11 Music and liturgy
One of the earliest surviving accounts of Christian liturgy is a report of Sunday Eucharist written at Rome around AD 150 by Justin Martyr : 1
And on the day named for the sun there is an assembly in one place for all who live in the towns and in the country; and the memoirs of the Apostles and the writings of the Prophets are read as long as time permits. Then when the reader has finished, he who presides speaks, giving admonishment and exhortation to imitate those noble deeds. Then we all stand together and offer prayers. And when, as we said above, we are finished with the prayers, bread is brought, and wine and water, and he who presides likewise offers prayers and thanksgiving, according to his ability, and the people give their assent by exclaiming Amen. And there takes place the distribution to each and partaking of that over which thanksgiving has been said, and it is brought to those not present by the deacon.
A description of a solemn mass as celebrated at Rome some six centuries later runs to several thousand words. The account of the introit , cited here in abbreviated form, is in itself longer than the entirety of Justin Martyr's earlier description. 2
Then [the choir] rises up and passes in order before the altar, and the two rows arrange themselves in this manner: the men-singers on either side without the doors [of the presbytery], and the children on each side within. Immediately the precentor begins the anthem for the entry: and when the deacons hear his voice, they at once go to the pontiff in the sacristy. Then the pontiff, rising, gives his right hand to the archdeacon, and his left to the second [deacon], or whoever may be appointed: who, after kissing his hands, walk with him as his supporters…After this the pontiff passes on, but before he comes to the choir the bearers of the candlesticks divide, four going to the right and three to the left; and the pontiff passes between them to the upper part of the choir, and bows his head to the altar…Then turning towards the precentor, he signs to him to sing, Glory be to the Father, and to the Son etc.; and the precentor bows to the pontiff, and begins it…
The differences between the two accounts are striking. Justin Martyr gives no explicit indication that music was used as part of the Eucharist, whereas in the later account a separate body of singers is assigned a distinct role within an elaborate ceremony. There is also a marked difference between an apparent fluidity – according to Justin, readings last ‘as long as time permits’ – and the detailed prescriptions for all aspects of ceremonial conduct in the later document. It is tempting on the basis of such reports to propose a history of music and liturgy that passed from simplicity to complexity, from fluidity to fixity. The problem with such a linear approach is that the surviving documents from this era present a partial and highly inflected picture of liturgical practice . 3 Justin Martyr's description, for example, occurs within the context of a letter to the Emperor Antoninus Pius in defence of Christianity; the absence of reference to ceremony and vagueness over certain details can be read as much as an attempt to present Christian worship as both sober and reasonable as a summary of the celebration of the Eucharist at that time. There must also remain doubts as to whether Justin Martyr is describing the Eucharist as practised at a specific institution, a standardized pattern abstracted from Roman practice, or a generic outline of Christian custom.
For all their differences, the points of agreement between these two Roman witnesses to the celebration of the Eucharist before the mid eighth century may be taken as a starting point for understanding the ways in which a shared framework was realized under differing conditions. 4 As can be seen from Table 11.1 , what is common to the two reports is a basic pattern of readings followed by prayer from the part of the service that has come to be known as the fore-mass or Liturgy of the Word, that is, up to but not including the presentation of gifts. From the Liturgy of the Eucharist, the elements of presentation, Eucharistic prayer, and communion are shared. 5
There would appear to be little continuity between the two reports concerning music, but when Justin Martyr's account is read alongside other extracts referring to music in the Eucharist during the early centuries of Christianity then a more varied picture emerges. For what may be detected in broad terms is a shift in kind from music as an optional amplification of the word to music as a constituent element of a discrete liturgical act. The difference may be illustrated with respect to the gradual. Clear evidence for a psalm routinely sung in the fore-mass as a discrete liturgical event dates only from the later fourth century; before this time, it is more than likely that one of the readings in the fore-mass could be a psalm that was on occasion intoned in a more or less lyrical manner. 6 A comparable shift can be traced in the expansion at points of liturgical action without words, that is, at the entrance, the presentation of gifts and the distribution of communion. Before the later fourth century, it seems that services began rather abruptly (Justin Martyr, for example, makes no mention of any items of greeting), there is equally hardly any evidence for singing during the presentation of gifts, and only sporadic and late evidence for singing during communion. Yet from this time onwards, Eucharistic celebration was expanded at these points by similar means, the addition of a chant followed by a prayer, resulting in discrete liturgical items with music as a key component, namely introit, offertory and communion chants. 7
Several factors in the rise of the formalized expansions including music may be highlighted here. 8 With the Emperor Constantine's legalization of Christianity in 313, imperial financial support enabled a significant expansion in the building of public spaces for Christian devotion. Christians adopted a basilican style for their new buildings, which in practice meant the erection of large buildings resembling assembly halls, with a long central nave ending in an apse and with side aisles. 9 The size of such buildings enabled the increasing numbers of those attending worship to be accommodated, while the long naves made provision for the expanding sequence of events at the beginning of the service. Processions between buildings, making use of the public spaces within the city, became increasingly popular after Constantine's edict; in the course of time, the Kyrie, which had been used as a litany accompanying processions between churches, was added to the opening rites within the church, while the Gloria was first added on solemn occasions presided over by the bishop.
A second important factor in the formalized expansion of liturgical action was an enthusiasm for psalm singing that peaked in the late fourth century. 10 The psalms, which were extensively quoted by church fathers of the time, seem to have solved two related problems that faced the early church. Amid the general movement to larger-scale worship and more formalized patterns of ecclesiastical life, the psalms provided an inspired language that retained a personal, expressive appeal through their frequent invocation of first-person experience. Equally, at a time of controversy over core elements of Christian doctrine, and most especially over the nature of the Trinity, the psalms provided canonically approved texts for singing, in direct contrast to the hymns used as vehicles for spreading the ideas of heretical movements. An important consequence of this preference for psalmody was an adoption of the formal possibilities inherent in psalmody, whether direct (singing verses straight through), responsorial (in which a soloist sang the verses and the congregation responded with a refrain usually drawn from the psalm), or antiphonal. In the early centuries of Christianity, the congregrational refrain of responsorial psalmody was divided between two choirs, who responded alternately to soloists singing the psalm verses. Through most of the Middle Ages, however, psalm verses were sung antiphonally with successive psalm verses sung by two choirs positioned opposite each other. 11
A third important factor in the expansion of formalized worship was the transition from lector to schola chant. 12 In the early church, psalms were led by lectors, who were adolescent or even younger boys of varied musical training. The relatively low status of these lectors is mirrored by the fact that the choice of psalms for particular days, aside from the recurrence of specific and obvious correlations, appears to have been determined on a largely ad hoc basis before at least the later fifth century. By contrast, the development of scholae to perform the psalms appears to have been related to the development of cycles of sung texts assigned to specific feasts. The Roman schola cantorum , for example, appears to have been founded towards the end of the seventh century, and it is from the same century that there is evidence for a fixing of sung texts across the liturgical year. 13 As for the composition of the Roman schola, the later document quoted above mentions four chief cantors and a further body of male singers and boys. It would seem no coincidence that the same document records a fully elaborated version of solemn mass as found throughout the Middle Ages; in other words, a complex liturgy and the institutional means to perform and sustain it appear to have gone hand in hand. 14
***
And he directed that same clergy, abundantly instructed in divine law and in Roman chant, to follow the custom and ritual order of the Roman church, which up to that time had scarcely been done in the church at Metz. 15
So wrote Paul the Deacon in the first half of the 780s about the achievements of Chrodegang (Bishop of Metz, 742–66), but what are we to make of this reported transfer of music and liturgy from Rome to the heartlands of the Carolingian kingdoms north of the Alps? The underlying motivations, means and relative success of this cultural transfer have repeatedly been questioned, and will no doubt continue to fascinate historians of music, if only because the gap of almost a century before the earliest surviving substantial notated chant sources allows ample room for informed speculation. Paul the Deacon's account nevertheless directs us to the nub of the matter: a Frankish clergy instructed in Roman chant (cantilena romana ) was directed to observe the custom (mos ) and ritual order ( ordo ) of the Church of Rome. In other words, a Roman way of doing liturgy including music was to be imitated, a manner of celebration for which cantilena romana was not simply central but, as implied by Paul the Deacon's phrasing, a prerequisite.
Several documents shed light on the attempt by both Chrodegang and his successor to introduce a Roman form of music and liturgy, from which it would appear that the formation of a schola cantorum , an ordered clerical lifestyle, Roman architectural features and an imitation of the Roman use of ritual space were all intrinsic to the process of imitation. 16 By the mid ninth century, a pattern of worship for the mass through the year, revised from Roman models, is recorded in a sacramentary belonging to Charlemagne's son, Drogo (Bishop of Metz, 823–55). 17 Sacramentaries, or books recording the prayers used by the celebrant in the mass, are of crucial importance: containing the prayers used by the highest-ranking individual at the most important service, they were taken by the Carolingians as the book that provided the outline for the liturgical year. Drogo's Sacramentary is something of a special case in so far as it is a collection reserved for ceremonies presided over by the bishop; in containing texts only for the main feasts, it provides an ideal introduction to the framework of the liturgical year as it had developed by the mid ninth century ( Table 11.2 ).
The annual cycle recorded in Drogo's Sacramentary represents not only the outcome of centuries of development, but also the outline of the pattern that was to endure through the Middle Ages and beyond. The liturgical year it records is formed out of two main overlapping cycles celebrating events in the life of Christ (the Temporale or Temporal cycle) and Saints (the Sanctorale or Sanctoral cycle). The two main poles of the Temporale are the feasts of Christmas and Easter. By the fourth century, Easter was routinely celebrated on a Sunday with a following fifty-day festal season, in which the fiftieth day was associated with the gift of the Spirit ( Pentecost) and the fortieth day with Ascension. 18 The celebration of Christmas on 25 December was also established by the fourth century, but only in Rome and North Africa; Christmas was celebrated elsewhere on 6 January. The differing emphasis given to Christmas in these celebrations came to be reconciled as complementary celebrations of the nativity (25 December, Christmas) and the manifestation of Christ (6 January, Epiphany). Preparatory periods to the main feasts of Christmas and Easter are also first attested as widespread phenomena in the fourth century, with the emergence of both a Lenten forty-day fast and a short period of fasting before Christmas known as Advent, which seems to have been developed into a more extended observance sometime during the seventh century. By the mid ninth century, octaves of individual feasts were also celebrated, thus either the week of the feast was celebrated (for example the Easter octave, whose eight days continue through to the following Sunday), or the eighth day of a feast was celebrated in its own right (for example the octave of Christmas on New Year's Day, which was only later established in Rome as the Feast of the Circumcision). Specific days within the week were specified by a mainly numerical system: feria II (Monday, the second day of the week) through to feria VI (Friday), followed by sabbato (Saturday).
The commemoration of individual saints both on the anniversary of the date of their death and at the place of their burial dates back to the mid second century, but such celebrations were necessarily local and, given that under Roman law it was illegal to open graves in order to transfer relics, the spread of cults was inevitably slow. By the time of Drogo's Sacramentary, however, a fully developed Sanctorale was in place, celebrating saints of local importance (e.g. Saint Arnulf of Metz, supposed ancestor of the Carolingians), as well as saints of regional significance (such as Saint Remigius, so-called Apostle of the Franks) alongside widely recognized saints (e.g. Saint Stephen, the first Christian martyr). Material collected together in Drogo's Sacramentary immediately after the Sanctorale represents a further important category of liturgical material, that of the Common. In this section, which is not listed in Table 11.2 , were copied prayers for services of dedication and for groups of similar feasts that share texts; thus the commemoration of individual apostles, virgins and martyrs would draw upon general texts for this category of saints in the absence of any specifically composed texts. One final category of information requires explanation, namely the churches listed in association with each feast. A similar listing of churches, otherwise known as stational indications, is also found in the earliest unnotated antiphoners (chant books) dating from ca800 onwards. These refer not to local churches but to churches in Rome and, as is evident from a comparable list that survives for Metz from the late eighth century, the system of celebrating the Eucharist under the bishop at different churches within Rome was translated into a parallel system within Frankish towns and abbeys, with major feasts of the Temporale celebrated in the main churches, and feasts celebrating saints in lesser churches. 19
The liturgical cycle represented in Drogo's Sacramentary reflects not only Roman practice, but adaptation to Carolingian tastes. In specific terms, the model for this book is the Gregorian Sacramentary as requested by Charlemagne from Pope Hadrian (772–95), which was then revised according to the standards of classical Latin that were taken as normative in the Carolingian renaissance and to which were added texts for specific occasions and other liturgical formulas omitted from the Sacramentary in a supplement prepared by Benedict of Aniane (d. 821). 20 In more general terms, Drogo's book provides an early example of the accommodation between an imported Roman rite and its local implementation that remained a feature of liturgical practice through the Middle Ages. By the late ninth century, the Roman rite had spread across the Carolingian kingdoms, with other rites and their associated chant traditions used outside of this axis: chiefly, the Visigothic rite in what is now central and northern Spain and Portugal, the Ambrosian rite at Milan, the Beneventan rite from the region of Benevento in southern Italy, and the Byzantine rite in areas that once formed the eastern Roman Empire. By the end of the Middle Ages, the Roman rite had become so widespread in the Latin West that only the Visigothic rite as followed in Toledo and the Ambrosian rite used in Milan remained.
In contrast to the mass, the divine office was rooted in the passing hours of the day, marking out the cycle from darkness to light and back again. Since the medieval history of the office is far more complex than that of the mass, only the most general observations will be made here. 21 Crucial for understanding the later development of the office is the fact that, by the fourth century, two separate traditions of daily prayer can be identified in the West: ‘cathedral’ practice, the normal practice of local Christian churches, and ‘monastic’ practice, as observed both by individual ascetics and by monks in urban monastic centres. 22 The ‘cathedral’ office consisted of daily morning and evening worship and was characterized by a selective use of psalms and hymns; the nucleus of morning prayer, for example, was the Laudate psalms (nos. 148–50), which led to the name lauds for this service. In addition, a variety of occasional vigils were followed within this tradition. An alternative model of daily prayer was followed in many urban monastic communities, where a fivefold pattern of daily prayer corresponding to the principal divisions of the day in the Roman Empire was the normal pattern, thus morning, third hour, sixth hour, ninth hour and evening. Night prayer was also observed. This ‘monastic’ practice drew on both the selective approach to psalmody of the ‘cathedral’ office and the more or less continuous recitation of psalms practised in desert monasticism. Saint Benedict (ca480–550) was not the first to combine these two traditions, but it was his achievement to draw them together in a way that was not only manageable but also lightened previously sober Roman practices through the introduction of hymns.
The pattern of the daily office outlined in Benedict's Rule was initially only one among many patterns encoded in monastic rules, but by virtue of its promotion under the Carolingians it achieved widespread (if not uniform) adoption in the West by the ninth century. For all its combination of traditions, Benedict's pattern nevertheless remained a monastic use distinct from what was known from the twelfth century onwards as ‘secular’ use, that is, the liturgical pattern followed by clergy not bound by vows to a religious order, such as those serving in collegiate and parish churches. The main difference between the two traditions was that monastic use was more elaborate, especially in the nocturns of Vigils (see below). Both uses nevertheless had their roots in Roman monastic practice and were treated as part of the Roman rite which spread rapidly from the ninth century onwards.
Any attempt to provide a summary of the structure of the office is complicated by the fact that variation was intrinsic to its design. The pattern of worship changed in response not only to the nature of the institution (monastic or secular), but also to the natural cycle with its changing hours of daylight and seasons, and to the many-layered liturgical cycle with its observation of ferial (daily) and festal worship, itself divided into Temporale and Sanctorale, proper and common. When extensive local variation and development over time is added to this mixture, the only way to proceed is to provide a summary based on a single document. Although lacking certain indications and in all probability a combination of varied sources, one of the earliest full accounts of the structure of the office is provided by Amalarius of Metz (ca775–ca850). 23 His outline (given in Table 11.3 ) combines features from the Rule of Benedict with the Roman (secular) rite, and remained at the core of the pattern of daily offices up to the reforms of the twentieth century.
Certain aspects of Table 11.3 require further elucidation. A versicle is a short dialogue between the minister and those present; antiphons and responsories consist of short texts taken, at least in the Roman tradition, from scripture, sung (respectively) either in conjunction with a psalm or canticle, or after readings. The Gloria patri is a Trinitarian doxology (Glory be to the Father) customarily sung at the end of psalms and canticles, and the preces are a sequence of prayers later fixed as comprising a lesser litany or Kyrie, the Lord's Prayer, and versicles and responses. Although not always mentioned by Amalarius, each office apart from Vigils ended with a collect or prayer (variable or fixed) and a blessing ( Benedicamus Domino). It should also be noted that later developments standardized a difference between monastic nocturns (on Sundays, three nocturns with twelve lessons and responds, with Old Testament canticles replacing psalms in the third nocturn) and secular nocturns (on Sundays, three nocturns with nine lessons and responds), as well as replacing the last response in the final nocturn by a Te Deum.
Leaving to one side the many daily variations on this pattern, what may be appreciated by reading across the columns in Table 11.3 is that the offices observed a basic structure of interspersed singing and reading, concluded by prayers, that is, in the most general of terms, the same elements of worship as seen in the fore-mass. As for the realization of this pattern of worship on any given day, Vigils was held during the night and was followed almost immediately by lauds, which took place at dawn; prime, terce, sext and nones (the Little Hours) took place at the first, third, sixth and ninth hours of the day respectively; vespers was held at dusk and Compline took place before retiring to bed. The distribution of psalms should also be noted: psalms 1 to 108 were assigned to Vigils, while psalms 109 to 150 were reserved for vespers. The complete Psalter could thus be recited through the week at these two services; those psalms with fixed assignments to lauds and the Little Hours were excluded from this cycle.
To appreciate quite how varied the realization of this framework might be, it is worth recalling Amalarius's own trials in trying to bring order to the traditions of chanting the office. 24 A vivid picture of prevailing diversity is painted by his report of a journey to Rome in 831 to compare the antiphoner in use at Metz to the Roman Antiphoner. In Rome, Amalarius was referred to Corbie, to where, several years previously, the antiphoner of Pope Hadrian I was sent as a present for Abbot Wala (the cousin of Charlemagne). Amalarius returned to Corbie, where he compared the two antiphoners and was surprised to find that the two repertories differed in construction, texts, and in the number and order of antiphons and responds. Faced with such diversity, he adopted the approach of a compiler, indicating more recent Roman additions by R, whereas those introduced by Chrodegang to Metz he indicated by M; he also indicated numerous personal options by the letters IC. For all that Amalarius represents an early stage in the history of the office, the difference between two traditions supposedly directly linked and the provision of a number of texts for individuals to choose between remained a feature of the office throughout the Middle Ages.
While the basic structures of worship used throughout the Middle Ages were in place by the end of the ninth century, the subsequent history of medieval music and liturgy was characterized in the most general terms by expansion, most notably in the proper. This was achieved in part through an exponential increase in the number of feasts, especially those commemorating individual saints, to the extent that by the end of the Middle Ages few days in the whole year were without a festal observance in the Roman Missal. New departures of particular interest in this expansion included the elaboration of offices with newly composed texts and music both for individual saints and for particular feasts. 25 Expansion in music for the mass occurred primarily through accretions to existing liturgical items, principally by means of the introduction of tropes and sequences. The addition alongside chant of what were often poetic sung texts, but could also be solely musical interpolations, took place soon after the introduction of Roman chant into the Carolingian kingdoms; indeed, given reported difficulties in learning Roman chant, it is quite possible that these so-called accretions were introduced alongside the formalization of the repertory through the ninth century. While even a short introduction to tropes and sequences lies beyond the scope of this survey, what may be said is that the texts served in general to expound the significance of the feast being celebrated. 26 Many tropes, for example, begin with the word ‘Hodie’ (‘Today’), before making explicit the themes of the day that had often remained implicit in the base chant texts. 27
Although it is possible to identify an overall tendency to prolixity, the development was by no means linear or uniform, as may be illustrated by two examples from the era of monastic reform, the tenth and eleventh centuries. The Benedictine monastery at Cluny, which lies near Mâcon in Burgundy, was founded near the beginning of the tenth century: while operating independently of any external control, it developed a rich pattern of life and worship, the grandeur of which was symbolized by the abbey church itself, which remained the largest church building in Western Christendom until the sixteenth century. As for its liturgy, the abbey of Cluny and the international network of houses that shared its customs were famed by the eleventh century for a seemingly ceaseless liturgical round, yet the repertory of sung items at Cluny appears to have been conservative for it appears that tropes were scarcely used and that texted sequences were sung only on the highest feasts. 28 Liturgical expansion was realized instead through intercessions offered for specific intentions, whether commemorations of the dead, or the recitation of additional psalms (psalmi familiares ) at each of the canonical hours for patrons, benefactors and friends, or the requirement for each monk to celebrate mass daily.
A broadly contemporary movement stemming from Gorze was characterized by a full use of tropes and sequences. 29 The abbey of Gorze, which lies some nine miles to the southwest of Metz, had been founded by Bishop Chrodegang as a Benedictine house and was revived by Adalbero (Bishop of Metz, 929–62) with the intent to reclaim the former glory of Metz through a return to strict adherence to the Benedictine Rule and a renewed intent to spread Messine chant. This reform proved so successful that the customs of Gorze spread throughout monastic houses in Lotharingia and the Eastern kingdoms ruled by the Ottonians. A certain degree of uniformity between houses associated with Gorze was ensured by the fact that it was the abbot (who was appointed from outside the community and remained in touch with the major centres of reform), rather than the cantor, who decided which chants were to be sung. Even so, there is little evidence to suggest a centrally coordinated reform of liturgy and chant within any of the monastic reform movements associated with individual houses in the tenth and eleventh centuries.
While musical and liturgical exchange during the tenth and eleventh centuries appears to have occurred largely through individual networks of association, overall development was shaped by a shared concern to seek renewal through a return to the Benedictine Rule, albeit as filtered through the Carolingians, under whose influence manual labour was replaced by increasingly elaborate communal worship. The Cistercian movement of the twelfth century stands out in sharp relief against this background, for the monks who settled at Cîteaux (near Dijon) in the closing years of the eleventh century sought a return to the letter of the earlier Benedict's Rule and thus to a pattern of worship free of the accretions of intervening centuries. Unlike earlier reform movements, there was at the same time a centrally coordinated attempt to revise the melodic repertory by returning to supposedly authentic musical sources, which in practice meant an attempt to reform chant on the basis of a gradual and antiphoner from Metz (whose legend of exemplary Roman practice endured) and a hymnal from Milan (the seat of the father of Latin hymnody, Saint Ambrose). 30 The results, however, were soon found to be unsatisfactory, in no small part because of the inadequate state of the sources used, prompting a second reform in the mid twelfth century based on a different model of authority. In this instance, the emphasis was placed not so much on authentic texts as on an intent to apply a priori rules in the reform of chant, namely that authentic and plagal modes should be kept separate, chants should not exceed a ten-note range, B flats were to be avoided in notation, and melismas should be cut down. Despite the apparent severity, it should be remembered not only that Cistercian chant reforms were not applied systematically, but also that the overall aim of Cistercian reform was to promote a simple and sober manner of worship that would encourage both unity and inner devotion.
A growing emphasis on individuals and inner experience can be detected in other movements with their roots in earlier medieval practices that achieved widespread prominence by the twelfth century. The shift can be traced most clearly with respect to the celebration of the mass and can be divided into two related tendencies. The period from the ninth to the twelfth century saw a rapid expansion in ‘private masses’, in which ceremonial splendour was reduced to a minimum (although it would still seem extensive to many modern observers): these masses, which made little or no use of music, were held by smaller groups in addition to the main community celebration as an act of intercession (often for the departed, or as part of devotion to the Blessed Virgin Mary). 31 In apparent opposition to this movement, there was a parallel tendency to expansion in all aspects of ritual gesture, including genuflexion, kissing the altar, kissing the Gospel, signing the cross and beating the breast. What these two tendencies shared was a growing conception of the mass as an act of piety offered by the priest on behalf of the assembled people, a notion expressed in the increasingly common practice of the priest celebrating facing East and himself reciting all parts of the mass, even when these were duly executed by the proper ministers or the choir. At the same time, physical separation increased between on the one hand the priest and attendant ministers, whose observances mainly took place around the altar in the sanctuary, and on the other hand the people, gathered in the nave, and separated not only by the chancel, in which were seated the choir, but increasingly by choir screens. The main participation open to the laity in such circumstances was through interior devotion, a state of affairs which goes some way towards explaining the growth in popular and affective practices centred on Mary and Jesus from the eleventh century onwards. 32
A parallel emphasis on both specialized roles for select individuals within worship and an inner devotion that was the counterpart of an ever more elaborate ritual spectacle can be traced in the rapid expansion of polyphony in the liturgy up to the thirteenth century. Polyphony had been cultivated within worship for many centuries previously: indeed, any attempt to speak of an introduction of polyphony into the liturgy has to be tempered by the fact the earliest witnesses to polyphony are of the same date as the earliest notated repertories of monophony, namely the late ninth century. With that said, the polyphony recorded at this date is not a codified liturgical repertory, but examples of performance practice in treatises of music theory. Indeed, there has to be caution in treating these examples as polyphony and drawing a straight line between them and the repertories of the later Middle Ages, for what is recorded in these ninth- and tenth-century treatises is a method for amplifying a single line. As for the liturgical role of polyphony, it is hard to extract a picture of wider practice from the selection of office antiphons, psalm tones and a portion of the Te Deum contained in the earliest theory treatise to include polyphonic examples. 33 In one sense, however, the paradigmatic appearance of polyphony in the earliest treatises is significant in itself, for all the indications are that an earlier multiplicity of performance options only gradually crystallized into discrete items. One of the key features in this development was the elaboration and increased density of compositional rules such as can be seen in the eleventh-century repertories of polyphony from Winchester and Chartres. While both these centres are cathedrals, it should be remembered that polyphony was also sponsored by monasteries: Winchester itself was staffed by Benedictine monks under an abbot-bishop; polyphony was also practised at the substantial Benedictine monasteries of Fleury (reformed via Cluny) and St Maur-des-Fossés in Paris. Rather than being characterized by any divide by personnel or institutional type, polyphonic experimentation, formalization and latterly codification took place at larger institutions associated with the new currents of learning fostered especially in northern France during the eleventh century.
When considered as an element of liturgy, it is the ceremonial grandeur of polyphony that is most impressive. The execution of polyphony by soloists wearing silk copes on major feasts of the liturgical year at Notre Dame in Paris sits easily alongside the use of extra candles, incense and the richest vestments and vessels on such occasions. The element of outward display was further heightened by the use of polyphony originally associated with other liturgical functions (such as alleluias and responsories) in processions. Turning to stylistic developments customarily associated with the twelfth century, the consistent placing of the polyphonic voice above the base chant might also be considered a feature of increased outward display, the prominence of the new line a far cry from monastic prohibition against self-promotion and placing chants too high in the voice. In a related vein, the twelfth century also saw the development of techniques of extension and elaboration in the form both of polyphonic pieces with a highly melismatic voice above a slower-moving plainchant (organum purum ) and compositions with up to four separate voices, thus announcing a decisive break with the previous insistence on music as a projection of the text akin to reading. At the same time, innovation was announced in polyphonic repertories by an increasingly self-confident distancing from liturgically assigned material, whether in the use of non-liturgical tenors in the Aquitanian repertory or in the increased expansion around the edges of the liturgy in the Parisian repertory in the form of the conductus. As for its underlying motivation, the development of polyphony has frequently been linked to Gothic splendour and a new-found urban self-confidence. Less often mentioned is the extent to which polyphony also served the immediate needs of the church, for liturgical polyphony soon centred on those parts of the chant repertory traditionally reserved for more elaborate singing by soloists (chiefly the verses of graduals, alleluias and responsories), thereby continuing, while also renewing, the functions already established in monophony. Viewed from a different perspective, polyphony served to foster the growing interest in inward and affective devotion, especially among the increasingly distinct and numerous crowd whose place it was to listen. The justification provided for such a new departure was an equally new-found emphasis on the pastoral role of the church, for polyphony was concentrated on those services to which the people of Paris had access and may be interpreted as ‘a conscious attempt by the canons to demonstrate their superior abilities, or those of their subordinates, and to heighten the spirituality of the common folk’. 34
The polyphonic repertory developed at Paris rapidly spread across Europe and remained the most elaborate liturgical music in circulation until the development of the Ars Nova in the early fourteenth century, yet during the same period it was Rome that led developments in liturgy. Symptomatic in this regard is the attitude taken by the new preaching orders in seeking to adopt new liturgies. A century after the Cistercians, the Dominicans turned to Metz in search of an authentic and standardized liturgy; they soon realized, however, that despite the continuance of legends, Metz no longer maintained an authoritative practice. The Dominicans therefore elected to follow Cistercian chant as codified in a master-exemplar of music and liturgy compiled around the mid thirteenth century. At about the same time, the Franciscans turned to a new source of authority in matters of chant and liturgy, namely the revised versions of the Breviary and Missal as drawn up at the papal court or curia in a series of revisions beginning with Pope Innocent III (1198–1216). 35 The return to music and liturgy as practised in Rome by the Franciscans was undoubtedly in part a matter of political expediency, but more fundamentally reflected a renewal of the political authority of the papacy.
The basis of the Roman rite adopted by the Franciscans was the so-called Romano-German Pontifical, a heterogeneous collection of liturgical rites which first appeared in Germany in the late tenth century. Following the coronation in Rome of Otto I as Holy Roman Emperor in 962, the liturgy associated with the Romano-German Pontifical became rapidly established at Rome: in a related process, the chant tradition as developed by the Franks also began to return to Rome, eventually replacing the local tradition now known as Old Roman chant. An attempt to revise and reclaim this received liturgy and music as distinctively Roman received significant impetus under Gregory VII (1073–85), who in attempting to reassert papal authority over territorial rulers claimed that the period of Germanic domination of the church had come to an end. While his desire to restore a pristine Roman liturgy proved idealistic, Gregory's efforts led to a curtailment of other rites, most notably the Visigothic. At the same time, Gregory's insistence on the authority of the priesthood above all laity (including rulers) served to further the developing divide between the clergy, who administered sacraments, and the laity, who were reduced to the role of observers.
The final stage in the spread of Roman liturgy occurred with the adoption of the practices of the curia by the Franciscans. Since the mendicant friars and the clerks of the papal court were both mainly concerned with affairs in the world, what was required was a liturgy that could be performed with the minimum of elaboration, and, crucially for the Franciscans, one that could be reduced into practical and portable collections. The eventual result was the codification of shortened forms of services in the form of a uniform Breviary (for the office) and Missal (for the mass) as regulated by the Ordinal of the Papal Court (ca1213–1216). 36 Again, however, it must be noted that at the same time as movement was taking place towards uniformity, there was an increase in popular devotions, especially those sponsored by the Franciscans, whether the Stations of the Cross in Holy Week, or the singing of simple religious songs in the vernacular (laude spirituali or carols).
Although no major liturgical reforms followed from the thirteenth century until the Council of Trent, two developments should not go unmentioned. First, the later Middle Ages saw a spread to the laity of forms of worship developed in monasteries; thus Books of Hours transferred both corporate hours and additional devotions, such as to the Virgin Mary or to the dead, into the sphere of private piety. Second, in the closing years of the thirteenth century a new pontifical was assembled in Avignon by perhaps the most famous liturgical commentator of the later Middle Ages, Bishop William Durandus. This compilation, which represented the clearest arrangements of materials to date, eventually superseded even the Pontifical of the Roman curia to become the model that was used until the Second Vatican Council. However, as Durandus noted in his commentary on the divine office, liturgical diversity persisted despite the spread of a uniform Breviary and full Missal of the Roman rite: 37
The reader should not be disturbed if he reads about things in this work which he has not found to be observed in his own particular church, or if he does not find something that is observed there. For we shall not proceed to discuss the peculiar observances of any particular place but the rites that are common and more usual, since we have labored to set forth a universal teaching and not one of particular bearing, nor would it be possible for us to examine thoroughly the peculiar observances of all places.
A striking example of the variety that could occur within the Roman rite is provided by Salisbury (Sarum) Use, not only because the various local uses that constitute Salisbury Use were in themselves highly varied, but also because the ceremonial elaboration within Salisbury Use stands in contrast to the pruning that took place in the Roman rite as observed in the papal chapel in the Lateran. 38 Sarum Use, namely the version of the Roman rite associated with Salisbury, the earliest surviving documents for which date from the thirteenth century, gradually spread through England, Wales and Ireland so that by the Reformation only the local uses of Bangor, Hereford, Lincoln and York remained. A comparison of the Roman rite of the papal court, as filtered through the order and ceremonial prepared for the Franciscan order in around 1250, with a summary of Salisbury Use highlights several points of difference ( Table 11.4 ). 39 Most notable is the greater use of processions in Salisbury Use, especially before High Mass on Sundays, although it should be recognized that processions were only infrequent in the papal court and places that adopted its use. The grander use of space in Salisbury Use is also apparent in the restriction of the celebration of mass in the papal court to the area around the altar, whereas Salisbury Use raises up certain activities to the height provided by the choir screen. A further difference is the amount of censing of the altar, which in Salisbury Use took place even during lessons at Matins. A final distinction which cannot be appreciated from the tabular comparison is the number of ministers involved: Salisbury Use could feature as many as seven deacons and subdeacons, two thurifers, and two to four priests in copes acting as cantors; the custom of the papal court as codified in Franciscan documents specifies for solemn mass on Sundays only a single priest, deacon, subdeacon and acolyte.
***
For things should not be loved for the sake of places, but places for the sake of good things. Therefore select from each of the Churches whatever things are devout, religious, and right; and when you have bound them, as it were, into a Sheaf, let the minds of the English grow accustomed to it. 40
These words, attributed to Pope Gregory the Great (590–604) in response to a question from Saint Augustine of Canterbury about the justification for variant local customs, may be taken as a spur for bringing together the strands addressed in this chapter. The importance of the first millennium for establishing the structures that lie at the basis of the medieval liturgy has been stressed, as well as the way in which diverse practices were gradually aligned. Similarly, the importance of Rome as a centre for disseminating a model of practice, yet nevertheless one that not only arose from varied practices, but also allowed different expressions of an established framework, has also been emphasized. The extent to which liturgy was shaped by the identity of individual places (Rome, Metz, Paris), as well as the way in which local places were articulated through a liturgical mapping out of space and formalization of communities, has also been touched upon.
Told in this way, the story is one of growth and circumscribed diversification, but a less positive evaluation of the spread of practices across the Latin West could also be told. A narrative based on the rise of a coherent historical, cultural and geographic Latin West overlooks not only the practices of different faiths within this domain, but the parallels and exchanges between Eastern and Western practices. While the comparative lack of sources makes it difficult to address Jewish and Islamic worship, 41 increasing awareness of the fundamental discontinuities between even the Latin liturgical sources that survive, as well as increasing suspicion of their witness to current practice, is prompting renewed interest in the stories that have been told about them. 42
A more general reflection on Gregory's words could begin from his use of the word ‘things’ to refer to the contents of liturgical exchange. On the one hand, the transfer of material goods is intrinsic to the history of music and liturgy, whether in the form of the paraphernalia needed for worship (fine vestments, vessels, candles, incense, books etc.), or the ongoing supply of wealth needed to maintain institutions (including, centrally, choirs). On the other hand, the terminology is indicative of the fact that, for all that has been said about the relation of music and liturgy up to this point, no single word for ‘liturgy’ and no clearly separable word for ‘music’ was in use in the Middle Ages. 43 The customary liturgical distinction between rite (texts or things uttered) and ceremonial (actions or things done) highlights instead the way in which an attempt has been made here to address music as an aspect of ceremonial. In considering music in the liturgy as ceremonial agenda or things done, the extraordinary power that it wielded over both the self and others is finally brought into consideration. It is precisely this power that caused the earlier Saint Augustine to hesitate before, in a move indicative of the widespread acceptance that music was later to find in Christian worship, sanctioning its use as divine enchantment: 44
I vacillate between the peril of pleasure and the value of the experience, and I am led more – while advocating no irrevocable position – to endorse the custom of singing in church so that by the pleasure of hearing the weaker soul might be elevated to an attitude of devotion.