a pitch area that might coincide with the psalm recitation tone; and free-form melodies consist of somewhat standard melodic phrases employed as needed, with melodic adjustments for textual requirements. Further, the melodic invention evident within the antiphon repertory has not taken place independently of the characteristics of the modes themselves, for certain modes tend to produce melodies of a particular form or style. Mode 1 and 2 melodies, for instance, move within a more limited melodic ambit, and mode 2 antiphons could even be described as having a propensity towards recitation. On the other hand, mode 8 abounds in melodic patterns, some circumscribed and others expansive, and has yielded melodies that are especially favourable to multiple adaptation (Hucke, 1951; Halmo, 1995).

As the repertory continued to develop, composers of antiphon melodies, like all composers of chant, borrowed the elements of their compositions from pre-existing models rather than seeking complete originality. For Gregorian chant the melodic models seem to have been fixed in the second half of the 8th century. In the period that followed, the expansion of the repertory involved various processes, including two seemingly contradictory ones: model melodies with the greatest adaptive possibilities generated a variety of others; yet at the same time a relatively small number of model melodies came into use more and more frequently but with particular exploitation of their own musical potential (see Crocker, 2/1990). Because of the size of the antiphon repertory, these and other kinds of development are readily discernible. Among those who have explored such musical interrelationships within the vast antiphon repertory, two scholars were pioneers in the field: F.-A. Gevaert, who identified 47 ‘themes’ based on the incipts of antiphons (Gevaert, 1895); and W.H. Frere, whose analysis of antiphons also took account of the interior phrases of their melodic structure (Frere, 1901–24).

Antiphon, §3: Origins and composition of melodies

(ii) Adaptation.

In the process of melodic adaptation, an existing antiphon melody was given a new text similar in shape and structure to its old one. There is thus a correspondence of melodic accentuation in the two texts. A melody used in this way is termed a ‘prototype melody’ and is cited by one of its texts, most often the first in order in the antiphoner. This process was extremely common: sometimes two successive antiphons in the same Office could be sung to the same melody, for example, the Veterem hominem antiphons for the octave of Epiphany (see CAO, i–ii, no.25) and three antiphons for Passion Sunday (no.66; see ex.2). The alleluiatic antiphons of Septuagesima, for the Office of ‘farewells’ to the alleluia, and of Paschaltide, which consist of numerous repetitions of the word ‘alleluia’, are other examples of the procedure. The oldest antiphoners, which lack neumatic notation, explicitly gave the incipts of the prototype antiphons as indications of the melodies to be used (CAO, iii, nos.1327–38).

The number of prototype melodies is fairly limited. There are none in the modes on D, except among the ‘O’ antiphons sung to the Magnificat during Advent. Some of the prototypes had more texts adapted to them than others. One of the most common prototypes in the antiphoner is in the 4th mode (ex.3, transposed up a 4th). This melody was used for a number of texts consisting of four clauses, and was adapted to them by syneresis (the contraction of occasional notes, indicated in the example with a horizontal bracket) or dieresis (the expansion of formulae with extra notes, here placed in parantheses). It is found with some 90 different texts in an antiphoner copied in about 1000 ce, even though it is totally absent from the series of antiphons to the Psalter, and no similar melody occurs in the Ambrosian repertory. This melody must have been utilized by the Gregorian composers when they needed to organize quickly a complete series, such as the antiphons for the weekdays of Advent and those of Lent. Of the 50 antiphons for the weekdays of Advent, 21 (i.e. 42%) are set to this prototype; the proportion is smaller for the weekdays of Lent (seven out of 56, or 12.5%) but almost as great for the weekdays of the two Passion weeks (16 out of 40, or 40%).

This prototype presents difficulties from the modal point of view, and was often criticized or even altered by theorists. Its modality is ambiguous, especially when it is adapted to short texts in which the note above the final does not occur at the end of the fourth phrase. The first two phrases lack a semitone; the third has a semitone, which creates a 1st-mode cadence (g–f–e–d–c–d); only in the fourth phrase is the mode clear, and even then only if the text is long enough to require the full number of notes. In the late 9th century the Commemoratio brevis proposed the use of the 2nd psalm tone rather than the 4th with this antiphon, a solution entirely justified when the antiphon text has a short fourth phrase, as in Benedicta tu and Ex Aegypto.

Another approach to this problem was proposed by Regino of Prüm and 12th-century Italian theorists, and was adopted in 1134 by the Cistercians. These theorists noted that the first phrase of the antiphon is identical to certain intonations of the 7th psalm tone (ex.4). Since any chant should be modally unified, it was proposed that this chant be modified so as to continue and conclude in the 7th mode. The Cistercians even went beyond the obvious logical changes required by these proposals and reshaped the melody completely, in the same way as they modified the antiphons using the tonus peregrinus (see §5(i) below).

Of the two prototype melodies in the F mode, the one set to the text Alleluia, Lapis revolutus est is used for several antiphons in Paschaltide. It may be of Gallican origin: similar chants framed by alleluias occur among Gallican fragments in certain antiphoners written in insular script, such as Alleluia, Quem quaeris mulier (CAO, iii, no.1350) and Alleluia, Noli flere Maria (no.1348). At a later date further texts of similar structure were set to the same melody, such as the following, which occur in German antiphoners: Alleluia, Resurrexit Dominus; Alleluia, Ego sum vitis vera; Alleluia, Quoniam in eternum; and Alleluia, Nimis exaltatus es (CAO, iii, nos.1352, 1342, 1351, 1347). This melody, recurring as a Paschaltide antiphon and frequently associated with the evening Offices, may suggest that some melodies are linked to a specific liturgical role. As another example, the 4th-mode melody discussed above (ex.3) appears in liturgical times that the medieval church considered to be sombre and penitential, that is, Advent, Lent and Passiontide. At least in some period of the development of the repertory, then, certain melodies must have been considered capable of conveying a special meaning or spirit appropriate to a particular liturgical observance.

The second prototype melody in the F mode is used for about 20 antiphons for Paschaltide and for six from the Christmas cycle, and is associated with the 6th psalm tone. Its structure, identical with that of the lesser responsories in Paschaltide, is characterized by repetition in its two phrases (see ex.5, with the text Ego sum vitis vera, alleluia, CAO, iii, no.2604). Notably, high festive seasons are the context for this melody.

There are three prototypes in the G mode for four-clause antiphons. That of the antiphon Tu es Petrus was adopted for some 60 other texts and uses the 7th psalm tone, whereas the other two use the 8th. The Ecce ancilla prototype occurs with about ten more texts, and the Omnes de Saba prototype with more than 20 different texts (see comparative table in Ferretti, 1934, pp. 112–13).

Antiphon, §3: Origins and composition of melodies

(iii) Centonization.

Another way in which new antiphons could be created without original composition in the proper sense has often been termed Centonization. According to Ferretti and others who have adopted his theory, this process occurred in the early liturgical repertories, in which the material was transmitted orally, and can only be fully understood in this light. The proponents of the centonization theory explain that once a psalm tone had been chosen (along with a psalm tone ending), possible formulae for the intonation came straightaway to the mind of the ‘composer’. If, for example, the first tone had been selected, together with the ending whose cadence is on the 3rd above the final of the mode, the choice of a readily compatible intonation would be possible (ex.6). After the intonation and some of the melody, the singer would reach an intermediate cadence. Intermediate cadences, which were given particular attention by theorists such as Pseudo-Odo (see Odo; also GerbertS, i, 257–8) or Wilhelm of Hirsau (GerbertS, ii, 172), are those that fall on the final of the mode or on a degree compatible with it.

The final cadence is the most important of all, for it introduces the intonation of the psalm and also stands at the close of the chant, thus defining the mode. It must therefore fall on the final of the mode. Nevertheless, the singer again would have a choice of several possible formulae (ex.7 shows the possibilities in the D mode). Of the formulae in ex.7 the last occurs only in later chants such as the antiphons Montes Gelboe (CAO, iii, no.3807) and Angeli archangeli (no.1398) for All Saints. The same cadential formulae occur in introits and communions, together with others that are more ornate despite being merely elaborations of simple cadences (ex.8). Many cadential and intonation formulae also appear in the same form in Byzantine chant (Huglo, 1966).

It would be possible to number the formulae used and then, by substituting numbers for the formulae, look at how chants were allegedly centonized; all possible combinations could be analysed by computer if the formulae were to be replaced by a series of numbers for each mode. However, it must be emphasized that Gregorian composition was not simply a matter of juggling formulae, for many and various factors could modify the structure of a chant. In particular, important words tended to attract to themselves the melodies to which they were sung in other chants, in spite of the change of context. This practice is common in repertories whose transmission is oral. For example, the word ‘ascendit’ in Assumpsit Jesus (CAO, iii, no.1501) is sung to the same figure as ‘ascendam’ in Vos ascendite (no.5493). Even the opening phrase of the antiphon may be affected by this practice (see ex.9a: In velamento clamabant and In velamento clamavi, CAO, iii, nos.3306–7; and 9b: the antiphons Dixit autem paterfamilias, Dixit autem Pater and Dixit Jesus discipulis, nos.2281, 2280 and 2296). Even a single syllable, regardless of sense, can attract the same melody, as occurs in the three antiphons Pater manifestavi, Vadam ad Patrem and Pacem relinquo (ex.10: CAO, iii, nos.4237, 5299 and 4205). Sometimes a similar melodic fragment may appear in chants of different categories (see ex.11, with the antiphon O vos omnes, CAO, iii, no.4095; and the responsory O vos omnes, CAO, iv, no.7303), or in antiphons of both Mass and Office (ex.12, with the antiphon Spiritus Domini, CAO, iii, no.4998; and the introit Spiritus Domini, in Hesbert, Antiphonale missarum sextuplex, no.106; and with the phrases ‘hoc autem’ and ‘hoc facite’ from the antiphon Solvite templum, CAO, iii, no.4982; and the communions Ultimo and Hoc corpus, in Antiphonale missarum sextuplex, nos.105 and 67b). These correspondences presuppose an amazing familiarity with the whole repertory, and they testify to the unified style of composition governing the redaction of the Gregorian gradual and antiphoner.

The concept of centonization in the strict sense has come under scrutiny as to whether it is an appropriate means of melodic analysis, and the validity of applying principles of the literary technique of centonization to musical compositions has been questioned. Scholars continue to seek a better understanding both of the vast expansion of the antiphon repertory and of its many areas of interrelatedness. László Dobszay and colleagues, through their extensive investigations of Central and Eastern European chant sources, have applied the principles of classification and analysis gained from the experience of systematizing folksongs. They have explored, for example, the process by which basic melodic structures can evolve into progressively more complex ones, according to textual needs, by the emergence of cadences from within the pitch structure: that is, a pitch that in one antiphon is merely a part of the melodic sequence assumes in another antiphon a new role as a cadential point or as an opening pitch in the subsequent phrase. When more musical material is needed to accommodate the text further, additional melodic elements are incorporated into the framework of the original sequence of pitches. This continuing process of understanding antiphon composition and melodic development involves collecting and ordering hundreds of antiphon melodies of varying length and complexity, and analysing and comparing them for aspects of internal stylistic articulation and other devices of composition.

Further aspects of compositional processes reflected in the more elaborate antiphons of the original Gregorian corpus, and of the evolution in musical style that began in the 10th century and became more marked in the 13th, are discussed below (§§5(iii) and (vi) in particular).

Seven main categories of Gregorian antiphon may be distinguished, and these can be grouped primarily according to their texts as related to liturgical function: antiphons of Matins, Lauds and Vespers; antiphons to the Psalter; antiphons to the Benedictus and Magnificat; Mass antiphons; Marian antiphons; processional antiphons; and versified antiphons. These, along with a smaller subgroup – antiphons with verses – are discussed individually below.

Antiphon

4. The Gregorian antiphons to the psalms at Matins, Lauds and Vespers.

The antiphons to the psalms at the major Hours of Matins, Lauds and Vespers are the most characteristic of the repertory and form the Gregorian antiphoner's basic corpus promulgated in the late 8th century. They were transmitted in two main branches of the antiphoner tradition: the secular and the monastic.

Secular antiphoners, used by diocesan canons, canons regular (those bound to a community rule, or regula) and Dominican and Franciscan friars, prescribe for Matins nine antiphons for the psalmody, distributed over three ‘nocturns’, each of which contains three psalms, lessons and responsories. Five antiphons are assigned to Lauds, and these are repeated at Vespers (except at certain festivals such as Christmas, St Peter, etc., when specific antiphons are laid down). Monastic antiphoners, on the other hand, which were used by members of monastic houses (Benedictines, Cistercians and Carthusians), prescribe 13 antiphons for Matins, distributed over three nocturns (6 + 6 + 1), five for Lauds and only four for Vespers.

Because of the differing structure and length of individual Hours in the two traditions, secular and monastic antiphoners agree in their series of antiphons only at Lauds, when both have five antiphons. Here, indeed, there is fairly close agreement, especially for the festivals of the Proper of the Time and of the Saints, as may be seen in Hesbert's concordance of autiphoners (CAO, i–ii). For Matins, monastic antiphoners give the nine antiphons of the secular antiphoners as their first nine; they then include a further four borrowed from the Common of the Saints for the feasts of the Proper of the Saints and from a series of interchangeable pieces for the feasts of the Proper of the Time. Occasionally the extra antiphons were composed following pre-existing models. An exception to the monastic–secular distinction occurs in the major Hours from Maundy Thursday Matins to Easter Sunday Matins inclusive, when, for this most sacred time of the Church year, an intentional uniformity of structure prevails in the Offices. The brief comments above partly explain the differences between secular and monastic antiphoners; no such differences characterize the graduals of the two traditions.

In modern editions of the chant it is, unfortunately, impossible to distinguish the antiphons of the original Gregorian repertory from those composed in later centuries. At least 200 or 300 years separate the two groups. Although composition may have proceeded along broadly similar lines during that time, the resulting forms are sometimes very different, as can be seen by comparing the 13th-century antiphons for the feast of Corpus Christi or those of contemporary versified Offices (e.g. of St Dominic or St Francis) with the old repertory of antiphons for the greater festivals of the Proper of the Time (Christmas, Easter and Pentecost) or of the Proper of the Saints (24 and 29 June, 10 August etc.). Study of the Gregorian antiphon must begin from the original repertory.

The texts of the antiphons of the original repertory consist of between one and four clauses. Some are very brief – between three and eight words – and are performed as a single phrase; some have two clauses separated by a caesura; some contain three clauses; and some, with four, seem to be modelled on the distich of Latin prosody. The following examples illustrate a variety of possibilities:

Omnis plebs ut vidit/dedit laudem Deo (two clauses: CAO, iii, no.4149); Dixit angelus ad Petrum/Circumda tibi vestimentum tuum/et sequere me (three clauses: no.2268); Hic vir despiciens mundum/et terrena triumphans//divitias coelo/condidit ore, manu (four clases: distich; no.3069); Solve jubente Deo/terrarum Petre catenas//qui facis ut pateant/coelestia regna beatis (four clauses: two hexameters; no.4981); In praesepio jacebat/et in coelis fulgebat//ad nos veniebat/et apud Patrem manebat(four clauses: assonanced prose; no.3272).

By contrast with this group of antiphons, the antiphons to the Magnifcat and Benedictus normally tend towards a greater number of clauses (see §5(iii) below). Other lengthier antiphon texts also exist, in particular those from the late 9th century onwards composed in honour of the patron saints of dioceses and of monasteries.

Antiphon

5. Other antiphons in the Gregorian repertory.

(i) Antiphons to the Psalter.

(ii) Antiphons with verses.

(iii) Antiphons to the ‘Benedictus’ and ‘Magnificat’.

(iv) Antiphons for the Mass: introits and communions.

(v) Marian antiphons.

(vi) Processional antiphons.

(vii) Versified antiphons.

Antiphon, §5: Other antiphons in the Gregorian repertory

(i) Antiphons to the Psalter.

The Psalter is distributed over the Divine Office of each week in ‘cursus’ or a generally consecutive order, and most of the antiphons used during its recitation are rather more simple than those assigned for specific feasts and seasons at Matins, Lauds and Vespers. Usually each psalm has its own antiphon with a text derived from that psalm, but some psalms are linked as pairs (e.g. Psalms lxii and lxvi in the Roman Office of Lauds) or as groups of three (e.g. Psalms cxlviii–cl, also at Lauds) under a single antiphon. In the Little Hours, on the other hand, Psalm cxviii is subdivided into 22 sections of eight verses each, with a particular antiphon for each section. The invitatory (Psalm xciv) is a special case (see §1 above and see Invitatory).

For the Offices of Sunday and the days of the week, the antiphon texts are drawn from the psalm; for the feasts of the Proper of the Time and of the Saints, the texts are of ‘ecclesiastical composition’ (i.e. they are non-biblical), but similar in spirit to Psalm xciv – in other words, an exhortation to praise.

In early psalters, the antiphons are located in relationship to the beginnings and ends of their psalms; in antiphoners, they are grouped together with a body of chants for use in the post-Epiphany period; in noted breviaries, they are found with the Psalter, usually at the front of the manuscript. The various invitatory melodies often occur in the final section of the antiphoner, sometimes together with the tonary, as was the custom in Cistercian antiphoners.

Most of the Psalter antiphons are very short and their melodies almost entirely syllabic. Many could be regarded as melodically bipartite but have texts that are virtually indivisible. It is perhaps the brevity of the texts that prevents melodic development as extended as in other antiphons, where the melody is able to rise to a climax and fall again over three or four phrases. The simple melodies of Psalter antiphons often consist of a few notes around a central pitch – sometimes the reciting note of the appropriate psalm tone – and confined within the range of a 4th or 5th. The simplicity of these melodies makes their transmission unreliable. An individual melody may have numerous variants; sometimes what is essentially the same melody has different modal assignments; and some antiphons exist in totally different versions. All of these factors pose many challenges for investigation of the sources.

When, in the late 8th century, the eight Gregorian psalm tones were adopted for the singing of the weekday Offices, it was found difficult to fit them to the Psalter antiphons, which more closely resemble Ambrosian antiphons than classical Gregorian ones. In Ambrosian practice, where the antiphon style itself determines the psalmody, the psalm tones had to be imposed on the antiphons indiscriminately, with the result that the tones and antiphons are at times at odds with one another. Even quite late in medieval times, theorists such as Frutolfus of Michelsberg (d 1103) continued to address the difficulties of classifying Psalter antiphons according to the eight-mode system. The musical evidence of those antiphons described as ‘of irregular tone’ should also be considered (see AM); although the tonaries and antiphoners designate the 4th psalm tone for these, Joseph Gajard in his edition prescribed a tone borrowed from Ambrosian psalmody, restricted in range to a 4th and better suited to them despite its lack of authenticity (ex.13).

The celebrated tonus peregrinus (‘wandering’ or ‘alien’ tone; the name appeared in the 12th century in Germany) is another irregularity within the system, since it has two different reciting notes and thus disobeys the rules of Gregorian psalmody. It was prescribed for six antiphons (ex.14). No doubt it was taken over into Gregorian psalmody from the Gallican rite, but it may ultimately be of Jewish origin (Huglo, 1971). It was adopted in the tonary not as a part of the 8th tone (as is commonly believed) but originally as an appendix, quite separate from the tonary proper and added after the last psalm tone ending of the 8th tone. The tonus peregrinus then, omitted from the antiphoner but recorded in the tonary, is further evidence of the conservative attitude of the organizers of Gregorian chant: in their unwillingness to touch the Psalter antiphons they even tolerated elements that disturbed the system.

Nevertheless, the tonus peregrinus antiphons were reshaped in various ways during the course of time in order to bring the final note into conformity with the intonation. The most far-reaching of these changes occurred in about 1134, when the Cistercians, pursuing their ideal of modal unity within each chant, modified the tonus peregrinus antiphons to fit the 1st psalm tone.

Very few additions were made over the centuries to the Psalter antiphons. For the Sunday Offices of the year, however, there is a series of versified antiphons, Pro fidei meritis (CAO, i, no.36; cf iii, no.4383), created perhaps by King Robert the Pious (d 1031), to whom some chants have been falsely attributed. This series has also been associated with Drogo, abbot of Bergues in Flanders; or, as another possibility, with an anonymous composer in northern Italy, since the series appeared first in Italy in the Silos Antiphoner (see CAO, i, iii).

Antiphon, §5: Other antiphons in the Gregorian repertory

(ii) Antiphons with verses.

Antiphons that are sung with verses occur in a few places, such as in three festive Offices in the Proper of the Saints: the two feasts of St Paul, 25 January (CAO, i–ii, no.47) and 30 June (ii, no.102), and that of St Lawrence, 10 August (CAO, i–ii, no.103). These antiphons are followed by verses reminiscent of the versus ad repetendum of the introit. Such verses are also found in the antiphons of Gallican chant, especially the Mandatum antiphons sung during the Washing of the Feet on Maundy Thursday, as well as in the Ambrosian repertory. A given verse, using the same melody as the psalm, is sung after the doxology following the psalm, and again after the repetition of the antiphon; the antiphon itself is performed at the end as a conclusion (see Gallican chant, §12.)

Antiphon, §5: Other antiphons in the Gregorian repertory

(iii) Antiphons to the ‘Benedictus’ and ‘Magnificat’.

The antiphons to the Gospel canticles of Lauds (Benedictus) and Vespers (Magnificat) are very similar to the other antiphons of Matins, Lauds and Vespers. Even though the Gospel canticles are sung to a more ornate psalm tone than the psalms and other canticles (attested in the late 9th-century Commemoratio brevis and in the tonaries), the musical style of the antiphons is not analogously ornate. They are distinguished, however, by their liturgical function and by their literary sources. The antiphons to the Gospel canticles are normally settings of texts drawn from or summarizing the Gospel of the Mass of the day; this series of antiphons was originally probably intended to provide two antiphons for each Sunday, but additions to it may have been made quite soon. The series of antiphons to the Gospel canticles for the Sundays in summer, whose texts are drawn from the Gospels of the Sundays after Pentecost, often gives more than two antiphons per Sunday (see CAO, i–ii, no.144). The absence of some of these antiphons from the original repertory is also suggested by the fact that a number are polymelodic (i.e. they exist with different melodies in different geographical regions). It is a fundamental rule of criticism that such pieces cannot have been present in a general original source. An example of this kind of chant is Scriptum est enim quia, an antiphon with a verse (CAO, iii, no.4836), which occurs in most manuscripts with an 8th-mode melody (AM, p.600), but in German antiphoners with a 1st-mode melody, and in Cluniac manuscripts and in the gradual-antiphoner of Brescia (GB-Ob misc.lit.366, f.260) with a 4th-mode melody. The proliferation of the Gospel canticle antiphons is also related to the fact that the series of Gospels of the day varied, especially towards the end of the liturgical cycle, from one church to another. Further, over time, Gospel canticle antiphons were provided not only for Sundays but for increasing numbers of feasts and saints’ days, and also for some ferias, especially in Advent and Lent. Since antiphons for the Benedictus and Magnificat constitute a minimum set of Propers for a given day, a great many of these antiphons were required, and the resulting repertory is large, numbering several hundred.

Some antiphons to the Gospel canticles from the Sundays in summer contain melodic formulae that do not occur in the antiphons discussed in previous sections. For example, Scriptum est enim quia cited above has a melisma on the first syllable of ‘docens’. This syllable is the fifth from the end of the chant, that which in prolix responsories generally bears a melisma. The same melisma occurs on the sixth syllable from the end of Dixit Dominus ad Adam, the Magnificat antiphon for Septuagesima. Since this antiphon replaced another from the alleluiatic Office of ‘farewell to the alleluia’, it is not an antiphon from the original corpus of the antiphoner. Comprehensive melodic analyses have prompted the observation that, indeed, most Gospel canticle antiphons do not use ancient melody types but particular adaptations of melodic elements from the older musical material (Dobszay, 1988); and also that there is considerable melodic interrelationship within this category of antiphon.

Some of the antiphons of this group have intermediate cadences on the note below the final (e.g. d–e–f–c–c in the 1st mode). These ‘lowered’ cadences are comparable with inverted cadences (e.g. f–g instead of g–f) and serve to avoid the monotony of several identical candences in an extended chant. Examples occur on ‘thronos’ in Qui coelorum (CAO, iii, no.4460) and on ‘altare’ in Si offers (no.4903). These cadences are followed by the usual re-intonations, or by others, for instance a–g–g–f–g–a on ‘et’, ‘ut’ or ‘ex-’ in the antiphons Exi cito, Vide Dominum sedentem, Si offers and Qui coelorum (CAO, iii, nos.2785, 5404, 4903 and 4460).

In a few passages in the Gospel canticle antiphons, it is not unreasonable to speak of descriptive music in which, for instance, a descending melody is used to suggest the idea of descent or of humility. Examples are found in the antiphon Descendit hic (ex.15: CAO, iii, no.2158); and in Homo quidam (no.3131), where the melody moves downwards at the words ‘descendebat ab Jerusalem in Jericho’.

The antiphons to the Magnificat for Saturdays in summer, unlike those for Sundays, are settings of texts from the Old Testament readings for Matins of the following day. This is true of the feasts of the Proper of the Time and of many in the Proper of the Saints. In the latter, however, the antiphons to the Gospel canticles, like other chants in the Office, may have texts drawn from the martyrology; or the Magnificat antiphon at Second Vespers might be a text summing up the significance of the feast, as in the Hodie antiphons (CAO, iii, nos.3088–124). The latter texts may have been suggested by Byzantine stichēra beginning with the word ‘sēmeron’ (‘today’). Interestingly, the word ‘sēmeron’ in the Byzantine stichēra has the same rising melody as is given to the word ‘hodie’ in the Latin antiphons (ex.16). This represents a further example of a key word attracting similar or virtually identical melodic contours even in different contexts (see §3(iii) above). In other Latin antiphons the texts are translations of Byzantine chants, but the melodies are not necessarily dependent on Byzantine models (Huglo, 1966). This is the case in the following chants honouring the Nativity and Mary respectively: the Benedictus antiphon Mirabile mysterium (CAO, iii, no.3763), and the Magnificat antiphon Nativitas tua (no.3852).

Finally, identifiable groups of antiphons associated with the Magnificat and the Benedictus have texts drawn from or paraphrasing the canticles themselves and assigned to the ferial Office. In the case of those for the Magnificat, for instance, this body of antiphons is found in greatest abundance in older antiphoners and in fewer numbers in later sources, suggesting that this repertory was set before the mid-9th century and that it preceded both the ferial Office as an organized entity in the antiphoners as well as the assignment of Propers for every Sunday (Udovich, 1980).

Antiphon, §5: Other antiphons in the Gregorian repertory

(iv) Antiphons for the Mass: introits and communions.

In the Western Church, the eucharistic service, or Mass, provided another context for the development and use of a repertory of antiphons. Responsorial psalmody was originally the sole form of psalmody at the Mass and survives, for example, in the gradual. It was replaced by antiphonal psalmody at the entrance of the presiding celebrant and at the distribution of Communion. The chant at the entrance is termed the antiphona ad introitum, or introit; that at the Communion is called the antiphona ad communionem, or communion. In the introit, the psalm verses are sung to a more ornate psalm tone than the ones used for the psalms and canticles of the Office. The intermediate cadences are accentual (governed by the rules of Latin accentuation). This type of psalmody also influenced the composition of the antiphons themselves, which at times bear traces of embellished psalmody (see Introit (i); for the connection of the offertory with psalmody see Offertory). As with the other categories of the Proper (except the alleluia), the introit repertory grew very little, except in south-west France and central and southern Italy. Troped introits appear from the 10th century on, with the tropes either preceding the introit or inserted within it.

The communion lost its psalmody between the 10th and the 12th centuries, according to region (see Huglo, 1971). The series of Roman communion antiphons is related to the two series of Ambrosian communion chants (the confractoria, chanted at the Fraction, and the transitoria, during Communion). A study of the Frankish–Roman communion series reveals that overall compositional planning was operative in the development of repertory and that some borrowing from the corpus of Office antiphons took place for this part of the Mass (McKinnon, 1992). The communion chants are less subject to troping than the introits and, from a modal point of view, are not as unified (see Communion).

Antiphon, §5: Other antiphons in the Gregorian repertory

(v) Marian antiphons.

Even though the antiphon is almost by definition a chant linked with a psalm, the Marian antiphons, an important group within the Gregorian repertory, are independent of psalmody. Sung in veneration of Mary in devotional contexts that were often attached in some way to Hours of the Office, these chants are often called ‘votive’ antiphons. In antiphoners and processionals, they usually occur together, located with the Proper of the Assumption (15 August). The group as a whole consists of some from the early repertory of antiphons as well as later ones in a newer musical style. From the 13th century to the present, designated Marian antiphons have been sung at the close of Compline, the last Hour of the liturgical day.

Some of the early Marian antiphons have very ancient texts. Sub tuum praesidium (CAO, iii, no.5041), for example, was a part of the Ambrosian liturgy, and its original Greek text (‘Hypo tēn sēn eusplanchnian’) survives in a 3rd-century papyrus (see Mercenier,1940). The antiphon Sancta Maria (CAO, iii, no.4703) is derived from a prayer concluding a sermon formerly attributed to St Augustine but now to Ambrose Autpert (d 781), abbot of St Vincent on the Volturno; its earliest possible date is the 9th century. A similar date may be assigned to Ascendit Christus, which is given for 15 August in Anglo-Norman manuscripts. This antiphon is not Gallican as has sometimes been suggested, since the text is from the letter of Pseudo-Jerome now attributed to Paschase Radbert (d c865), abbot of Corbie. These two antiphons occur, moreover, with different melodies in different regions, a fact that supports a late date for them. The same is true of a series of antiphons based on the Song of Solomon and set out according to the numerical order of the modes, which is included in the antiphoners on either 15 August (CAO, ii, no.106f) or 8 September. Some Marian antiphons, such as O virgo virginum, are used independently and also as antiphons for the Gospel canticle.

The large-scale Marian antiphons are the most important of this group: Alma redemptoris mater (with a text in hexameters; CAO, iii, no.1356), Ave regina caelorum (no.1542), Regina caeli (no.4597) and Salve regina. Each became associated with that segment of the liturgical year for which its text is most appropriate: respectively, Advent to 1 February; 2 February (Purification) to Wednesday of Holy Week; Paschaltide and Pentecost week; and Ordinary Time, from Trinity Sunday to the day before Advent. The Cistercians chanted Salve regina daily from 1218; the Dominicans at Bologna chanted it daily at Compline after a miracle in 1230, and the custom was adopted by the entire Order in 1250. The general chapter of the Franciscans at Metz in 1249 prescribed all four of these antiphons for Compline, though not in the same way as would the Roman breviary of 1568. Indeed, practice has varied considerably in this matter, as may be seen in Table 1, which shows distribution of the Marian antiphons in four churches in the 15th century.

Of the four great Marian antiphons, Alma redemptoris mater seems to be the oldest; its text, along with that of the Ave regina caelorum, may have been used originally with psalmody. The Regina caeli, linked with devotions to Mary in Paschaltide, appears always to have been independent, as was the Salve regina also. The latter, due to its widespread adoption in medieval times and its liturgical placement in Ordinary Time (which can take up nearly half a calendar year), is the best known of these four Marian chants.

The unanimous adoption of the Salve regina for Saturdays suggests that this antiphon was composed around the time when the Saturday votive Mass of the Virgin was developing during the 11th century. The literary style and vocabulary of its text (‘regina misericordiae’, ‘advocata’, ‘eia ergo’) reflect the spirit of this period, although it has been suggested that the date of its composition could be as late as the very early 12th century (Colette, 1992). Attributions of text or music to individuals such as Peter of Compostela (d c1002) or Hermannus Contractus (d 1054) are still debated. However, the musical style of the 3rd-mode melody in certain German manuscripts is quite different from music of the Office of St Afra by Hermannus. A 1st-mode melody for the antiphon, and one widely known, makes its earliest appearance in the pontifical of Aurillac (F-Pn lat.944). This melody resembles the musical style common in 11th-century Offices in that it uses the entire theoretical ambitus of the mode (here A to d). Motifs are repeated from one clause to the next, and there is a melisma of significance near the end of ‘O dulcis Maria’. It is possible that this version from the pontifical of Aurillac, which stems from the geographical area near Le Puy and St Martial de Limoges, may in fact represent the region of origin of the Salve regina.

Antiphon, §5: Other antiphons in the Gregorian repertory

(vi) Processional antiphons.

Another segment of the repertory lacking psalmody is the group known as processional antiphons. These are given in medieval manuscripts for certain occasions, including the feast of St Mark (25 April), the three Rogation days before Ascension Thursday, Palm Sunday and other days of Holy Week, Purification and some lesser celebrations. While the earliest sources may have as many as 90 or 100 such antiphons, this number diminishes to about half by later medieval times.

The oldest group is associated with the Major Litanies of 25 April, a procession at Rome which replaced a pagan procession on the same day (see Litany and Processional). An antiphon from this series, Deprecamur te, was cited by Bede and is alluded to in the biography of St Augustine of Canterbury. Augustine, sent to England by Gregory the Great (d 604), is said to have chanted this antiphon together with his companions as they entered Canterbury; and the whole series probably dates from this period (see Hesbert, Antiphonale missarum sextuplex, p.CXXI). Among the processional antiphons of Roman origin there are, however, some from the Gallican liturgy, such as the antiphon Collegerunt with its verse (op.cit., no.213b) and the other Palm Sunday antiphons, whose musical and literary style is distinct from that of the rest of the repertory.

Many of the processional antiphons tend to be musically lengthy and embellished with melismas, and their texts are longer than those of regular Office antiphons. The literary sources include oration-like formulae, Greek kontakia in translation, and what appear to be free compositions (see Bailey, 1971). In the Ambrosian repertory there is considerable overlapping between the rich corpus of processional antiphons and Office antiphons (Bailey and Merkley, 1989). Both musically and textually, and in all liturgical usages, this repertory has greatly varied origins and is bound together primarily by its common liturgical function.

Processional antiphons were originally copied at the end of graduals (see, for example, the early group of manuscripts for the Mass represented in Antiphonale missarum sextuplex, nos.200ff). From the 10th century onwards, however, they were transferred to a liturgical book of their own, the Processional.

Antiphon, §5: Other antiphons in the Gregorian repertory

(vii) Versified antiphons.

By the 9th century, and before in some instances, many antiphons began to reflect the influence of poetic techniques, including assonance and the use of metrical patterns. The texts were often taken from contemporary accounts of saints' lives which had been written in rhymed prose or with other poetic devices. Assembled into groups, these antiphons (along with responsories) functioned as hagiographical chants to be inserted, on the saints' feast days, into the usual places with psalmody and other regular elements in the Offices. Rhyme, though not invariably present (thus the term ‘versified Office’ is to be preferred to ‘rhymed Office’; see Versified Office), eventually emerged as the most prominent poetic feature in the construction of these antiphon texts. One of the earliest versified Offices is that for St Fuscianus, contained in the Mont-Renaud Antiphoner from the second half of the 9th century (PalMus, 1st ser., xiv,1931–6/R; see Jonsson, 1968).

In the 13th century, modal theory – by now fully developed – was applied to these chants in particular, with the result that the antiphons of versified Offices follow the numerical order of the modes. Also evidently cultivated within the Office at this time was a type of ‘crescendo’ effect, achieved through the progressive use of ‘beautiful’, ‘more beautiful’ and ‘most beautiful’ intervals (‘pulcher’, ‘pulchrior’ and ‘pulcherrimus’, as they were classified by Jerome of Moravia in his Tractatus de musica).

The antiphons of versified Offices evolved in musical style as well as in literary form, because new Office texts created to replace older ones were as a rule longer and more elaborate than those they supplanted; an example is the Office of St Martin composed by Odo, abbot of Cluny (Odo, §1). The melodies of versified antiphons are more symmetrical in structure than their non-versified counterparts and have somewhat stereotyped intonation and cadential formulae. The melodic range also tends to be greater than in the standard Gregorian repertory, with the complete octave more fully exploited and its contrasts of range highlighted. Melismas at the close of the Benedictus and Magnificat antiphons are similar to those in the final responsory of each nocturn at Matins. From the first half of the 12th century, the antiphons to the Gospel canticles were extended by a final melisma in the same mode as the antiphon and borrowed from modal formulae in the tonaries. This practice was attested by cathedral ordinals, by liturgists such as John Beleth (d c1165), Siccardo of Cremona (d 1215) and Guillaume Durand of Mende (d 1296), and by the English priest Alfred in Practica artis musice (1271), and was also popular in the late Middle Ages. These melismas or neumae were used as tenors in polyphonic works of the Notre Dame and Ars Nova periods. Much later, in the 16th century, such sung neumae were replaced by organ playing (Huglo, 1971).

Most of the versified Offices in which these antiphons occur honour local saints and were composed for dioceses or abbeys, generally for the occasion of the translation of relics or for a new dedication. Thus their area of diffusion was limited. They are often found in only one or two manuscripts, except where the saint was particularly popular. Examples of well-known Offices include that for St Nicholas composed by Regimbold of Eichstätt (966–91) – this saint was even more highly esteemed after the translation of his relics to Bari in 1087; and the Office of St Thomas of Canterbury, who was canonized in 1172, only two years after his death, and whose Office is found in nearly 300 extant manuscripts from across Europe (see Hughes, 1988). Most versified Offices, however, remained unknown outside their area of origin, for example, the Office of the Transfiguration by Peter the Venerable, abbot of Cluny (d 1156), which is preserved in a unique manuscript from St Martin-des-Champs (F-Pn lat.17716), and the chants in honour of the Holy Cross composed by Nicholas de Montiéramey, secretary to St Bernard, between 1140 and 1160 (GB-Lbl Harl.3073, ff.108f).

Antiphon

BIBLIOGRAPHY

general literature

mostly concerning Office antiphons

MGG1 (B. Stäblein)

MGG2 (E. Nowacki)

S. Bäumer: Geschichte des Breviers (Freiburg, 1895; Fr. trans., rev., 1905/R)

F.-A. Gevaert: La mélopée antique dans le chant de l'église latine (Ghent, 1895/R)

P. Wagner: Einführung in die gregorianischen Melodien: ein Handbuch der Choralwissenschaft, i: Ursprung und Entwicklung der liturgischen Gesangsformen bis zum Ausgang des Mittelalters (Leipzig, 2/1901, 3/1911/R; Eng trans., 1901/R); ii: Neumenkunde: Paläographie des liturgischen Gesanges (Fribourg, 1905, 2/1912/R); iii: Gregorianische Formenlehre: eine choralische Stilkunde (Leipzig, 1921/R)

W.H. Frere, ed.: Antiphonale sarisburiense (London, 1901–24/R)

P. Ferretti: Estetica gregoriana ossia Trattato delle forme musicali del canto gregoriano, i (Rome, 1934; Fr. trans., 1938); ii completed and ed. P. Ernetti as Estetica gregoriana dei recitativi liturgici (Venice, 1964)

H. Hucke: Untersuchungen zum Begriff ‘Antiphon’ und zur Melodik der Offiziumsantiphonen (diss., U. of Freiburg, 1951)

H. Hucke: ‘Die Entwicklung des christlichen Kultgesangs zum Gregorianischen Gesang’, Römische Quartalschrift für christliche Altertumskunde und Kirchengeschichte, xlviii (1953), 147–94

H. Hucke: ‘Musikalische Formen der Offiziumsantiphonen’, KJb, xxxvii (1953), 7–33

R.-J. Hesbert, ed.: Corpus antiphonalium officii [CAO] (Rome, 1963–79)

W. Lipphardt: Der karolingische Tonar von Metz (Münster, 1965)

M. Huglo: Les tonaires: inventaire, analyse, comparaison (Paris, 1971)

J. Claire: ‘Les répertoires liturgiques latins avant l'octoéchos, I: L'office férial romano-franc’, EG, xv (1975), 5–192

R. Taft: The Liturgy of the Hours in East and West: the Origins of the Divine Office and Its Meaning for Today (Collegeville, MN, 1986)

J. McKinnon, ed.: Music in Early Christian Literature (Cambridge, 1987)

A. Turco: ‘Melodie-tipo e timbri modali nell' “Antiphonale romanum”’, Studi gregoriani, iii (1987), 191–241

L. Dobszay: ‘Experiences in the Musical Classification of Antiphons’, Cantus Planus III: Tihány 1988, 143–56

L. Dobszay and G. Prószéky: Corpus antiphonalium officii ecclesiarum centralis Europae: a Preliminary Report (Budapest, 1988)

J. Dyer: ‘Monastic Psalmody of the Middle Ages’, Revue bénédictine, xcix (1989), 41–74

J. Dyer: ‘The Singing of Psalms in the Early-Medieval Office’, Speculum, lxiv (1989), 535–78

R. Crocker: ‘Chants of the Roman Office’, The Early Middle Ages to 1300, NOHM, ii (2/1990), 146–73

D. Hiley: Western Plainchant: a Handbook (Oxford, 1993)

T. Bailey: Antiphon and Psalm in the Ambrosian Office (Ottawa, 1994)

special studies

C. Marbach: Carmina scripturarum scilicet antiphonas et responsoria ex sacro scripturae fonte in libros liturgicos sanctae ecclesiae Romanae derivata (Strasbourg, 1907/R)

F. Mercenier: ‘La plus ancienne prière à la Sainte Vierge: le “Sub tuum praesidium”’, Questions liturgiques et paroissiales, xxv (1940), 33–6

J.M. Canal: Salve regina misericordiae: historia y leyendas en torno a esta antífona (Rome, 1963)

O. Strunk: ‘The Latin Antiphons for the Octave of the Epiphany’, Mélanges Georges Ostrogorsky, ed. F. Barisic, ii (Belgrade, 1964), 417–26; repr. in Essays on Music in the Byzantine World, ed. O. Strunk (New York, 1977), 208–19

M. Huglo: ‘Relations musicales entre Byzance et l'Occident’, Byzantine Studies XIII: Oxford 1966, 267–80

V. Fiala: ‘Eine Sonderform der O-Antiphonen’, Archiv für Liturgiewissenschaft, xii (1970), 261–7

T. Bailey: The Processions of Sarum and the Western Church (Toronto, 1971)

J. Udovich: ‘The Magnificat Antiphons for the Ferial Office’, Journal of the Plainsong and Mediaeval Music Society, iii (1980), 1–25

M. Huglo: ‘Le répons-graduel de la messe: évolution de la forme, permanence de la fonction’, Schweizer Jb für Musikwissenschaft, new ser., ii (1982), 53–73

D.G. Hughes: ‘Variants in Antiphon Families: Notation and Tradition’, IMSCR XIII: Strasbourg 1982, ii, 29–47

R. Steiner: ‘Antiphons for the Benedicite at Lauds’, JPMMS, vii (1984), 1–17

R.L. Crocker: ‘Matins Antiphons at St. Denis’, JAMS, xxxix (1986), 441–90

T. Bailey and P. Merkley: The Antiphons of the Ambrosian Office (Ottawa, 1989)

T. Bailey and P. Merkley, eds.: The Melodic Tradition of the Ambrosian Office: Antiphons (Ottawa, 1990)

M.N. Colette: ‘Le “Salve Regina” en Aquitaine au XIIème siècle: l'auteur du Salve’, Cantus Planus IV: Pécs 1990, 521–47

R. Steiner: ‘The Parable of the Talents in Liturgy and Chant’, Essays in Musicology: a Tribute to Alvin Johnson, ed. L. Lockwood and E. Roesner (Philadelphia, 1990), 1–15

S. Gasser: ‘Les antiennes O’, EG, xxiv (1992), 53–84

R. Steiner: ‘Marian Antiphons at Cluny and Lewes’, Music in the Medieval English Liturgy, ed. S. Rankin and D. Hiley (Oxford, 1993), 175–204

J. Halmo: Antiphons for Paschal Triduum-Easter in the Medieval Office (Ottawa, 1995)

E. Nowacki: ‘Antiphonal Psalmody in Christian Antiquity and Early Middle Ages’, Essays on Medieval Music in Honor of David G. Hughes, ed. G.M. Boone (Cambridge, MA, 1995), 287–315

mass antiphons

R.-J. Hesbert, ed.: Antiphonale missarum sextuplex (Brussels, 1935/R)

J.A. Jungmann: Missarum sollemnia: eine genetische Erklärung der römischen Messe (Vienna, 1948, 5/1962; Eng. trans., 1951–5/R as The Mass of the Roman Rite)

J. Froger: Les chants de la messe aux VIIIe et IXe siècles (Tournai, 1950)

R. Crocker: ‘Chants of the Roman Mass’, The Early Middle Ages to 1300, NOHM, ii (2/1990), 174–96

J. McKinnon: ‘The Eighth-Century Frankish-Roman Communion Cycle’, JAMS, xlv (1992), 179–227

D. Hiley: Western Plainchant: a Handbook (Oxford, 1993)

versified offices

FasquelleE (‘Offices rhythmiques’)

G.M. Dreves and C. Blume, eds.: Analecta hymnica medii aevi [AH], v (Leipzig, 1889/R); xiii (1892/R); xvii–xviii (1894/R); xxiv (1896/R); xxv–xxvi (1897/R); xxviii (1898/R); xlia (1903/R); xlva (1904/R)

R. Jonsson: Historia: études sur la genèse des offices versifiés (Stockholm, 1968)

M. Lütolf, ed.: Analecta hymnica medii aevi: Register (Berne, 1978)

A. Hughes: ‘Chants in the Rhymed Office of St Thomas of Canterbury’, EMc, xvi (1988), 185–201

A.M. Nilsson: ‘Some Remarks on Melodic Influences in Swedish Rhymed Offices’, Cantus Planus VI: Éger 1993, ii, 635–51

A. Hughes: Late Medieval Liturgical Offices (Toronto, 1994–6) [with computer diskettes]

electronic resources

A number of computer-assisted reserach projects have been established to enable scholars to locate chants and texts in various sources and to compare bodies of repertory. The most important are listed below, together with the name of the institution at which each project was first established, the name of the scholar responsible and the area of repertory concerned.

databases with particular focus on the antiphon; data also available in published form

Corpus Antiphonalium Officii-Ecclesiae Centralis Europae [CAO–ECE] (Budapest, Hungarian Academy of Sciences; L. Dobszay) [antiphoners of Central and East European origin]

CANTUS: a Database for Gregorian Chant (Washington DC, Catholic U. of America; R. Steiner) [on-line database of indexes of antiphons and other chants in important Office sources of European provenance; some material pubd in Musicological Studies, lv/1–, Ottawa, 1992–]

Late Medieval Liturgical Offices [LMLO] (U. of Toronto, Pontifical Institute of Mediaeval Studies; A. Hughes) [computer diskettes with selected versified Offices texts issued with Late Medieval Liturgical Offices, Toronto, 1994–6]

other materials previously in manuscript and/or in print

Antiphonale Missarum Sextuplex [AMS] (U. of Regensburg; D. Hiley) [on-line database of R.-J. Hesbert's text concordance for Mass Propers, Rome, 1963–79]

Thesaurus Musicarum Latinarum [TML] (Indiana U; T.J. Mathiesen) [on-line database of medieval and Renaissance theoretical treatises, many pertaining to chant and some particularly to antiphons]