As with so many other aspects of Hispanic culture in the early Middle Ages, the musical life of the Iberian peninsula was distinct from that of other parts of Europe for many centuries. Though Christianity reached the peninsula as early as the third century, most of the country came under Muslim rule in 711, to be reconquered by Christians in a series of campaigns through the Middle Ages. Not only did this bring Arabic music into Europe, but the Muslims were also uniquely tolerant of the Jews, allowing a Judaeo-Spanish musical tradition to flourish, which has had wide influence elsewhere in later centuries. Until the eleventh century, the church in the Spanish kingdoms remained independent from the Roman rite and employed a separate liturgical structure, with its own musical tradition. The later Middle Ages saw much greater assimilation of musical traditions from the rest of Europe, but the peninsula also developed its own discrete musical genres, particularly in secular music making.
Though some evidence survives for music and dance from the Stone Age onwards, 1 it will be as well to begin this survey after the Visigoths had established a kingdom centred on Toledo and extending across almost the whole of modern-day Spain, Portugal and the south of France, by the end of the sixth century. Isidore, Archbishop of Seville (ca559–636), is a figure of major importance in all branches of learning. He compiled an encyclopaedia, the Etymologiarum sive originum libri xx , which enjoyed an extraordinary Europe-wide influence throughout the Middle Ages, surviving today in more than a thousand manuscripts. 2 The Etymologiae are largely a synthesis of the writings of various classical authors, and Isidore's definition of music as a discipline concerned with numbers as sounds directly follows Cassiodorus. As well as defining music as a subject of study within the Quadrivium, Isidore provides much information here and in his practical text on the liturgy, De ecclesiasticis officiis , on the role of music in the divine office. 3
Isidore played an important part in the fourth council of Toledo in 633, one of the canons arising from which prescribed a single order of prayer and singing (‘orandi atque psallendi’) to be kept through all of Spain and Gaul (Gaul here meaning the Roman province centred on Narbonne). Neither his own writings nor the decrees of the fourth council suggest that Isidore himself was directly involved in the composition of the chants for the newly reformed liturgy, but his writings enhance our understanding of the surviving liturgical books, all of which date from later centuries. Isidore and others do, though, mention various composers of chant, including his brother Leander, whom he succeeded as Bishop of Seville and who ‘composed many works of sweet sound’.
The liturgy of the Spanish church is often misleadingly referred to as the ‘Mozarabic rite’, since it is known today from books used by Christians living under Moorish rule, the Mozarabs (from the Arabic ‘making oneself an Arab’). The more general term ‘Old Spanish’ or ‘Hispanic’ is nowadays preferred, since it is clear that the liturgy was well established before the Muslim invasion in 711, at least from the fourth council of Toledo and probably in some form long before this. It was also used for some time in parts of the Christian north of Spain that never came under Muslim rule, and other parts that returned to Christian rule in intervening years. The rite did not go unchallenged within Christian territories, however: in parts of Catalonia the Roman rite was preferred from a very early date, and a growing desire for liturgical uniformity (and imposition of papal authority) through the tenth and eleventh centuries led eventually to the suppression of the Old Spanish rite in favour of the Roman throughout reconquered Spain in the treaty of Burgos in 1080. 4 Meanwhile the Portuguese territories had used the Roman rite from as early as the sixth century, and the Old Spanish rite was employed for only a somewhat briefer period than in the northern Spanish kingdoms. 5
Only one liturgical book, an orationale written in Tarragona, survives from as early as the time of the Muslim occupation. 6 This does not include any musical notation, but the chant texts it records concur with later sources, confirming that the liturgical prescriptions were broadly fixed by this time. Notation is preserved in more than twenty complete manuscripts of the Old Spanish rite, mainly from the tenth and eleventh centuries, and a further twenty or so fragments. 7 This is a large number when compared with other chant traditions from this time, but because of their early date, almost all of the manuscripts are written in a notation that is not decipherable. The suppression of the rite unfortunately occurred just before pitch-specific notation was becoming common elsewhere. The notation bears some relation to other neumatic systems, but is in many respects quite different, the extravagantly florid style of some complex neumes being unlike any other traditions in western Europe. It may be divided into two broad categories: about one third of the surviving manuscripts employ an oblique, almost horizontal ductus, associated with manuscripts from the school of Toledo, and the remainder use the ‘northern’ notation especially associated with León and the Rioja, which has a more vertical aspect. 8
It is a particular shame that no manuscripts survive from the southern half of the country, since those parts which remained under Muslim rule into the fifteenth century might well have continued to use the Old Spanish chant late enough to have notated it on a staff. As it is, we are fortunate that in a single tenth-century manuscript from the monastery of San Millán de la Cogolla (Rioja), the Old Spanish notation of eighteen chants for the Office of the Dead was erased in the eleventh century, to be replaced by heightened – and therefore decipherable – Aquitanian neumes. 9 Similarly, three antiphons for the washing of feet on Maundy Thursday were replaced with Aquitanian notation in MS 4 of San Domingo at Silos, a monastery from where many other important liturgical manuscripts also survive. This may leave us with only twenty-one decipherable melodies out of a complete repertory of many thousands, but scholars have extrapolated from this slender evidence to reveal many general similarities with other Western chant traditions. 10
The manuscripts preserve a large repertory of antiphons for the mass and office, which Isidore tells us were sung by two choirs alternating, as in the Gregorian tradition. There is no evidence of a structured system of eight modes, but certain formulas are found regularly in particular contexts, which again shows a certain degree of resemblance to other Western chant traditions. The responsory chants are closer in form and structure to their Frankish equivalents, but show a considerable degree of variation in their psalm-tones between sources from the different regions. The various chants for the mass use a separate terminology from their counterparts in other rites, but have broadly similar functions in the service.
The gradual suppression of the Spanish rite in favour of the Roman was generally brought about through new liturgical books using Aquitanian notation, though a few examples survive from the time of transition of manuscripts of the new rite using the old notation. 11 The extent to which Spanish plainchant demonstrated uniformity across the country and remained true to its Roman sources in succeeding centuries is not a subject that has yet received much attention. There were certainly national and local characteristics in particular liturgical forms, some of which are likely to be a legacy from the old rite. For example, though the majority of hymn texts found in the Spanish kingdoms were widely disseminated in other countries, almost two thirds of the melodies are not generally known outside the peninsula. Many of these are likely to be native compositions, and some are probably remnants of the Old Spanish rite. 12
It seems improbable that tropes or sequences were ever employed in the Old Spanish liturgy. 13 There are references to prosaria , a term often used to describe books containing both tropes and proses, in various Catalan documents listing the books required for the Roman liturgy from as early as the mid tenth century, but the earliest trope manuscript to survive from the Iberian peninsula dates from no earlier than the late eleventh century. MS 105 in the Cathedral Archive of Vic (north of Barcelona) contains tropes for the ordinary of the mass, as well as several introit tropes and a large number of sequences. 14 Only around a dozen such manuscripts survive in all from the Iberian peninsula through to the fifteenth century, but it seems likely that the practice of troping was far more widespread than the surviving sources might lead us to believe. 15 In later sources, a far greater number of ordinary tropes are found than of proper, and while the repertory used in Vic appears to have been compiled locally, sources from elsewhere are often derived directly from Aquitanian practice.
A similar pattern of dissemination may be assumed in the case of liturgical drama. A later portion of the same manuscript in Vic, dating from the early twelfth century, includes a play of the three Marys, and similar plays are known from elsewhere in Catalonia from a relatively early date. 16 It is clear that while Catalonia was one of the great centres of creativity in the practice of liturgical drama, these forms of liturgical expression were much less widespread elsewhere in the peninsula. Brief versions of the Visitatio sepulchri are found in two eleventh-century manuscripts from the abbey of Silos written in Mozarabic script and notation yet forming part of the Roman rite, but other examples from outside Catalonia are few and far between. It seems that their place was taken by a complementary tradition of vernacular drama on ecclesiastical subjects, beginning in the late twelfth century with the Auto de los Reyes Magos , which cannot properly be considered as liturgical drama and for which music does not generally survive.
In Catalonia in the later Middle Ages several dramatic innovations took place in the context of the liturgy. A customary or consueta from Girona, dated 1360, describes eight different representational ceremonies, including a particularly sophisticated Visitatio sepulchri , the first of three plays to take place at Easter. Though the texts used at Girona do not for the most part survive, this book provides much useful information as to how the plays were enacted. Among the Christmastide representations was one on the martyrdom of Saint Stephen, a theme which appears to have been a Catalan speciality and which extended in Barcelona to a realistic stoning of the protomartyr with imitation stones. 17 In the fourteenth and fifteenth centuries the feast of the Assumption of the Virgin came to be celebrated in ever more elaborate processions, often including dramatic elements sometimes in the vernacular. The most famous of these, the Mystery of Elx (or Elche), is an extended drama in Catalan which is still performed each year. Though the present format and music largely derive from the seventeenth century, it represents a continuous tradition from the end of the Middle Ages.
Another ceremony that sometimes took on a dramatic form in the Spanish church is associated with matins on Christmas morning. The song of the Sibyl, prophesying the Second Coming at the Day of Judgement, has its origins in the first years of Christianity and began to be used in a liturgical context in the twelfth century, as the sixth or ninth lesson of matins for the Nativity, in certain monasteries as well as a few cathedrals. This practice was especially common in the Spanish territories, though examples are also found in France and Italy. The foreboding refrain, ‘Iudicii signum: tellus sudore madescet’ (‘Sign of Judgement: the earth will grow wet with sweat’), is sung in alternation with a series of verses. In some churches the prophecy would be sung by a boy dressed up as the Erythraean pythoness. Some of the Spanish sources from the fifteenth and sixteenth centuries give the words in Catalan or Spanish instead of Latin, thereby making this remarkable chant even more direct and immediate to its audience. 18
Just as tropes and sequences were unknown in the Old Spanish rite, let alone such dramatic ceremonies as the Sibylline prophecy, so it seems that the practice of polyphony was unknown to the Visigoths and Mozarabs. At the very least we can say that no notated examples survive, nor is there any testimony to the practice in documentary sources. The sparse material that has survived postdates the imposition of the Roman rite by some considerable time, and much of it shows the strong influence of French polyphonic styles.
The earliest significant manuscript in this context was probably made in France, possibly around Vézelay, in the mid twelfth century. It is a compilation of historical, hagiographical and liturgical material connected with Saint James, and was used at his shrine at Santiago de Compostela. 19 At the end of the book are found 21 polyphonic pieces, a mixture of mass chants, responsories, Benedicamus settings and conductus, all in two parts except for one in three parts. Very unusually for this period, the pieces are fictitiously ascribed to various named individuals, all of them French and including several bishops. Even if the manuscript was not created in Spain, its presence at Santiago from such an early date provides important evidence of the practice of polyphony in the Iberian peninsula in the twelfth century.
Our knowledge of polyphony in the Iberian peninsula in the following century derives largely from a diverse range of fragmentary sources from various monastic and secular institutions spread throughout Catalonia and Castile, 20 but in addition to these we are fortunate that two of the principal sources of polyphony prior to the Ars Nova are of Iberian provenance. The origins of the first, the Madrid Codex, are unclear. It was kept in Toledo Cathedral until being transferred in 1869 to the Biblioteca Nacional in Madrid, where it now has the number 20486, but there is no evidence to connect it directly with Toledo before the seventeenth century. It was probably written in northern Spain around 1260, and comprises some 61 conductus and 35 motets, as well as a few other polyphonic compositions, notably including texted and untexted versions of the famous quadrupla attributed to Perotinus, Viderunt omnes and Sederunt principes (some parts of which are now missing). 21 Many of the compositions in the Madrid Codex are also known from Parisian sources, but there are a good number of pieces unique to this manuscript, several of which may reasonably be assumed to be native Iberian compositions.
The second major source of polyphony is the Las Huelgas codex, which is still housed in the Cistercian convent for which it was written towards the end of the first quarter of the fourteenth century, the Monasterio de Santa María la Real de Las Huelgas, outside Burgos. 22 The manuscript is chiefly the work of a single scribe, and presents a substantial repertory of 48 organa (including several ordinary tropes in two or three parts), 31 sequences (eleven of them polyphonic), 58 motets and 29 conductus, to which another nineteen pieces in various genres have been added by eleven later hands.
The singing of polyphony, let alone tropes and sequences, was strictly forbidden by the Cistercian order, but Las Huelgas was in many respects an exceptional convent. It was founded in 1187 by Alfonso VIII and his queen, Eleanor of England, as both a monastic house for noblewomen and a royal mausoleum. It is very likely that this book, which brings together music written throughout the thirteenth and early fourteenth centuries, from France and elsewhere in Europe as well as a large number of Spanish compositions, was intended as a complete repertory of the music required at the convent outside that of the regular liturgical books. It is not certain that the nuns themselves sang the music: the foundation had a large retinue of chaplains who may well have formed a choir.
Unlike several of the major sources of organa and motets from other countries, the Las Huelgas codex is not a beautiful manuscript; it is written on parchment of poor quality and the pages are often heavily and inelegantly corrected. This makes it all the more interesting, as it was clearly intended as a practical manuscript, representing the music exactly as it was performed. The notation appears clumsy and unsystematic when compared with other manuscripts, but in fact shows considerable subtlety and nuance in matters of performance, particularly of rhythm. The repertory of pieces unique to this codex, at least some of which we may assume were specially composed at or for the convent, includes some attractive sequences in honour of the Blessed Virgin, and four planctus or songs of mourning for deceased kings and an abbess which show particular melodic extravagance.
As in the thirteenth century, so in the fourteenth various fragments give testimony to the widespread practice of Ars Nova polyphony in the Iberian peninsula, albeit on a less sophisticated scale than is found in other countries. The most famous source from this period is the so-called Llibre Vermell (‘red book’), which was compiled in the last years of the fourteenth century as a collection of materials relating to the pilgrimage site of the shrine of the Black Madonna at the abbey of Montserrat in Catalonia, where the manuscript remains to this day. As well as accounts of miracles, homilies and many other miscellaneous texts, seven folios are devoted to ten pieces of music. 23 An annotation makes it clear that these were written not for use in the liturgy but for pilgrims to sing for their own amusement and edification during the night vigil and in the courtyard by day. Four of the pieces are monophonic, two in two parts and four in three parts; eight are in Latin and two in Catalan. All but the last are songs in honour of the Virgin; three are simple canons, and another three are in virelai form.
The Llibre Vermell is an exceptional survival of a fairly simple type of songwriting from the end of the fourteenth century. Evidence for the creative development of liturgical polyphony in the Iberian peninsula in the following century is relatively limited. Simple polyphony is found on occasion in books of predominantly monophonic chant, very probably reflecting a more widespread tradition of improvised polyphony, but surviving polyphonic choirbooks do not demonstrate the degree of contrapuntal sophistication found in the Low Countries and elsewhere at the time. Studies of the rich archives of the Catalano-Aragonese court have shown that music had an important place in the life of the court, both inside and outside the royal chapel, and other archives are now being mined for similar information. 24 However, it was not until after the marriage of Isabella of Castile to Ferdinand of Aragón in 1469, and the subsequent union of the two kingdoms, that we may observe a conscious attempt to bring the latest polyphonic styles from central Europe into the peninsula. At first this was effected more by importing music than by commissioning new works from native composers. An unnumbered manuscript in Segovia Cathedral written in 1495–7 combines many works of such northern composers as Obrecht and Josquin with much simpler native Spanish compositions by Juan de Anchieta and others. 25 But in the early years of the sixteenth century fully fledged northern-style counterpoint came to be done with confidence in the peninsula, and such composers as Peñalosa and Morales may be ranked with the most sophisticated and talented of their contemporaries elsewhere. 26
As with liturgical polyphony, our knowledge of secular music in the Iberian peninsula in the Middle Ages depends on a very small number of manuscripts all of which are of immense importance in an international context. A tenth-century manuscript known as the Azagra Codex contains versus some of which are supplied with simple neumes. 27 In subsequent centuries we know that there was a tradition of professional musicians in courtly and perhaps also more lowly circles, but only very few musical fragments survive as testament to this tradition: it was not until the fifteenth century that romances (or ballads) and the increasingly popular genre of the villancico came to be written down. 28 A small bifolio now in New York is the sole musical source for six love songs, or cantigas de amigo , by a minstrel named Martin Codax, dating from the second half of the thirteenth century. 29 The stories told in the songs are set in Vigo, on the border of Spain and Portugal, and the texts are written in Galician Portuguese. A comparable manuscript fragment was recently discovered with the music of seven Galician love songs by the Portuguese king Dom Dinis (1261–1325). 30
Among these piecemeal testaments to a widespread tradition of vernacular monophonic song, pride of place goes to the most famous medieval secular song repertory to survive from any country, the Cantigas de Santa María of Alfonso X of Castile and León (1221–84), known as Alfonso el Sabio, Alfonso the Wise. Some 420 cantigas survive in four luxurious manuscripts, principal among them codex b.I.2 in the library of El Escorial. 31 The language is again Galician Portuguese, and most of the cantigas tell stories of miracles effected by the Blessed Virgin; every tenth song is a cantiga de loor in praise of the Virgin. Whether the king actually composed any of the cantigas himself is open to question: illuminations in the Escorial manuscript show him overseeing its production, and in some of the cantigas the king speaks in the first person, but it is more likely that his responsibility took the form of closely involved patronage.
The context in which this large repertory came into being has been hotly contested. No immediate precursors survive in written form, and it has been suggested that the melodies, which are notated rhythmically, may derive from Spanish folksongs, from the music of French or Spanish troubadours, or even from an Arabic song tradition. 32 Most of the cantigas are in the form of the virelai (ABCCABAB). Only a single melodic line is written, but the manuscripts are richly illuminated with pictures of musicians playing very accurate depictions of various instruments, most probably implying that they should be sung to the accompaniment of one or more instrument, and maybe also with dancers. Many modern reconstructions of medieval instruments have been made on the basis of the illustrations in codex b.I.2. The repertory as a whole is a remarkable survival and has spawned a considerable diversity of modern realizations.
The Muslim invasion of 711 was driven by a fundamental difference in religion, and yet it seems that Christians living in the Moorish territories (known as al-Andalus) were not severely persecuted. The church naturally lost its administrative authority, many people were converted to Islam and the majority assumed the language and dress of Muslims by becoming Mozarabs, but Christian worship was not proscribed; even some monasteries were allowed to continue to operate. Likewise, if to a lesser extent, the Christian reconquest in subsequent centuries left room for a considerable degree of acculturation between Christian, Muslim and Jew.
No musical manuscripts survive from Muslim Spain, but we know a good deal about music making at court and elsewhere in Moorish society from literary and other documentary sources. 33 The Arab influx brought many new instruments to the peninsula. Several of these, such as the ‘ū d (lute), rabā b (rebec) and naqqā ra ( nakers), subsequently enjoyed wider influence throughout Europe. The exceptional skills of Moorish craftsmen were used to full advantage after the reconquest by the Christian rulers; Moorish instrument-makers, as well as players and dancers, enjoyed high status in courtly circles. The cultural achievements of the Muslims were cherished, not spurned.
In the ninth century, a great number of classical Greek texts were translated into Arabic. This gave birth to an Arabic tradition of speculative writing on music which reached a level of great sophistication in its interpretation of the nature of harmony, melody and rhythm, providing an interesting counterpart to the more theologically charged discussions in the Latin West. 34 One of the greatest cultural influences from the Arab world on the reconquered territories came in the twelfth century, when many scientific and philosophical works were translated from Arabic into Latin. These included many musical writings, but also Greek texts which had until then remained unknown to Latin-speaking scholars. 35 Though the Andalusian frontier provided a conduit for the dissemination of these newly available sources throughout Europe, surprisingly little survives in the way of Latin music theory written in the peninsula before the fifteenth century. 36 There is nevertheless enough evidence to suggest that the study of music held a position within the Spanish scholastic tradition comparable with other countries, and the university of Salamanca was the first to have a chair of music, established in 1254 by Alfonso el Sabio.
Like the Muslims, the Jews were finally expelled from the Spanish kingdoms in 1492, when Ferdinand and Isabella's newly combined forces conquered Granada and ended the Muslim rule that had been in place there since 711. In the preceding centuries, the Jews of Spain (known as Sephardi from a Hebrew word interpreted as meaning ‘Spain’) had actively participated in the musical culture of the peninsula, both Christian and Muslim, incorporating elements of the villancico and other secular genres into their own music making.
Though no medieval sources of Jewish music survive from Spain or Portugal (from where the Jews were similarly expelled in 1497), the Jews’ musical traditions were maintained in exile, in particular by the Sephardi who settled in North Africa and parts of the Ottoman Empire. Much speculative work has been done to evaluate the extent to which later sources from other countries retain vestiges of Iberian musical practices. 37 The influence of the Castilian ballad, a secular genre that flourished in the fifteenth century, persists to the present day, and though textual similarities with medieval Iberian sources can be shown to exist, it is impossible to say with any conviction that the melodies known from later sources bear any relation to those that may have been sung across the peninsula in the fifteenth century or earlier. 38 Formal vestiges of certain of the Galician cantigas may likewise be found in the texts of some later, post-exile Jewish songs. The close ties between Arabic and Jewish culture brought about new styles of Hebrew verse founded on Arabic models, as well as a developed understanding of the theory of music in Hebrew treatises, again inspired by Arabic rather than Latin sources.
The tremendous diversity to be found in the musical life of the Iberian peninsula in the Middle Ages has perhaps been unduly neglected in the standard histories, which have often seen Spain as a peripheral concern in the context of the wider story of music making in Europe. While it is true that many aspects of the peninsula's musical practices had few ramifications beyond its borders, the numerous innovations and idiosyncrasies that characterize its history are increasingly coming to be understood as an integral part of European cultural history. Spain and Portugal are both countries where important new sources continue to be discovered, and the process of reassessment will continue in years to come.