ONE field in which the Carolingians made an advance had long since been taught theoretically as within the circle of the liberal arts, namely, music. But a theory of music, even a scientific one, is not the same thing as the practice of music, and it was in sung music the Carolingians made so notable a contribution. The ninth century saw a musical efflorescence due both to the fusion of an oriental tradition with Greco-Roman music, and to the existence of a great musical instrument, the monastic choirs. Benedictine monks were bound to a recitation of the divine office as the opus dei, and, since St Benedict’s own day, to a more frequent or even daily attendance at the solemnities of the mass. A body of men cannot recite the psalms or canticles intelligibly unless they make a ‘voiced’ sound, that is, unless they chant them (not necessarily with variation of musical notes): a multiple use of the ordinary speaking voice produces only a confused murmur. Even the single cantor or lector, reading a prayer, must make a voiced sound, i.e. chant, for his voice to carry across a quite moderate sized cella or chapel. Hence, from the days of the public celebration of the Christian mysteries at least, and presumably from the very earliest days when Jewish and Syrian and Greek and Persian Christians gathered themselves together, there had always been some chanting. Throughout the middle ages the parish priest and his little altar boy were always said to ‘chant’ the mass, i.e. sing it, though this might mean no more than that the priest said the audible portions of the rite on a note. But while, in the Carolingian period, the rural or town priest would be conservative about such chanting, the monastic and cathedral choirs, singing mass and office day in, day out, would tend towards musical enrichment of the office as naturally as towards the adornment of the chapel and altar. Not only the monks but the cathedral clergy now sang the mass and office: for Charlemagne had ordered that
the clergy shall have a good knowledge of the Roman chant, and use it in the night office and the gradual office [the singing at mass], even as our father Pepin of blessed memory decreed when he did away with the Gallican (office) for the sake of uniformity with the apostolic see.
Much light has been thrown in recent years by the Byzantinists on the origin of church music; like Byzantine music in general, it was entirely vocal and homophonic (without parts). While the Greek philosophers from the time of Pythagoras in the sixth century B.C. had discussed musical intervals and the octave and the whole scientific connexion of music with mathematics, and while by the time of Constantine the rhetors were teaching a classical arrangement of Greco-Roman music in fifteen modes or scales, cult music was, in fact, mainly local. The succession of intervals in the mode, and the melodies of the mode, that is, were local. The Christian musical phrases and the method of chanting antiphonally went back to eastern origins: to the antiphonal chanting in the synagogue, and possibly to the melodies and intervals used by peoples outside the eastern frontier of the Roman empire. This was not strange: since the Christians continued the liturgical reading of the Old Testament scriptures, and the traditional form of the worship of the synagogue for the first part of the Eucharist, it was natural that antiphonal chanting should form part of Christian worship: Eusebius in his Ecclesiastical History assumes that this was so.
It is also natural that in the west, where Christian worship was necessarily brief and private, information about early chanting is lacking. The only churches built for public Christian worship were outside the empire, in the east; here, the remains of such churches, built before Constantine, still survive. There was no persecution in the Persian kingdom, and a good deal is known about the church of Seleukia-Ctesiphon (see p. 170). It therefore fits into the historical picture that our first reference to antiphonal chanting should come from eastern sources. It appears that the congregation formed two choirs, the one of men and the other of women and children, and that when each choir had sung a verse, a short refrain was chanted by both choirs. When Ambrose died in 397, such antiphonal chanting at Christian worship had already passed from the Greek east to the Latin west, like the forms of the centrally-planned Christian churches: music in Milan was Greek or oriental.
Gregory the Great forbade the deacon to sing more than the gospel at mass (595), and gave the cantor the charge of the singing in churches; he cut down (see p. 518) the Ambrosian iubilus and other melodic formulae. In the singing of the creed the music became ‘syllabic’, one note to each syllable, which, as Iso at Saint-Gall knew, was the Roman musical rule. There was in the west a real reform, redirection and pruning of the church music then in general use in the Byzantine empire. But further musical advance waited upon the discovery of some method of writing music; in this respect too the west borrowed from the Byzantines.1
Up till the Carolingian period, music in the west had been unwritten, dependent on the memory of the cantor: there had been no notation above the words to be sung in books used for the mass or office. Nevertheless, the Byzantines by about 500 had developed a system of writing music from the old prosodic signs described by the sixth-century encyclopedists. These signs had denoted voice-inflexions, a rise or fall in the pitch of the notes, as well as time duration, a long or short pause. They were used for secular music, in the acclamations required by court ceremonial, and in ecclesiastical music, and by about 500 the Byzantines were using two sets of signs, the ecphonetic signs, used in reading the lessons, and the musical signs proper. While the ecphonetic signs were set at the beginning or end of a group of words, the musical signs were set over the syllables of the text. The ecphonetic signs included the high tone or acute (rising) accent, the low-pitched tone or grave (falling) accent, and the ligature or circumflex, consisting of a rising and falling accent; they included also marks of quantity, for long and short syllables, rough and smooth breathings, and certain declamation marks. The musical signs proper at this early stage had no distinct interval value: they were called neums (pneumata). In form, both sets of signs consisted of lines and dots, single or grouped, the lines set at an angle and sometimes hooked. Their use appears to have spread throughout the Mediterranean countries. Isidore of Seville, at the far end of the Mediterranean, denied that music could be written: music, he said, ‘is a thing of sense: it passes along into past time and is impressed on the memory. For unless sounds are held in the memory of man they perish, because they cannot be written’. Yet elsewhere he implied that music could be written, when he wrote of one or two churchmen that they made and published musical compositions, and of his brother Leander that ‘he made many compositions of sweet melody, for the offertory of the mass, and the sacrifice and the alleluia’s and the psalms’. It would seem that Leander used the Byzantine notation; but no western liturgical manuscripts using such notation have survived.
When pope Hadrian (772–795) sent to Charlemagne at his request musicians to teach his prelates the Roman use and liturgical practice for the course of the year, they brought with them books which the Franks said ‘signified the notes by letters of the alphabet, either upright or inverted or forward or backward’, i.e. had the Byzantine ecphonetic and musical signs. Though for long after the eighth century this system was too difficult for most western singers, and antiphoners and tropers continued to be written with the words only, leaving the melody to be passed on by the memory of the singers, yet there were some cantors in the song schools who understood the notation and began to write it. The prosodic signs had originally been written in lines of equal thickness on papyrus; but when the Caroline scribes wrote with their quill pens on vellum, the up strokes became long and thin and the down strokes short and fat: the virga, punctum and circumflex became musical notes, and, in combination, groups of notes. From this system of ‘neums’ our own musical notation has developed. There is no manuscript evidence of its use among the Franks in Charlemagne’s own day, though its introduction must date from the copying of liturgical books with the new Roman rite; but it was certainly in use in the ninth century in the song schools of Jumièges and Limoges, Corbie and Saint-Gall and the rest. Its use is especially associated with the ninth-century singers of Saint-Gall.
Neither in east or west was there as yet, apparently, an exact interval notation. The old Greek music discussed by the rhetors had depended upon an arrangement of intervals, tones and hemitones, in the fifteen Greek modes or scales. Smaller intervals and different arrangements would seem to have been used in local music within the empire and oriental music beyond the eastern frontiers; but by the sixth century the smaller intervals were not used in Greek music, nor were they used in the later Byzantine music or transmitted to the west. By the eighth century the old fifteen Greek modes had been reduced to the eight modes of Byzantine church music, and were so transmitted (with the tonus peregrinus) to the west: they were known as the oktoékos and were set in a compass of two full octaves known as the ‘greater complete system’. The Arabs learned the eight modes from Christian churches and monasteries in the east and gave the system an Arabic name. In the early tenth century the monk Hucbald of Saint-Amand wrote or revised a notable hand-book explaining the system in a work entitled the Musica Enchiriadis.
The interval system in western chant was now firmly established: but it was not yet exactly indicated in the notation. The western system developed independently of Byzantine music: the use of a staff line above the written words, the position of the note groups on, above or below the staff, and, finally, the addition of three other lines to the staff (now distinguished by a clef) made possible the exact recording of the intervals used in the melody. This method was in use in the west in the early tenth century, for it is described in the Musica Enchiriadis. The golden age of the liturgical chanting of mass and office, with the year-long communal meditation on the mysteries of the life of Christ, had come.
In the sphere of art, the Carolingian renaissance produced few large monuments and almost no sculpture in the round. Artistic skill found outlet rather in objects of comparatively small size, particularly those used in the service of the church. As in the case of learning, there was an effort to recover and imitate the Greco-Roman past, or rather, since architectural and art-forms could not be completely recovered from old manuscripts like classical texts, to imitate such Byzantine buildings, ivories and reliquaries as could be seen in Italy or brought from the east Mediterranean. Byzantine and eastern influence tended to oust the abstract art of the Celt and the old remnants of Germanic animal ornament. Of the two Hellenistic styles of east Mediterranean art, the Celts and Northumbrians had adopted the neo-Attic in the portraits of the evangelists in their gospel books (see p. 237): from the time when the Alexandrian style, with its impressionistic figures and landscape, had become domesticated in Italy, it had grown more popular than the space-circumscribed, timeless, passionless neo-Attic. ‘Hellenistic’, when used of ninth-century east Mediterranean art, usually implies a lively, naturalistic use of the human figure and an impressionist rendering of incident and of the background. It was this form of east Mediterranean art that the Carolingians finally chose, in place of the stylized, abstract art of the Celt.
Though the Franks thus revered the Greco-Roman past and the Byzantine present, they had neither the slave labour nor the numbers of skilled craftsmen to build many great stone buildings, with carved capitals and colonnaded halls, as the Greeks and Romans had done, nor did their public life demand municipal basilicas, libraries and temples. It did however demand a few palatia for their rulers, some stone churches, many fine ivory chests and reliquaries, and fine church books: which the Greco-Roman pagan temples had not needed at all.
With regard to architecture: it was natural that Charlemagne and his court, with their eyes on the re-establishment of a Christian empire, should have planned the imperial chapel at Aix in imitation of a fine Greco-Roman church, St Vitale at Ravenna: of imperial buildings in Constantinople they had no first-hand knowledge. St Vitale had been built for Justinian and completed in 547: it was an octagon with a two-storeyed octagonal ambulatory and a chancel and apse at the east end. The niches enclosing the central space beneath the dome were not enclosed with walls but opened into arcades on both ground and first floor, and thence into ambulatories: ‘the central space flows into the ambulatory’, creating ‘a sensation of uncertainty, a timeless floating’ (N. Pevsner). On the walls, the glowing surface of the mosaics with austere figures in sombre tints seemed just as immaterial. These spatial subleties and complexions were however beyond the understanding of Charlemagne’s architects. The Frankish builders copied the octagonal plan of St Vitale for their chapel: but they flattened the curved-out niches, eliminated the columns on the ground floor, and built simple openings with short, sturdy piers. Yet for the upper floor they provided polished antique columns superimposed in two orders, which ‘re-echo something of the transparency, and the floating of space from one unit into another, which make the beauty of Justinian’s churches’. They placed an equestrian statue of Theoderic, looted from Ravenna, in the colonnaded forecourt of Aix: they believed it to be a figure of Constantine, the Christian emperor.
Italy provided models, not only for royal villas in the classical, inward-looking Roman house, but for important early Christian churches and basilicas, particularly those of Rome. The churches of Saint-Denis and Fulda, begun in 760 and 802 respectively, were planned on the model of St Peter’s and the other Roman basilicas. The unusual plan of the church at Saint-Riquier surviving in a vellum copy made c. 835, shows it as having two transepts, one at the east end and one at the west, the crossings with the nave crowned by towers in both cases. The opening of the ninth century saw the building or enlargement of many abbey churches in stone; for the country churches and the rural villas or halls of the nobles and officials wood must have been largely used.
With regard to sculpture in the round: the only surviving specimen is the equestrian statuette, in bronze, now in the Louvre. Critics accept this as ninth-century work, an effigy either of Charlemagne himself or of one of the later Carolingian kings. It is surprising that when Byzantine illuminated gospel books reached Northumbria and inspired a sudden efflorescence of figure sculpture on the Ruthwell and Bewcastle crosses (though, as far as we know, no figure in the round), no such efflorescence of figure work in stone was inspired among the ninth-century Franks. Possibly the Northumbrians, working with a north Germanic tradition of timber architecture, had wood carvers who could turn their skill to sculpting the stone crosses; while the Franks, who had long looked to the Mediterranean for inspiration, had no such artistically skilled wood carvers to turn into stone masons. The Franks had armourers, metal workers, and their single surviving figure in the round is this bronze. It looks as if the timber halls and boats of Northumbria may have had much more decorative carving than the halls of the Franks.
For illuminated gospel books and psalters, used in every abbey and great church in the solemn celebration of the liturgy, there was a great demand in the Frankish empire, and the Caroline illuminators evolved a style of their own. They used mainly Byzantine models, which were at hand in Italy, adding something of their own less civilized vigour and directness. The Irish monasteries on the continent still provided decorated manuscripts with pages of richly coloured ‘abstract’ design, and their stiff, formalized figures of evangelists. The gospel books of Saint-Gall and Echternach, and the Book of Bobbio, were all the work of Irish scribes, and as Celtic in character as the Book of Kells. The influence of the Celtic style survived, however, in the great majority of Caroline manuscripts only in detail: in borderings, interlaces and solemn initials. Caroline illumination as a whole was modelled on Byzantine and east Mediterranean forms (see p. 423).
The work of the Caroline illuminators is usually considered in schools, though these have no sharply dividing lines. The earliest, the so-called Palatine or Aix group, of which the Vienna, Brussels and Aix gospel books are examples, all belong to the reign of Charlemagne himself; they represent the evangelists as clothed in white togas carefully modelled by shading, and usually without their symbols, as was customary in Byzantine painting. The Vienna gospels used a border of the classical acanthus, and is said to have been the book found on the knees of the dead Charlemagne by Otto III, when he had the tomb opened in 1001; it is the book on which the German kings took their coronation oath. The Berlin and Aix gospel books have conformed to western usage in providing symbols for their evangelists: but there is still an impressionist sense of sky and landscape, with which the evangelists still have some relation.
The manuscripts of the so-called Ada group (from a dedicatory inscription in one of them to a supposed sister of Charlemagne) show more Celtic influence in borders and decorations; the group seems to have been focussed at Trier. An example of the work of this school, made at the end of the eighth century, was the gospels of Godescalc, who had (apparently) been to Ravenna and copied the representation of the Saviour from the walls of St Vitale: the classical shoulder-curls of the figure became however in Godescalc’s painting the stringy locks of the northern warrior (R. Morey). The text was written (781–783) in gold on purple vellum, as in models from Constantinople: all the Ada books made apparently for Charlemagne or royal patrons, used rich and luxurious material, like the Harley Gospels in the British Museum, and the early ninth-century Codex Aureus. The evangelists of this school sit within a circumscribed architectural frame, often in an apsidal niche between two columns, with their symbols above their heads; they are, as it were, effigies, like the Celtic evangelists, but the treatment is more humanistic, and the colour very rich. The later Ada manuscripts became increasingly Byzantine.
The work of the school of Tours flourished under the abbacy of Alcuin in the two monasteries of St Martin: that within the city of Tours, and Marmoutier (see p. 71). Alcuin’s chief preoccupation was the revision of the text of the Vulgate, not its decoration, but in his time the beautiful Caroline minuscule came into use at Tours and was associated with an illumination of mixed insular and Morovingian type. Under his successors it reached its peak of beauty, as in the Bible presented by the lay abbot of St Martin’s, count Vivian, to the emperor Charles the Bald (875–877): and the Bamberg Bible, which reverts to the method of continuous narrative in narrow strips. The interest of these and other Caroline manuscripts lies partly in the evidence that they were inspired by some antique model. The illuminated scenes in Genesis and Exodus in count Vivian’s Bible are reminiscent of some late fifth-century model like the Vatican Vergil, for instance in the clumsy rendering of an antique, graduated sky and, in the background of the group representing the delivery of the Law by Moses, the placing of an antique temple (cf. p. 423).
The school of Tours had been eclectic in its use of material: that of Reims, perhaps the largest and most important of the Caroline groups, produced purer examples of the antique style, more nervous and linear than the illumination of Tours. It is not certain that all the books of this school were actually produced in Reims itself: the style is named after one of its examples, the gospel book of Ebbo, archbishop of Reims (816–835). He had been librarian at Aix earlier, and seems to have brought skilled illuminators or fine books from the palace school to Reims. The Utrecht psalter is a fine example of the work of this school: it has the words of the psalm illustrated by bands of small scenes, interspersed with lines of the text (in rustic capitals). Compared with the illumination of the palace school, the figures have exchanged a meditative calm for a lively, almost agitated rendering: the small pictures are literal, lively and sometimes poignant. The artist has made a complete break-through from the stiff, stylized treatment to an Alexandrian naturalism. The Drogo sacramentary from Metz has beautiful examples of storiated initials.
Round the court of Charles the Bald a new school of illumination developed, perhaps focussed in the scriptorium of Saint-Denis. Most of the manuscripts were written and decorated for Charles, who personally assumed the function of lay abbot there in 867. The library or treasury of the abbey now included count Vivian’s Bible and some of the Ada manuscripts, and doubtless specimens from other schools, all of which lent models to the Saint-Denis illuminators. They used both Franco-Saxon initials and the fine, humanistic drawing of the Reims school and its acanthus borders: they borrowed and adapted from all Caroline styles: they completely filled whole vellum pages with rich, illuminated ornament. Fulda was another eclectic school which used line drawing on a larger scale than the Utrecht psalter, and also the massively elaborate compositions of the Ada school.
Caroline decoration was not confined to vellum books, but used also in the minor arts, goldsmith’s work and ivory carving particularly. Surviving examples include the gold covers of gospel books, e.g. the Ashburnham gospels, in the Reims style as handled at Saint-Denis, in Charles the Bald’s golden gospels: and the portable altar presented by Odo, count of Paris to Arnulf of Carinthia, a precious object he had taken from the treasury of Saint-Denis. The late ninth-century golden altars show the Caroline love of linear effect and ‘instinct for material splendour’ (R. Hinks). They can be compared only with the Byzantine diptychs in their use on the same surface of gold, silver, coloured enamels and figure work in relief. The golden altar of Sant’ Ambrogio in Milan, given by archbishop Angilbert about 835, and the cover of the Munich Codex Argenteus, produced at Saint-Denis about 870, both show a mixture of Byzantine and Frankish art. The figures are distinctively Frankish, the jewelled footstools and architectural detail, among which they move, distinctively Byzantine; in all this modelled figure work the fluttering drapery, the craning necks, the expressive gesture all suggest movement, and even violent disturbance. Similarly, in the engraved crystal disc eight inches across, in the British Museum, which bears the name of the emperor Lothar, the story of Susanna and the Elders is engraved in a series of lively episodes; they are like the sketches in Hellenistic illumination, not connected in any coherent design.
The earliest Caroline ivories come from c. 800, and are parallel to the early Ada manuscripts in design: they merely imitate old Italo-Gallic models, though with more emphasis on line than plastic form. A book cover in the Bodleian Library has Christ treading upon the lion and the adder; the beautiful pierced ivory cover of a gospel book of the Metz school, now in the Bibliothèque Nationale, has on one side three gospel scenes, the figures beautifully modelled, with fluttering drapery and limbs in motion; even the shepherds here presenting their gifts to the infant Christ are represented half-kneeling, as if in a movement caught by a camera; the whole cover is bordered by a rich, undercut and pierced design of growing grape-clustered vines.
Of Carolingian wall-painting north of the Alps nothing has survived, though there is plenty of literary evidence that its subjects were both scriptural and drawn from classical antiquity. The murals of some ninth-century Italian churches have survived, and can perhaps be taken as evidence of the general character of Frankish wall-painting, if only because their landscape and figure work share the general character of scenes in contemporary illuminations and ivories, which would have supplied models for some of the Frankish palace wall-paintings; contemporary writers speak also of the painted palatia of bishops and great abbots. In the lower church of San Clemente at Rome a fresco of the Ascension includes a group of apostles obviously amazed, and showing their wonder by hiding their eyes or waving their arms. It is all very fluttery: and it is possibly the work of Benedictines dispersed from Monte Cassino; the refugees from iconoclasm had spread a Byzantine style in Italy rather earlier. In any case, the great Benedictine abbeys north of the Alps, possessors of ivories and miniatures, would seem to have been the chief agents in the development of late Carolingian wall-painting and figure work. The Carolingian rulers were their patrons: but the artists in many cases were monks.
Written references to Frankish murals show that all doubtfulness at mentioning or depicting the old pagan gods had completely disappeared: gone were the days when Augustine had apologized for referring to a pagan goddess because it might have implied that he believed in her existence. The paganism which Frankish capitularies and canons condemned was the Germanic or Scandinavian paganism of a people recently converted, not the old gods of Greece and Rome. The very constellations in the night sky had been named after the old gods and heroes who were now painted on the walls even of episcopal palaces: Mercury and Jupiter, Perseus and Andromeda and Hercules. It was fashionable to paint on palace walls the Biblical personages with their pagan antetypes: but sometimes the pagan personages appeared without any antetypal excuse.
The episcopal palace of Theodulf at Germigny-des-Prés had a lively painting of Earth on the refectory walls: she was represented as a strong, vigorous woman among a profusion of chariots, animals and serpents. The palace of Saint-Denis had paintings of the seasons as women, and Saint-Riquier had figures representing the different parts of the world (R. Crozet). At Saint-Gall the sages of Greece faced the Christian saints. The palace at Ingelheim had paintings in a great historical cycle: the pagan conquerors, Cyprus, Ninus, Romulus, Alexander, Hannibal, etc. were set over against Constantine, Theodosius, Charles Martel, Pepin and Charlemagne. Some monastic walls had paintings of the occupations of the months (those from Cosimo de’ Medici’s study at the Victoria and Albert Museum had a very long ancestry), Æsop’s fables and hunting scenes with deer and dogs, all in the Byzantine manner. There must have been many more than those whose memory is preserved in literary reference, and the choice of pagan subjects was natural when so many Carolingian scholars were studying Vergil and Ovid, and when manuscript illuminators of liturgical books even slipped into the detail of decorative borders Bellerophon mounted on Pegasus or the chimaera, Apollo the healer, or a laurel-wreathed bacchante. The sacramentary of Drogo had Earth as a woman suckling two children: and Ocean, riding a dolphin. Even in subjects not directly pagan, the traditions of antiquity intruded: some manuscripts from Tours have the apostle John as an old man with a wing-filled banner above his head, in a manner in which pagan deities used to be portrayed. As to detail: towns were represented as women crowned with a walled and towered headdress, and antique weapons, costume, standards, temples, columns, the masked actor and the paraphernalia of the theatre all appeared in pictured landscape and scenery. Together with the allegorical draped women, feeling for plastic form and fluttering drapery, there was transmitted from pagan antiquity ‘the sense of grandeur and the sense of humanity’ (R. Crozet).
BIBLIOGRAPHICAL NOTE