10
Easily the largest single roofed building of the Roman empire, and larger than any subsequent building in Europe until Seville and Cologne cathedrals in the thirteenth century, was Justinian’s Great Church in Constantinople, dedicated to the Holy Wisdom (Hagia Sophia). It was built in under six years after a fire damaged the city’s earlier cathedral during the Nika riots of 532, and was dedicated in December 537: an unheard-of speed, then or later, for such an ambitious building. It was, all the same, built with considerable care, from the best materials, and has lasted little changed until the present day; the most significant modification was early, for the dome partly collapsed after an earthquake in 557 and was rebuilt, slightly higher, in the next years, allowing a rededication in 562, when Justinian was still alive. Subsequent emperors only tinkered with the building, for example adding a ceremonial door on the south-west porch (the work of Theophilos around 840), or else, later, adding external buttresses (Andronikos II in the 1310s and Sinan in the 1570s for the Ottomans; the Ottomans also added minarets for the Aya Sofya mosque that the church had become). The interior space remained the same, however; the only major Byzantine change here was the addition of figural decoration in the mosaics covering roof and upper walls, from the ninth century onwards, for Justinian’s decorative programme had above all been gold mosaic, sometimes interrupted by crosses or floral motifs, and coloured marble.
Hagia Sophia from the outside looks like a giant brooding spider, thanks to the Ottoman minarets. Inside, its central space shows itself at once as the major architectural innovation it was, with its great dome, 100 Roman/Byzantine feet (31 metres) across, balanced on four arches each 120 feet (37 metres) high, creating an unparalleled single volume, unbroken by pillars, which was further extended to the east and west by half-domes and then, to the east, a smaller half-domed apse. The whole is stamped with Justinian’s identity, for very many of the capitals have his monogram on, or else that of his wife Theodora. Most of the columns and all the capitals were cut especially for the building, unusually for the late Roman empire, where the reuse of building material was normal even for major monuments. Justinian intended the building to be innovative; he employed academic geometricians, Anthemios of Tralles and Isidore of Miletos, to build it, not, as was normal, master-builders. And people were duly amazed. In the context of the second dedication of the church, Paul the Silentiary wrote a verse description of it, which as a work dedicated to a single construction is unusual in our period, and a decade earlier, Prokopios’ On Buildings, written to praise all of Justinian’s building projects, starts off with an eleven-page eulogy to Hagia Sophia. Both writers stress its size, of course, and the effect of the gold and marble (the green marble was a meadow with flowers to Prokopios, fresh green hills and vines to Paul), particularly given the relationship between the gold of the roof and the light from the windows. ‘Whoever lifts his eyes to the beautiful firmament of the roof can scarcely keep them on its rounded expanse, sprinkled with dancing stars . . . whoever puts foot inside the sacred temple would never wish to leave, but would lift up his head and, with his eyes drawn first this way and then that way, would gaze around’ (Paul). And Prokopios was also well aware of the building’s architectural originality, for he spends two pages describing the dome-work with considerable technical detail, ending up by pre-echoing Paul with the observation - a cliché, but still true today - that it is hard to concentrate on one detail, given the arresting complexity of the whole. These descriptive works had their own literary tradition (artistic and architectural descriptions are called ekphraseis in Greek), and they were, furthermore, commissioned by or at least written for Justinian himself, but they at least tell us how the building was intended to be seen, the impact it was intended to have. It was an impact that lasted; Hagia Sophia was almost the first rectangular church focused on a central dome, but almost all later Byzantine churches used this model, in a simpler version at least, and so did Sinan’s mosques for sixteenth-century Istanbul.
Hagia Sophia was not just a huge, expensive and innovative building, one of many which Justinian erected, as Prokopios tells us at length. It also sat at the apex of the ceremonial life of the east Roman empire. In Rome itself, the new churches of the Christian empire were built outside the walls or on the edge of town for a long time, decentring the old public focus of the forum complexes, the imperial palace on the Palatine hill above them, and the great racecourse, the Circus Maximus, to the latter’s south. In Constantinople, Constantine’s new foundation, these public spaces could be and were put together, with the forums leading in a line along wide colonnaded streets to the Great Church, and the palace and the Hippodrome just to its south. The people of the city regularly met in the Hippodrome, and, although access to the church was often more restricted, many thousands could get into Hagia Sophia. The ceremonial of imperial life had as its centre movements between the palace and the church, which were watched by an audience, and public processions regularly proceeded through the forums to the church- palace, attended by even more spectators. The church that was there before 532 was already large for these same reasons, but the size and ambition of Justinian’s church set his own mark on the entire public and ceremonial space of the largest city in Europe, for close on a millennium. Justinian’s church was remembered by later generations in the same breath as his legal codifications and his conquests, and if there is one act which sums up his desire to be recognized as the ideal or archetypal Roman emperor, Hagia Sophia could be seen as that.
This might seem a lot of weight to put on a single building, but the Romans intended their constructions to be seen as representative of their power and wealth, and, judging by the numerous reactions we have in written texts, they indeed were so. People could build buildings with quite complex inter-textual references, too; in Hagia Sophia’s case, the rotunda of the Pantheon in Rome, or that of Galerius’ palace at Thessaloniki, were models to be surpassed, as also was the private church of Hagios Polyeuktos, built on a huge scale in Constantinople only a decade before by the imperial heiress Anicia Juliana, this time in a more conventional basilica form, which Hagia Sophia could displace simply by being so different. The force of the politics of building was not restricted to the Roman empire, either. All the societies described in this book recognized it and respected it, in fact; and the differences between the buildings which powerful people erected in those different societies is one quick way to understand the variation in their aspirations, both in their scale and in their aesthetics.
This central chapter, accordingly, is intended to be comparative. It sets societies against each other through their different uses of material culture, particularly architecture, for the purpose of display. We seldom have as clear an idea of the intentions of the patrons of a building as we have for Hagia Sophia, thanks to Prokopios and Paul the Silentiary; but we do have many of the buildings themselves, or at least their archaeological vestiges, and we can reconstruct some of these intentions. I cannot do justice to all the societies in this book in a single chapter, of course, but I can at least give a sample of the sort of comparative analysis of display that can be achieved. We shall look in turn at four buildings: Hagia Sophia, already discussed; the Great Mosque of Damascus; the Northumbrian palace complex of Yeavering; and the church of S. Prassede in ninth-century Rome. These buildings are mostly religious, for the survival of secular buildings is much more patchy (Yeavering, indeed, only survives as a set of post-holes), but at the end of the chapter we shall look briefly at the varying structures of royal palaces, and also - outside the restricted world of kings, emperors and bishops - at the changing spatial patterns of villages, for these too are a guide to power, on a smaller scale.
Caliph al-Walid I (705-15) had the Great Mosque built in his capital at Damascus in 705-16, finished after his death. It was not the first mosque in the territory of the caliphate, most of which had been conquered by the Arabs sixty to seventy years before, between 636 and 651; but, together with contemporary constructions in Medina and Jerusalem, it was the first large-scale monumental mosque, and it set a pattern which would be largely repeated in subsequent building projects, in Fustat (Old Cairo), Kairouan, Córdoba and many other cities. Mosque architecture used many elements of Roman (and also, in Iraq and Iran, Sassanian) architectural style, including colonnades; indeed, the columns were for a long time characteristically spolia, taken from Roman buildings and reused. The Damascus mosque also had a marble vine-frieze, much praised by medieval writers, which has clear Constantinopolitan antecedents. But the overall effect of an early medieval mosque was quite different from that of any Roman building. It consisted of a walled rectangular courtyard, part of which was open to the sky and part roofed, the latter making up a deep space held up by lines of columns. Sometimes the roofed space was quite small by comparison with the courtyard; sometimes (as at Damascus) it was around the same size, with three lines of columns in that case. (The famous forest of columns in the later Córdoba mosque, with thirty-four lines of eighteen columns in its last phase, is atypical.) The effect was, however, of a relatively unhierarchical space, the open courtyard running into the roofed area without a complete break, with only the mihrab, a niche pointing towards Mecca, operating as a focus. Islam is not a religion with an organized priesthood, and it puts great emphasis on a community of believers. Inside the courtyard, opened and roofed alike, the community could meet in prayer.
The Damascus mosque also had a specific political and spatial symbolism, by no means only directed to Muslims. For a start, al-Walid built it on top of the demolished cathedral of the city, which still had a Christian majority, in a particularly overt assertion of Muslim supremacy. This formed part of the monumental rhetoric he inherited from his father ‘Abd al-Malik (685-705), who was, as we shall see in Chapter 12, the first caliph to publicize Islam on a large scale in material form, in coins and monumental buildings; al-Walid simply developed it further, including by bringing it to the capital itself, right at the start of his reign. Like Hagia Sophia, the Great Mosque is very large, with a courtyard 157 by 100 metres in size, and was hugely and visibly expensive. The courtyard used the walls of the precinct of the pagan Roman temple of Jupiter, which the Christians had already left around their cathedral, but that enclosure was now turned into a specific walled-off Muslim religious and political space, reserved for the new Arab ruling class of Damascus. The walled courtyard constituted a typical element of the mosque for ever after. Al-Walid put four minarets at its corners, perhaps to show to all that the old Roman space had a new function; but this was the only important feature of the mosque not to have a later history, for a single tall minaret is characteristic of most later major mosques.
The effect of the Great Mosque was not, however, restricted to its scale and to its appropriation of a former sacred site. Al-Walid had the monumental upper parts of his roofed space, looking out onto the open courtyard, covered in mosaics, probably the work of Byzantine mosaicists; mosaics also covered much of the walls of the roofed space, and the other walls of the courtyard. Sections of these survive; they consist of trees and foliage, interspersed on the courtyard walls with buildings and a river - paradise imagery in all likelihood - of remarkably high quality, but with no representations of humans or animals. This marks out a new style of visual programme. Mosaic decoration was normally figurative in the Roman world, in public buildings and churches alike (Hagia Sophia was atypical in this respect), and vegetation was at best used as a background, or as a divider between scenes. Here, the caliph was making very obvious indeed the fact that the new Islamic religion was beginning to avoid human representation in public spaces (it matched the new coinage developed under ‘Abd al-Malik, too, which abruptly abandoned pictures of caliphs in 696). The importance which representation came to have in both the caliphate and the Byzantine empire will be looked at again in the next two chapters; but the Great Mosque is one of its earliest signs.
The other important feature of the Great Mosque was that, as a space, it was closed off to the outside. Roman cities were structured by wide streets leading to central forum areas, to which processions led and where public participation could be considerable, as continued to be the case in Constantinople for centuries. Amphitheatres (in the West), theatres and racetracks were other major venues for public activity, and the Hippodrome of Constantinople carried on this tradition for a long time. In the Islamic world, the mosque courtyard took over from all of these; major political events, like collective oaths of loyalty, took place there, not in any secular location. And the Arab states did not use processions as a major part of their political legitimization; the assembly in the mosque courtyard was sufficient for that. The need for wide boulevards ended; pre-Islamic Syrian and Palestinian colonnades were quite quickly filled in with shops in the eighth century, some of them commissioned as public amenities by caliphs. The narrow streets of Islamic cities resulted directly from this, for there was no public interest involved in keeping them clear from obstructions like vendors’ stalls, beyond a certain minimum (enough for two loaded pack animals to pass each other, later jurists said). Public display came to be focused on the mosque, and, secondarily, rulers’ palaces and city gates, rather than on the cityscape as a whole. The impact of al-Walid’s mosaics would have been all the greater as a result, although that would be a future development, only set in motion in the eighth century. The caliph and his advisers were nonetheless making a set of conscious symbolic and political points by organizing the Great Mosque as they did; and the way the public space in Islamic cities changed, to focus so exclusively on mosques, although less conscious as a process, would have seemed to them auspicious and fitting. In a time when the population of Syria was still mostly Christian, and Greek- or Syriac-speaking, these changes were also probably the most immediate signs they had of the content of the Muslim religion of their new rulers.
At the other end of the former Roman world, in the Cheviot hills of Bernicia (now Northumberland) just south of the modern Scottish border, King Edwin (616-33) of Northumbria had a court (villa) called, according to Bede, Ad Gefrin. There is no serious doubt that this villa was at Yeavering, where in 1949 air photography allowed the localization of a complex Anglo-Saxon site, which was excavated in the 1950s and published in 1977. This site had lost its topsoil and floor levels, and with them most of the small finds one would normally expect, though it has to be said that the site was, even then, unusually poor in finds for such an important centre, which underlines how limited the resources of early Anglo-Saxon kings were. But in compensation the post-hole foundations of a variety of wooden buildings were identified, which show us a much more elaborate picture of an early Anglo-Saxon palace complex than researchers had previously expected.
Literary images of royal palaces in Old English texts concentrate on a single wooden hall, like Heorot in Beowulf, where kings and their retainers met, feasted and slept. Yeavering was both less and more than that. In the late sixth century the Anglo-Saxons had found an earlier stone circle, a Bronze Age barrow, and a large fortified enclosure, some of which seems to have made up a British cult-site. This was further developed by pagan Anglo-Saxons, with small buildings which may have been temples. In the middle of the site, around 600 or so, a building unparalleled in Anglo-Saxon England was set up, consisting of a dais and banked seats looking down on it, 20 metres from front to back, the whole looking in plan (all that survives of it) like the cross-section of an orange segment. This construction most resembles a section of a Roman theatre, imitated in wood, and its parallels are firmly Roman. It is generally and convincingly interpreted as an assembly point for the Bernician, and later Northumbrian, aristocracy when they and the king came to the Yeavering cult-site. A few years later, the site turned into a more typical palace complex as well, with the construction of a line of large halls, some 70 metres long in total, pointing straight at the apex of the ‘theatre’. This was the setting for a set piece of Christian conversion and baptism in the 620s by Paulinus, an early missionary to Northumbria, which explains Bede’s references to Ad Gefrin. In these halls, which were occupied until around 650, Edwin could easily have lived the sort of life described inBeowulf and similar literature; but they were surrounded by a network of earlier architectural representations looking in other directions too.
The pre-Christian Anglo-Saxons settled in a British landscape, but took relatively little from their predecessors and neighbours by way of material culture, even though the British were overwhelmingly dominant numerically. Yeavering was right at the edge of a relatively narrow Anglo-Saxon coastal settlement around the Bernician royal centre of Bamburgh; it may not be so surprising that we find here one of the few documented examples of a British site (and British religious practices, maybe also pagan in this area, north of Hadrian’s Wall as it was) having a cultural influence on an Anglo-Saxon one. But, given that, the ‘theatre’ is all the more striking. We are not so far from the wall here, and Roman material culture was thus at least physically available to the Bernicians; but for Anglo-Saxons living north of the Roman province of Britannia deliberately to adopt a Roman-influenced construction for something as emblematically Anglo-Saxon as a public assembly point sheds considerable light on royal aspirations, particularly because it seems to predate Christianization, which would make Roman influences more obviously culturally attractive. Indeed, this may go some way to explaining the readiness of Anglo-Saxon rulers to be converted relatively quickly. And that Roman imagery presumably made sense to an aristocratic and possibly also popular audience too. The early Anglo-Saxons are sometimes depicted as finding the Roman past incomprehensibly grandiose, as in the Old English poem The Ruin, plausibly about Bath, which refers to the Roman buildings of a city as ‘the work of giants’. However that may be, they could deal with elements of that past with the same sort of creative and expressive bricolage that we find in Arab Syria. Cosy primitivist readings of Anglo-Saxon ‘barbarism’ are out of place here. The early Anglo-Saxons did not have access to a technologically complex material culture, but despite this the culture they did have could be manipulated in complex ways, with images of legitimacy taken from Anglo-Saxon, Roman and non-Roman British culture all at once.
For our next example, let us move to the Rome of the Carolingian period, by which time the Franks ruled most of western Europe, including a protectorate over the papal city. We shall look here at the building programme of Pope Paschal I (817-24), which was very extensive for what was not a very long reign. Paschal was a controversial pope, who built up an influential set of aristocratic enemies. Although his enemies could draw on Carolingian support, in 823 Paschal had several of them executed, and fiercely defended the executions to Carolingian emissaries. He was not afraid of much, it seems, and his church-building, which includes two of the three largest churches in Rome built between the sixth century and the twelfth, testifies to his confidence. I shall focus here on the earliest and best-surviving of them, S. Prassede, built around 820.
Paschal was not the first builder-pope of the late eighth and ninth centuries. There was probably no break in papal building in the early Middle Ages (and there was certainly no break in reconstruction and repair), but our sources, both written and material, concur that there was more new church construction than before, from S. Silvestro of Paul I (757-67) and SS. Nereo e Achilleo of Hadrian I (772-95) onwards, up to the 850s at least. These churches were all built on a standard basilica plan, looking directly to the great church of St Peter’s in the Vatican, originally founded by Constantine; they constituted a self-representation of the unbroken continuity of papal legitimacy and centrality. In three churches, S. Prassede, S. Cecilia and S. Maria in Domnica, Paschal simply did this on a rather larger scale. S. Prassede, some 50 metres long excluding its courtyard, has expensive internal finishings, such as good-quality reused columns and a good deal of marble, some of which is still in situ; it also had a remarkable quantity of gold and silver furnishings, as the near-contemporary biography of Paschal in the Liber Pontificalis informs us, including a silver canopy weighing 910 pounds, and a silver image of St Praxedis herself on her coffin in the crypt, weighing 99 pounds. The eye is caught today by the dramatic quality of the mosaics in the apse and triumphal arch, and in the side chapel of S. Zenone, a burial chapel for Paschal’s mother, Theodora. The apse mosaics, of the risen Christ and associated saints (including Praxedis), with a portrait representation of the pope, copy those of the sixth-century church of SS. Cosma e Damiano in the forum, built by Felix IV (526-30), and are a further sign of Paschal’s concern to show himself as part of an unbroken papal tradition. It is worth remembering, however, that the Liber Pontificalis, while mentioning the mosaics, puts rather more stress on Paschal’s gold and silver gifts, and also on the pope’s clearest innovation in S. Prassede, the moving of a large quantity of saints’ bodies from Rome’s catacombs to the church, which a contemporary inscription claims to have numbered 2,300 in total.
Paschal had a variety of audiences - one could well say targets - for his activities in S. Prassede. One was the Byzantine emperors, who in 815 had readopted Iconoclasm, a hostility to holy images of God and the saints (on which more in the next chapter) which the pope was in the front line of opposition to. Paschal wrote critical letters to Constantinople about it, and sheltered Iconophile monks in Rome; S. Prassede indeed was endowed with a community of Greek monks, who must have been part of the Iconophile observance. In the context of the material culture of the church, the numerous mosaic figures in S. Prassede’s apse were too traditional a set of motifs for their detail to be a specific response to Iconoclasm, but the expense of Praxedis’s silver image is quite likely to have been. It must be added that Paschal here could well have had an eye on Frankish Iconoclasts too (see Chapter 17). Only fifteen years before, Theodulf of Orléans (d. c. 826) had constructed his intriguing and unique monument to his Iconoclast beliefs, the private chapel at Germigny-des-Prés on the Loire, whose apse mosaics show two angels (not human, so acceptable to represent) and the ark of the covenant. These representations were a polemical response to some of the arguments of Byzantine Iconophiles, and had a complex relationship to Old Testament interpretation, as much of Theodulf’s own writings had; they thus show how theological positions could have quite a detailed effect on western visual imagery in this period. Paschal is unlikely to have known about Theodulf’s chapel (and his Roman audience is unlikely to have heard of Theodulf at all), but he knew of Frankish Iconoclast sympathizers such as Claudius of Turin, and he opposed them explicitly; S. Prassede could at least serve as a visual reassertion of the centrality of Roman and papal traditions and the superiority of papal positions on the matter of religious belief. Paschal’s buildings responded to a network of contestations of papal positions simply by, so to speak, repeating themselves, but louder.
Paschal had two other audiences for his building campaigns. One was the Frankish court itself, to whose power in Rome he was perhaps the firmest opponent in the Carolingian period. There were always Franks in Rome by now, as pilgrims (as we saw in Chapter 8), but also, at the political level, as emissaries and dealers; they were expected to see what the pope was doing, and to report it back northwards. They would have reported that Paschal’s churches were not just larger and pricier than those of his predecessors, but were as large as those of the Carolingians themselves; and they challenged monuments like the octagonal royal chapel in the palace at Aachen by, once again, their traditionalism - Rome had no need of Carolingian protagonism, including its moral reform programme; it was simply itself, and could carry on as before.
The other audience was the Romans themselves. Paschal was like Justinian and al-Walid in building big to impress a local audience, the people who would be inside or near S. Prassede most often; the church was indeed on one of the major processional routes of the city, leading out over the Esquiline hill to the basilica of S. Lorenzo fuori le mura. All the major popes of the century after Paul I were builders, indeed, and it is arguable that it was their collective intervention, above all in church-building, that did most to make Rome into the ‘papal city’ that it remained for the next millennium. But it is Paschal’s appropriation of so many relics which marks his position most clearly here. Rome had a highly dispersed array of cult-sites, scattered across the huge field of ruins that the city had become, and based on the burial places of numerous martyrs and other prominent Christians of the pre-Constantinian period; they extended, in particular, way out into the countryside in Rome’s extramural cemeteries. These were hard for popes to protect, as Paul I already recognized (he imported several saints into papal churches inside the walls after Lombard attacks); this became all the more pressing in the early ninth century, given a growing Frankish obsession with Roman relics, which by the 820s extended to outright theft (below, Chapter 17). The sites were also hard for popes to control politically; the churches associated with these scattered cults had local communities and aristocratic families as patrons, quite as much as they were under papal patronage. To empty them of 2,300 saints, who were to be transported to a new papal prestige foundation, was thus a notably authoritarian move. It cannot have contributed to Paschal’s popularity, which as we have seen led to contestation in 823 by some of the aristocratic officials of the papal hierarchy; but it was certainly an assertion of his power - and anyway he had defeated his opponents before he unexpectedly died in 824.
Aachen was only the biggest of a long sequence of Merovingian, Carolingian, Ottonian palaces across the centuries in the Frankish world. Most of the others do not survive, and have not even been excavated (exceptions include the Merovingian Malay and the Ottonian Tilleda, both fairly small, and some rather grander complexes, such as Carolingian Ingelheim and Compiègne, and Ottonian Paderborn); but they are described occasionally in detail, in written texts.
Palaces were long-standing sites of royal or imperial rhetoric, aimed to impress both royal subjects and ambassadors or other visitors from outside. Even in societies where kings lived in single wooden halls, these were seen as remarkable, ‘greater than the children of men had ever heard tell of’, as the Beowulf poet said of Heorot, and acting as metonyms for the fate of the kingdom itself, as with Cynddylan’s hall, ‘dark tonight, without a fire, without a bed’, as a ninth-century poet wrote of an eastern Welsh king after his death in battle. They were barred by élite guards who would only let in appropriate people, as with Hrothgar’s court-officer Wulfgar in Beowulf, or Arthur’s door-keeper Glewlwyd in Culhwch ac Olwen, a Welsh text of the eleventh century; this added to the honour involved in entering them and participating in the Königsnähe (‘closeness to the king’) inside. These are heroic texts, in which everything and everyone is larger than life; the east Roman ambassador Priskos was less amazed at Attila’s very similar palace complex in 449. But he describes it neutrally and with respect, as a splendid hall made of planed wood, surrounded by other buildings, including dining halls and colonnades, some carved and well constructed, the whole in a wooden enclosure with towers, ‘with an eye not to security but to elegance’. The furnishings inside, in Attila’s case in linen and wool, and multi-coloured hangings ‘like those which the Greeks and Romans prepare for weddings’, were also designed for effect. Yeavering probably had this sort of impact, too, on a smaller scale, which would have been all the greater for visitors from smaller centres than Constantinople.
Frankish royal palaces, or at least the major ones, were more complicated than this. Carolingian Ingelheim consisted of a set of large rooms (including a royal hall) built in stone, arrayed around a substantial apsed and colonnaded courtyard, 100 by 70 metres in size, some of which still stands. This was also (apart from the apse form) the case in Aachen, where the scale was larger. This can still be seen from one part of it, the palace chapel, with its internal marbling, nineteenth-century replacements for its rich mosaics and a throne standing in the gallery. The numerous rooms in these palaces, which visitors had to pass through, were doubtless set out for effect. Merovingian sources already make this clear, as when Duke Rauching was shown into King Childebert’s private chamber, probably in Metz, in order to be killed in 589 (as described in Chapter 5). But it is Carolingian sources which stress most clearly the intricacy of royal or imperial display. Ingelheim in the 820s was described in a poem as having ‘a hundred columns, with many sorts of buildings, a thousand entrances and exits and a thousand inner chambers’, as well as having an elaborate painted programme in the church drawn from the Bible, and an even more striking decorative programme in secular areas of the palace, featuring classical heroes and Christian Roman emperors, and leading up to narrative scenes featuring Charles Martel, Pippin III and Charlemagne himself. Notker of St. Gallen, in his Deeds of Charlemagne of the 880s, imagines a fantastic story in which Byzantine ambassadors to Charlemagne come into the palace (which palace, Notker does not say), and go through groups of nobles, four times convinced that the central figure must be grand enough to be Charlemagne - one is on a throne, another is in the emperor’s private apartments - before they are finally ushered into a separate room, and into the presence of Charlemagne, clad in gold and jewels and glittering like the sun. Notker never went to court, but he had talked at length to senior courtiers, and his image of the spatial complexity of the court rings true, given what we know of the big Carolingian palaces. It may be added that the material culture of display was here focused as much on the dress of human beings as on the walls of the palace (Priskos noted the same of Attila’s residence); but if Ingelheim was typical in its decoration, the walls all conveyed meaning too.
Palaces competed in order to impress. The Franks could not match the display of Constantinople, where Liutprand of Cremona, envoy for King Berengar II of Italy in 949, was so struck by the mechanical singing-birds and the mechanical lift under the throne of the Magnaura palace only one of many buildings in the palace complex. But inside the material cultural possibilities of Latin Europe, visiting Carolingian Aachen and Ingelheim was as complicated and overwhelming an experience as anything available. Notker also claimed that the ‘ever-vigilant’ Charlemagne could look down from the windows of his chamber at everyone in the palace, including at what was happening in the houses of his aristocrats, so as to see ‘everything they were doing, and all their comings and goings’. This precursor of Jeremy Bentham’s (and Michel Foucault’s) panopticon, even though once again doubtless an imaginative flourish on Notker’s part, shows the degree to which such palaces were expected, in all their complexity, to be under the direct control of the king/emperor as well. For that complexity itself made royal power visible, and therefore had to reflect, at least ideally, the concrete operation of that power, that is, knowledge, and, when necessary, coercion based on that knowledge.
Our evidence for village layouts is entirely archaeological, and here I can only discuss a few examples out of a hundred or more. They do not tend to demonstrate any conscious planning, and were built up out of individual farm units, by the peasants themselves. The way this build-up occurred varied from region to region, however, and also across time; it demonstrates changes in sociability, sometimes in village-level competitiveness, and in village hierarchies. In particular, in the last third of our period, the growing internal hierarchies of western European village society began to take material form.
In the western Roman empire, villages were relatively rare. The rural landscape was certainly hierarchical, with the villas of landowners operating as estate-centres for a dependent population, and indeed often acting as highly ambitious monumental complexes, designed to impress aristocratic peers; but the peasant majority in most areas lived in houses scattered across the landscape, without any obvious sociability. One has to move to the East to find nucleated village complexes, and some very striking ones still survive in the landscape, particularly in southern Anatolia, Syria and Palestine. The villages of the north Syrian Limestone Massif, rich from oil export, and lucky in their long-lasting and easy-to-carve stone materials, are the best-surviving of this set, and have been the most systematically studied. The village of Serjilla, for example, is a complex of a church, a community building, a bath-house, and nearly twenty houses, some still with a second storey and a roof, each in its own courtyard, in no obvious spatial order. The houses vary in their scale, most of them having four rooms or less, but some with substantial extensions; they have similar decoration, with regular (if severe) carved surrounds to doors and windows, and quite elaborate roof pediments, but they vary here too; many, for example, have internal colonnades. Building size may indicate family size, but it indicates resources too; architectural elaboration indicates ambition. So also does the village bath complex - a relatively unusual amenity for a village - which was put up in 473 by Julian and Domna, as a mosaic inscription tells us. There is no sign, all the same, that the inhabitants of Serjilla were anything other than peasants (or stone-workers); no house is typologically distinct, as would befit a residence or rent-collection centre for a landlord. But they must have been remarkably prosperous as a group, some of them doubtless with a few tenants of their own (Julian and Domna for example), and also mutually competitive on a substantial scale, above all at the height of Syrian oil production in the fourth to sixth centuries. For once, the density of surviving housing in the Limestone Massif allows us to track that competition through display in some detail. When we have written accounts of eastern village societies, they often appear as very fractious; the buildings of villages like Serjilla allow us to trace that tension on the ground. But the absence of a clearly marked élite housing is all the more striking. It may indeed have made the fractiousness much worse, for society was not formally stratified, and there was more to play for.
Village societies existed in the Germanic world north of the Roman empire; and similar villages also crystallized in the post-Roman kingdoms of the West, sometimes doubtless under the influence of incoming Germanic groups, notably in the case of the Anglo-Saxons in England, though the village-form also emerged in parts of the post-Roman West where there were relatively few incomers, such as southern France or central Italy. When villages did appear, they were in wood, overwhelmingly the dominant medium for rural housing in Europe until after 1200, except for parts of Mediterranean Europe, where stone came in a century or two earlier. We cannot track local relationships with the density possible for Serjilla, partly because houses only survive through their post-holes, and partly because they tended to be even more uniform. In a substantial area from Denmark to the Alps, and west into central France and England, villages were made up of farmstead blocks, centred on a main building (very long in Denmark, up to 40 metres sometimes, where it included living quarters and an animal byre; rather shorter in southern Germany or England), with subsidiary buildings and sunken-floored huts, which seem for the most part to have been outhouses for artisanal activity and storage, the whole set in a yard, usually fenced. The squares of each farmstead often created quite regular patterns for these villages, set on either side of a main road or else in a block around a crossroads, a regularity further enhanced because longhouses and other main buildings tended quite often to be parallel to each other. This regularity enhanced the sense of uniformity created by similar house and farmstead plans.
Farmstead units were not all alike, though. In some villages, they were; but there was often one rather larger house, often on the edge of the settlement. Vorbasse in Denmark is a good example of this, for this settlement, like many in Denmark and the area of northern Germany and the Netherlands, regularly shifted site inside the same agricultural territory, and changes in its patterning can thus be more easily compared from century to century, in this case (unusually) from the first century BC to the twelfth century AD when it settled down on its present site. From the third century to the start of the eighth, there was always one rather larger farmstead in Vorbasse, half again as large as its neighbours, with a bigger main building, and more imported goods found in it. In one of the village’s shifts, around 300, the rest of the village shifted but the larger farmstead stayed put, which marks the stability of the social position of its owner even more firmly. Vorbasse evidently had a leader, then; but we would be wrong to see him (or her) as a local lord, still less the village’s landlord. His house was larger and richer than those of his neighbours, but, as at Serjilla, it was not otherwise different. And it is interesting that around 700, when Vorbasse reorganized itself more substantially than usual on a new site, the larger farmstead disappeared. Leadership had been stable here for a long time, but was not so structurally solid that it could not be sidestepped, even if we cannot tell precisely how.
Lauchheim in Alemannia, in the upper Danube valley, settled from the sixth century to the twelfth, shows a more hierarchical pattern. Here there were around ten farmsteads along a road, but one was much larger already by the seventh century, and became twice the size by the eleventh. Here, the larger farmstead was indeed typologically distinct, for it had a much larger number of non-residential post-hole buildings, probably for storage, plausibly of grain and other produce collected in rent. It also had its own small cemetery, with rich burials, in the late seventh century, before burials moved to the churchyard of the eighth-century church. It would be fair to call this central farmstead of Lauchheim an estate-centre, and it is quite possible - even if not certain - that its holder was, or became, the landlord of the village as a whole.
Most excavated villages had some sort of identifiable hierarchy, at least of resources, which probably points to village leadership, too; but it was not always stable (different houses could be the largest one in a village in different centuries), and, even in Lauchheim, that hierarchy did not point to a radically different lifestyle for the inhabitants of the largest house. From the Carolingian period onwards, however, we begin to get signs of structural differences. The first innovation was characteristically the village church, often built in stone from the start; village church-building tended to begin after 700, and to gain pace in later centuries (see above, Chapter 8). Once a church was built in a village, the settlement tended to gain a more stable spatial structure (and, in Denmark, to stop moving site); and churches always had aristocratic or local patrons and, generally, resident priests, whose élite status was reinforced by the considerable investment church-building involved. The most striking change came, however, when local leaders or lords began to fortify their residences. This development, which can be summarized succinctly (if simplistically) as ‘the rise of the castle’, was rare before 900, and not widely generalized until after 1000 (below, Chapter 21), but came in the end to characterize most of Europe. It happened in different ways in different places; in some places, Lauchheim-type estate-centres gained bigger fences, then ditches, then stone walls, then stone residences, perhaps on an artificial hill or motte above the village; in others (as in England) some lords had moved into increasingly fortified residences, which had no necessary connection with still-fragmented peasant settlement, by perhaps 900, a long time before mottes appeared in the wake of the Norman Conquest. In central-southern Italy and other parts of southern Europe, villages were themselves fortified in the tenth to twelfth centuries (and called castra or castella, castles), with a lord’s residence developing as an internal fortification (a rocca or cassero) inside the village. In each case the relationship between the castle-dwelling lord and the village or villages around was different, the difference being very clear on the ground.
A good example of that Italian development is Montarrenti near Siena in southern Tuscany. Here, a village on the slopes and summit of a hill is documented from the late seventh century onwards; the houses were small and one-roomed, as is typical for Italy; the whole village already probably had a palisade around it, and the hill-top had a separate fence, although the houses there were of much the same size. Already in the early ninth century the hill-top palisade was replaced by a stone wall, surrounding a large wooden building (probably a granary), a grain-drying oven, and a mill-stone: as at Lauchheim, one can see an estate-centre crystallizing here. This burnt down later in the century, but the wall was rebuilt in the tenth. The hill-top still had wooden houses in the tenth century, but in the late eleventh stone towers began to replace them, to create a clearly seigneurial focus. Settlement continued lower down the hill, however (perhaps with breaks), and by the twelfth century the lower hill-slopes were themselves surrounded by a wall which included the whole village, by now mostly built of stone, although the upper cassero remained the seigneurial centre. It was this whole village that was called a castrum from the end of that century (I would guess by the early eleventh, judging by other Italian examples, but we do not have the documents for Montarrenti), but it had a clear settlement hierarchy in it, one which had begun already in the ninth century, and which was permanently fixed in the towers of the eleventh. This sort of articulated spatial hierarchy has plenty of parallels in the settlement archaeology of Tuscany and Lazio (two well-studied regions of Italy), and has, as we shall see in Chapter 22, clear analogues in our documentation for the increasingly militarized social hierarchies in tenth-century villages, for Italy and elsewhere. The social hierarchy, however, was made increasingly manifest and solid in village architecture. Once village élites moved to stone towers, they were making visual claims to status and lordship, which they could back up by armed force, and which were no longer negotiable, as village leadership had frequently been two or three centuries before. Display here was not intended to compete with neighbours, but to exclude them.
The display involved in building huge prestige constructions like Hagia Sophia and that involved in building a flashier village house was quite different in scale, but it had many of the same aims: to impress, to establish status and power, maybe to elicit fear and submission. (The two types of display were even sometimes linked; the decoration of the emperor Zeno’s huge church to honour Simon the Stylite in the 470s at Qa‘lat Sim‘an in the Limestone Massif was copied by village church builders all over the region in the next generation.) The frames of reference in which display operated varied very greatly from region to region and between different types of building, however; and it is that variability which tells us most about cultural assumptions. Justinian was bouncing his architectural references off earlier pagan buildings and near-contemporary churches, all of which he was aiming to surpass. Al-Walid was aiming to surpass, too, but was also aiming at establishing a fundamental difference from past styles of building-plan and mosaic decoration, to mark out the novelty and superiority of the Muslim religion. Edwin and his predecessors were making claims to links with a Roman past which evidently had local prestige, even though Yeavering was north of the former territory of the later Roman empire. Paschal was reasserting papal centrality through unbroken links with past architectural and decorative styles, inside and against a world with quite different political configurations. Charlemagne was offering his visitors a visual and spatial experience in his palaces which had no recent parallels in northern and western Europe, and which was intended, doubtless successfully, to mark him out as unique, at least to people who had no experience of Constantinople. The lords of Montarrenti were not just showing their local mastery and their defensive capability with their stone towers, but were also drawing on urban models of building, and thus transferring urban power and cultural prestige into the construction of rural lordship; and, in a less top-down and more competitive way, it was also urban prestige that was evoked by Julian and Domna’s bath-house.
The intervisuality of architectural style is one of the most powerful conveyors of meaning and visual effect. As remarked at the start of this book, archaeology, and the study of material culture in its widest sense in art history and architectural history, tends to tell us different sorts of things from the study of narrative and documentary texts. Material culture tells us more about the use of space, the function of spatial relationships, as well as, of course, stylistic and technological changes; written culture tells us more about human relationships, choices, conscious representations of the world around us. But the construction of visual meaning, by emperors and peasants alike, links these two worlds: it is material culture, not words, which tells us about the choices of al-Walid, or Paschal, or Julian and Domna in Serjilla. That is why this chapter is the central one in this book; it offers a way to compare the strategies of every actor in the early Middle Ages, rich or poor, and not - for once - just those who had access to the written word. And the audiences of buildings such as these were also far wider than those of any written text, save of the sections of the Bible and Qur’an most often read out in religious ceremonies, and these latter tended not to change much across time and space. The whole population of Europe was thus involved in the communication discussed in this chapter, and could even, if they chose, participate as communicators, not just as audiences. Indeed, as archaeology makes its inevitable advances in the future, this is a sector of historical knowledge which, for a change, we shall know progressively more about.