14
Romanesque Architecture
There is something very odd about the Romanesque style as it is currently defined, namely how it is supposed to have begun. As originally used, in the eighteenth century, the word “Romanesque” referred to the Romance languages, those which had become what were considered corrupted versions of late Latin over the course of the first millennium. Thus when William Gunn in 1819 first applied the term “Romanesque” to architecture, he used it to cover all the masonry buildings of Western Europe between the Roman period and the Gothic. To underline the parallel with the languages, he cited the difference drawn in the Rome of his day between a Romano, someone who was unarguably a citizen, and a Romanesco, an inhabitant of dubious origins. Romanesque therefore meant not properly Roman, or literally Roman-ish.1 The current definition, introduced in France in the late nineteenth century, is very different in that it restricts the style to the last two or three centuries of the longer period. Whereas the old long period was a continuation of the Roman arising directly out of changes in Roman culture, the new short one, while retaining a strong link with the Roman past, has no obvious historical context or period of social change to help explain it. In addition, for some scholars it begins as late as the middle of the eleventh century and for others as early as the second quarter of the tenth, while proposed places of origin lie as far apart as Lombardy, the Loire Valley, and Saxony.
Given these uncertainties it is worthwhile asking if the new style is a convincing historical phenomenon or merely the result of an academic exercise. It is supported by the clarity of its main characteristic, which is most often seen, in all the visual arts but especially in architecture, as the articulation of parts from smallest to largest, forming clear geometrical shapes which relate to one another in understandable ways.2 The early eleventh-century church of St Vincent at Cardona in Catalonia can be used to exemplify the style, in that it is composed of clearly readable masses and volumes and has an interior which can be determined from the exterior. Equally important is the consistency of the changes which take place within it, as for example with the differences between the late eleventh-century portal of St Etienne in Caen and the portal of a century later in the castle hall in Durham, two Norman buildings which illustrate a classic development from simple to complex. Patterns of change like these imply no mystical “life” of forms, but rather the psychology of use and enjoyment, of boredom and invention, as such establishing that the masons responsible were working within a tradition. This evidence leaves little doubt, then, that we are dealing with a recognizable phenomenon, and from now on in this chapter I shall be using the term Romanesque for this restricted meaning and period, with uncertainty only over the date of the start of the period.
The diffusion in time and space of the origins of the Romanesque suggests that they are likely to have been caused or accompanied by a major, all-pervading change, so it is appropriate to begin by looking at what has been called the most momentous change in the history of the West in the first millennium, that is, the end of antiquity (bearing in mind that all periods are artificial impositions on the past, which we need, as Wölfflin put it, to keep us sane). The end of antiquity in the West has traditionally been associated with the deposing of the last western Roman emperor in 476, but the old administrative, social, and trading structures clearly continued beyond this date. A more significant break has therefore been identified in the Islamic invasions of the seventh and eighth centuries. They are seen as a symptom in that they only happened because of the weakness of the Empire, but they are also seen as a cause: as the Roman Empire consisted not so much of an area of land as of the Mediterranean sea plus the territories around its shores, when the invasions cut control of the sea in two they severely affected the world-view of its inhabitants, one which had kept the structure, if not the name, of the western empire going in even the most dire of circumstances. For anyone used to thinking of Carthage, Alexandria, Jerusalem, and Antioch as part of their world, the invasions must have had a devastating effect (figs. 14-1, 14-2). The rupture had the effect of forcing the Frankish rulers of the northern parts of this lingering remnant of antiquity back on themselves, leading them to form a state which, under Charlemagne in the year 800 and in alliance with the Western Church, revived the western Roman Empire in Carolingian form, a new polity with its centre of gravity firmly north of the Alps rather than on the Mediterranean (fig. 14-3).3
For these reasons, consolidation of the Carolingian dynasty in the late eighth century can be said to mark the end of antiquity in the West and therefore the start of a new era. This new post-antique period is normally identified as the Middle Ages, but there is a daunting problem with the legacy of this concept. I am not referring to the pejorative connotations which make it a trough of low culture between antiquity and the Renaissance, as these have long been abandoned; it is rather the major break which is still implied between the Middle Ages and the Renaissance. On the contrary, there appear to be few if any grounds for the view that there was a change in the fourteenth and fifteenth centuries of equivalent scope and depth to that marked by the end of antiquity.
FIGURE 14-1 The Roman Empire, c.395. © by Eric Fernie and Chris Kennish.
FIGURE 14-2 The Caliphate of Cordoba, c.750. © by Eric Fernie and Chris Kennish.
FIGURE 14-3 The Carolingian Empire, 814. © by Eric Fernie and Chris Kennish.
The Renaissance makes more sense as part of a single development from the Carolingian era to the present day, which can, eventually, be identified as the culture of Europe.
This conclusion is supported by the evidence of the boundaries of the Carolingian Empire, which dissociate it from the past and associate it with what were to become the nations of Europe. While the prefectures of the Roman Empire provide a basis for the shapes of the future European political units of France, Spain, Italy, and Britain, the land boundaries of the Carolingian Empire nonetheless have no parallels at all with those, internal or external, of the Roman Empire, and even parallels between the boundaries consisting of coastlines are limited. Conversely, the Carolingian boundaries have a fundamental relevance for the future, both in terms of the extent of the states and of the continuity of administrative structures. In 843 Charlemagne’s empire was divided into three parts (fig. 14-4). Of these, the western and eastern kingdoms provided the basis for France and Germany respectively, while the kingdom between, Lotharingia, contained the areas of the smaller states, provinces, duchies, and counties which subsequently became the Netherlands, Brabant, Lorraine, Alsace, Burgundy, Lombardy, Tuscany, and so on.4
If the Carolingian state can be said to represent the political aspect of the formation of European culture, changes in its economic life were no less significant. It is difficult to exaggerate the poverty into which this part of the world had fallen, with the long decline of the western Roman economy finally reaching a nadir in the eighth century. Then, in the ninth century, an economic revival began, and, though it was curtailed almost immediately by the invasions of the Vikings and Hungarians, it resumed in the tenth century to grow almost unbroken until the thirteenth. The causes of this revival are difficult to identify, but technological innovation must have played a major role. Indeed, the five centuries from the ninth to the thirteenth deserve to be known as the first industrial revolution, as more machines and techniques were adopted or invented in the West in this period than at any time before the Industrial Revolution itself. Among many other things, the ninth and tenth centuries saw watermills being exploited on a large scale for the first time, the heavy-wheeled plough made its appearance, and the collar harness increased the pulling power of horses fivefold, aiding not only agriculture but also transport and the mining industry.5
FIGURE 14-4 The Carolingian Empire, 843. © Eric Fernie and Chris Kennish.
Along with these signs of a major break in politics and the economy, there is the evidence of the architecture. In the Carolingian period the articulation of forms and volumes which is a feature of some monuments of the Roman period takes a major step toward its consistent use, new features like the outer crypt, the westwork and the crossing provided the basis for some of the most characteristic developments of the tenth, eleventh, and twelfth centuries, and (in one of the most marked changes of any date to the Western architectural vocabulary), the silhouette of the building is for the first time treated as a feature of the design.
In addition to this evidence from the buildings, the building industry itself underwent a revolution. It has been estimated that “quarrying was by far the most important mining industry in Europe, more important possibly than all the others combined. The mining of stone in the Middle Ages can only be compared to the mining of coal in the nineteenth century and the drilling for oil in the twentieth.”6 In the centuries of greatest economic weakness, from the fifth to the eighth, stone buildings were largely restricted to the patronage of the ruling elite, yet by the eleventh century almost every parish church was built of stone, while dwellings and shops soon came to be constructed of stone as well as wood. This huge expansion in the provision of cut stone suggests that the industry may have seen an invention as influential as those in agriculture and manufacture, such as an improved metal for the cutting edges of chisels, just as between the ninth century and the eleventh wooden hoes and pitchforks were replaced by their iron equivalents. While high-quality masonry was produced throughout the poorer centuries, it was largely restricted to the dressing of corners, whereas once the working of stone had been established on a new scale it became easier to build whole walls of a regular character. Features like pilasters would have underlined the regularity by subdividing and paralleling the wall, with their corners expressing the exactness of the cutting. In other words the masons’ concentration on clarity and articulation, the chief characteristics of the Romanesque style, could have been strengthened if not caused by the opportunities offered by improvements in quarrying in the Carolingian economic revolution.
This theory of the dependence of the Romanesque on a masonry and hence Roman tradition differs from that proposed by Walter Horn, according to which Romanesque articulation derived from the bays of the tradition of building in wood. Such a source is certainly plausible, as many more structures were built of wood than stone at the time, and every stone building in addition required large amounts of carpentry, for construction, roofing, and fittings. The weakness of Horn’s theory lies in the fact that, apart from the bays, which are available in Roman masonry buildings in any case, there is little or nothing in the Romanesque masonry repertoire which can be derived from wooden construction, which has, for example, no clear equivalent for a feature as basic as the pilaster.7
If this hypothesis is sustainable, then the occurrence of the style in the other visual arts would be dependent on the changes in architecture. Architectural articulation would have made itself felt first in stone sculpture on the buildings, in the designing of capitals, tympana, jambs, and so on, as frameworks for foliage, figures, and scenes.8 The painting adorning that sculpture could then have taken on the same characteristics, and finally manuscript illumination, metalwork, and ivory carving, a sequence supported by the fact that, on the whole, the more precious the material the less Romanesque the style or the later its development. There is, however, a major problem with this sequence: while the link between building and sculpture is direct, that between sculpture and painting is not. It is not at all clear that those responsible for painting books would have appropriated styles from those who painted masonry, because of the greater delicacy of the work in the scriptorium and because the illuminators worked with scribes who were literate or were literate themselves, with all the status which that implies. That the production of manuscripts needs to be considered in a different context is also indicated by the history of lettering, in that the Carolingian minuscule, beautifully articulated in the Romanesque manner, was invented already in the early ninth century and used almost universally in the West up to the rise of black letter in the twelfth. While this complicates the question of the origins of the Romanesque style in different media, it is also, of course, another example of consistency in an aspect of the visual arts in the period c.800 to c.1200.
To sum up, if the Carolingian period marks the start of a new culture, not only in politics and economics but also in architecture, providing a historical context for the origins of the Romanesque style, it has to be asked what utility is served by maintaining a major hiatus between styles called Carolingian and Romanesque. If this and similar breaks are ignored, then the differences which have led to widely varied starting dates for the Romanesque cease to be contentious and take their place in a continuum of stages in the development of a style. The case is supported in many apparently minor ways, by, for example, the number of books on Romanesque subjects which use or mention the date brackets 800 and 1200, or by the words of Robert de Lasteyrie, writing in 1929: “Si donc je respecte l’usage qui est de faire commencer l’époque romane au XIe siècle, je prie mes lecteurs de bien retenir qu’une foule de détails propres à l’art roman se rencontrent déjà au IXe et au Xe siècles.”9
Political Units and Stylistic Subdivisions
This discussion of definitions provides a basis for attempting an overview of the areas and periods into which the Romanesque architecture of c.800 to c.1200 can be divided.10 In any attempt of this kind, it needs to be borne in mind that, while one of the central tasks of anyone studying the Romanesque period is to understand what is happening in terms of the political units of the time, most of the literature is ordered almost exclusively by modern countries and even on occasion with nationalistic motives.11
The Carolingian era begins with a period of political unity in the late eighth and early ninth centuries. Although many important monuments, including revivals of Early Christian basilicas, emulate buildings of the Roman period which avoid articulation, the period also sees the laying of the groundwork of the new style in all major aspects of its treatment of form and use of features. Thus the palace chapel built by Charlemagne at Aachen between 786 and 805 is traditional in the sense that it is based on Roman centralized structures such as the sixth-century buildings of SS Sergius and Bacchus in Constantinople or San Vitale in Ravenna. It is also, however, possible to describe it in strictly Romanesque terms, as the clarity of its spatial organization and the differentiation and articulation of its parts stand in sharp contrast to the billowing ambiguous spaces and filigree capital carving of the buildings in Constantinople and Ravenna. The late eighth-century gatehouse at Lorsch provides a similar example of very early Romanesque design.
The St Gall Plan, of the 820s, provides more varied evidence of breaks with the past, in at least four ways. The first is the relationship between the church, the cloister, and the monastic buildings. This is a masterpiece of articulation, with the cloister fitted into the angle between the nave and the transept and the ranges for the main functions set out around the sides of the square. While the ideal of order which lies behind the Plan is of course Roman, the arrangement of the three parts has no known precedents and conversely became the standard formula for monastic design in the Romanesque period and after. The second indication is the placing of the north and south passages of the crypt outside the main structure. This is a simple version of the outer crypt which in its fullest form is almost a separate building beyond the sanctuary and which provided the basis for the designing of eastern arms of churches from the tenth century on. The third is the representation of a crossing at the intersection between the axis of the transept on the one hand and that of the east arm and the nave on the other. The crossing is common in centralized churches from the earliest dates, but is new in the Carolingian period in unambiguously basilican churches. The fourth is the west end of the church, where there are two towers and a curved colonnade, an arrangement which creates a greater focus of importance than a façade which simply represents the cross section of the nave, and which again relates to more complex future developments.12
The period of unity came to an end in 843 with the triple division of the Empire into the East Frankish, Lotharingian, and West Frankish kingdoms. This division set the pattern not only in political terms but also in architecture, producing the main areas of the whole Romanesque period and underlining the appropriateness of beginning the period with the Carolingian dynasty. In the eastern, German, kingdom and subsequently Empire, in many buildings stress was laid on the west end of the church as well as the east. The westwork at Corvey on the Weser, after 870, with its dramatic silhouette, exemplifies the type. Alternation was equally important, as at Bishop Bernward’s church at Hildesheim, of 1010–33, which also has a western apse.13
In the course of the eleventh century in many of the noteworthy buildings built in the Rhineland, classicizing elements were increasingly employed, as in the Emperor Conrad’s cathedral at Speyer of 1030–61, and even more so in Henry IV’s reworking of the building in the 1080s. The chief aim seems to have been to support the legitimacy of the imperial line by using elements not only from the Roman period but also from contemporary buildings in Lombardy. Thus Henry’s Speyer has large groin vaults over the nave reminiscent of those in thermae, giant-order arcades like those of the Roman basilica at Trier, and sculptural decoration on eleventh-century buildings in Lombardy such as St Abbondio in Como. It is not clear whether these last parallels were mistaken for monuments of classical date or if they were simply considered classical because of their provenance. Cologne in the eleventh and early twelfth centuries also saw a great building boom in Romanesque churches. After this period of activity, the German tradition produced one of the clearest and most inventive instances of Romanesque baroque, as with the western apses of the cathedrals of Worms and Mainz.14
While the German Empire became the strongest political force of the time in the West, with an architecture markedly shaped by political considerations, the middle of the three kingdoms, the Lotharingian, had by the tenth century been absorbed into the other two kingdoms. Despite this, it is significant for two reasons, first because it included many of the economically most advanced parts of the Carolingian states, and second because of the substantial degree of overlap between it and what has been called the First Romanesque style (fig. 14-5). The largely passive role previously ascribed to this style has been replaced with a positive view of its contribution. It appears to have originated in Lombardy and the Po valley in the ninth and tenth centuries, developing directly out of late Roman sources, including the arched corbel-tables which were to become its leitmotiv. The style spread from Italy west to Catalonia which made precocious use of it from the middle of the tenth century, while the early eleventh saw the building of churches as ambitious and international, as those at Cardona and Ripoll. The importance of the Byzantine Empire for the First Romanesque is indicated by the similarities between tenth-century Constantinopolitan churches such as the Bodrum Camii and buildings of the late tenth and eleventh centuries on the Adriatic and Ligurian coasts and in Catalonia. The style spread from Italy up the valleys of the Rhône and the Saône, into Burgundy and the alpine regions, to the monastery at Romainmôtier among others, and to the Rhineland. There, First Romanesque features such as the arched corbel-table tend to be superficially applied to buildings of a more “imperial” character, including, for example, St Pantaleon in Cologne, which may be as early as the late tenth century.15
FIGURE 14-5 The extent of the “First Romanesque” style. © by Eric Fernie and Chris Kennish.
By the middle of the eleventh century the First Romanesque style had changed sufficiently, in terms of quality of masonry and use of half-shafts and sculpture, to warrant a new label, what has been called straightforwardly the Second Romanesque. North Italy again appears to have been in the vanguard. Architects in Lombardy and Piedmont contributed innovations in rib-vaulting, Milan, Pavia, and Novara, for example, having a number of buildings of the late eleventh and early twelfth centuries which have close connections in the north at Speyer and Utrecht and distant ones with Norman examples such as Durham. In most other areas of north and central Italy, wooden ceilings were preferred, in churches like San Miniato near Florence, of the mid-eleventh century, and the cathedral of Modena of 1099 on, with very similar elements, and of course most especially in Rome, where the conservative forms of the Early Christian period continued in use into the thirteenth century. Buildings of this kind indicate that, despite the popularity of vaults, wooden ceilings were in no way considered inferior. Southern Italy and Sicily, in keeping with their political history, developed an amalgam of the Byzantine, the Islamic, and the Norman. The Normans introduced northern French work at places like the cathedrals of Aversa and Acerenza, while the major churches of Bari in Apulia illustrate the importance of pilgrimage in the distribution of architectural forms.16
Italian commercial dominance of the Mediterranean, which was almost complete by the late eleventh century, depended entirely on the success of Italian towns, places such as Venice, Bari, Pisa, Genoa, Milan, Piacenza, Parma, and Verona. Venice was especially prominent, with St Marks, begun in 1063, being based on the sixth-century Holy Apostles in Constantinople, a model which may also lie behind San Nazaro, previously the Holy Apostles, in Milan (refurbished after a fire in 1075), and that of Canosa in Apulia (consecrated 1101).17
In the third kingdom, the western, the Carolingian dynasty lasted until 987 when it was succeeded by the Capetian, the first dynasty of the kingdom of France. Despite sharing a common base with the eastern kingdom in the architecture of the Carolingian empire, the architecture of the western kingdom proceeded along very different lines. The buildings indicate a special interest in the popularity of relics, accompanied by changes in regulations governing the use of altars. These led in the course of the tenth and early eleventh centuries to the reworking of the outer crypt into the fully articulated design of the ambulatory with radiating chapels. This was achieved by the simple geometric expedient of using the central point of the main apse for plotting both the curves of the ambulatory walls and the radiating axes of the chapels, as in the early examples of the cathedrals of Clermont-Ferrand, Orléans, and Chartres. In Anjou there is precocious work in Angers itself, at the cathedral of c.1040, with its aisleless nave, and the Ronceray with its complete set of eleventh-century barrel vaults and a vaulted crossing.18 The Normans, fast learners, adopted the architecture of these and other areas along with a fervour for monastic reform.19
While the French crown, centered on Paris and Orléans, theoretically controlled territory as far as and even beyond the Pyrenees, in reality the southern half of the country (broadly the Duchy of Aquitaine, south of the valley of the Loire) was virtually independent as well as being culturally self-confident. This division between north and south is fairly clear except at its eastern end, as the territories occupied by the Burgundians from the sixth century on are probably politically the most complicated region of the whole Romanesque period. This is due to the fact that their name is attached to a duchy, a county, and a kingdom, the first of which was in France and the other two in the Empire, and that the duchy has a large contingent of First Romanesque buildings closely related to those on the other side of the border. This complex position is simplified at least for the art historian by the fact that most of the Romanesque buildings of importance in the region, whether early or late, lie in the modern French province of Burgundy. Noteworthy First Romanesque buildings include St Philibert at Tournus, with an articulated crypt and experimental vaults of the earlier and later eleventh century respectively, and St Bénigne at Dijon, with its gigantic eastern rotunda, of 1001/2–18. The second church at Cluny, begun in the 950s, is important as the source of a plan type which, along with monastic reform, penetrated much of Western Europe, from Farfa near Rome to Bernay in Normandy and Hirsau in the German Empire.20 The late eleventh century in Burgundy saw dramatic changes from First Romanesque formulae, with the introduction of a newly interpreted repertoire of forms from antiquity. One might have expected these developments to derive from parallels in Provence, given the more extensive Roman remains there and the likely Islamic sources of the pointed arch in Sicily and southern Italy. Yet all the evidence places the Provençal monuments later than the Burgundian ones and leaves us to explain Burgundian primacy by the institutional power of the Cluniac order, exemplified by the vast, structurally daring and classically decorated third church at Cluny begun in 1088.
The most idiosyncratic group of buildings in the south is that vaulted with domes. These appear to be a response to the challenge of vaulting buildings with a tradition of wide, aisleless naves. The linking of the domes to the prestige of the Byzantine Empire is, however, suggested by the fact that two of the most striking structures, the cathedral of Cahors and St Front in Périgueux, are very similar respectively to the Hagia Sophia and the Holy Apostles in Constantinople. Finally, there are a number of eleventh-century buildings known for their location on the pilgrimage route to Santiago, based on north French structures like St Martin at Tours (large, wood-roofed with ambulatory and radiating chapels, transept arms with aisles, and galleries all the way round the church), with the addition of vaults from the First Romanesque tradition. They include St Sernin in Toulouse, St Foi at Conques and St Martial in Limoges. The main church buildings of the twelfth century in the Auvergne are of a related type.21
The Spanish kingdoms lay outside the boundaries of the Carolingian kingdoms and their successor states, but they were soon, through the involvement of the Cluniac order, brought into the French ambient. The cathedral at Compostela, for instance, goal of a pilgrimage which was as important as those to Rome and Jerusalem, is basically the same as St Sernin. There are also numerous buildings on a smaller scale along the route from the Pyrenees, the camina francés, for example, at Jaca in Aragon and Frómista in Castile. It used to be assumed that all Romanesque art in northern Spain was French in origin, and there is much to be said for the view, but the cultural power of the Caliphate in Córdoba made the courts of Christian Spain into centers of cultural importance in their own right.22 England in the tenth and early eleventh centuries developed its own version of the Romanesque on the basis of models from Germany, before also being brought into the French sphere by the Norman Conquest of 1066, with the Norman style then being imposed on or imported into Wales, Scotland, and Ireland.23 The styles of the German Empire, with their imposing, classicizing character, were exported to the newly converted countries of Bohemia and Poland from the tenth century on, and, via Bremen, to Norway, Denmark, and Sweden in the course of the eleventh and twelfth centuries, with Lund Cathedral, for instance, being closely modeled on Speyer. Italy exported its version of the style to Hungary and Croatia, in buildings such as the cathedrals of Pécs and Dubrovnik.24
Numerous aspects of the Romanesque style have their own literature, including monastic architecture, iconography, planning and vaults.25 The most important in terms of historiography, however, is secular architecture. This has been less studied than ecclesiastical architecture, probably because the buildings are less accessible and more altered, with a literature which is often difficult and obscure. In what may admittedly be an extreme case, the author of what was until recently one of the few books to cover castles across Europe was, in the words of one reviewer, “an Estonian who wrote in Swedish a book actually published in German, and the final translation into English cannot be trusted to express his mind.” The position has recently changed out of all recognition with, for example, a string of publications which treat castles as designs meeting the social priorities of a warrior caste and not just as steps in an arms race.26
Most if not all studies of Romanesque architecture as a whole understandably begin by concentrating on those parts of Western Europe in which the style is assumed to have had its origins and main development, leaving the study of the areas on the periphery to a series of slots toward the end of each major regional section. Thus the limits of the style are likely to be considered separately in the chapters on Spain, southern Italy, the German Empire, Scandinavia, and so on. Some form of this arrangement is necessary, but a case can be made for examining all the boundaries of the style together, both because this may indicate links between contiguous areas on the periphery and because of what it can reveal about the character of the style itself.
The boundaries of the Romanesque are natural to the north and west and cultural to the south and east, marked respectively by the Arctic, the Atlantic, Islam, and Orthodoxy. Restricting comment to the eastern boundary (because of the importance for the Romanesque of the architecture and politics of the Byzantine empire), while Byzantine culture had a marked effect across the southern parts of the boundary, in the north this was reversed, with Polish Romanesque showing little if any sign of adopting Orthodox elements. On the contrary, Romanesque forms are evident in the twelfth and thirteenth centuries as far east as Novgorod, Vladimir Suzdal, and Yuryev-Polski in Russia. Despite the extensive importance of the Byzantine Empire for the West, the Byzantine buildings which most closely parallel Romanesque examples, those of the fifth century on in Georgia and Armenia, have no obvious connections with their Romanesque counterparts. They are apparently instead similar responses in masonry (as opposed to the brick almost standard elsewhere in the Byzantine Empire) to the Roman past shared with the Romanesque tradition.27
Unlike the Romanesque, with its vague, broad, social, and economic beginnings, the Gothic style has specifically identifiable origins in a particular place and at a particular time, in the area around Paris in the 1130s and 1140s, and probably at Suger’s St Denis. It was developed out of the Romanesque of northern France and Norman England, as is evident from a comparison of, for example, the use of wall passages in the transept arms of St Etienne in Caen (of the 1070s) and Noyon Cathedral (of the 1150s), or the use of shafts and the stress on verticality at the cathedrals of Ely (1080s) and Laon (c.1160).28 From these beginnings the Gothic style was imported into all other parts of Europe, in some cases, as in the German Empire, to a resistance which appears to have been ideological, but by the middle of the thirteenth century the old style had almost everywhere been replaced by the new. In this the Gothic style was aided by numerous factors including the prestige of the French monarchy, though there is disagreement over the extent to which the spread of the Burgundian Romanesque architecture of the Cistercian order was instrumental in its success.29
The Romanesque is often described as an age of monasteries and the Gothic as one of cathedrals. This contrast is accurate in that there was a great deal more monastic reform in the Romanesque period than in the Gothic, and in that the Gothic as a style was formed, at least initially, primarily in cathedrals. Yet the implication that cathedrals were not important in the Romanesque period is seriously misleading. One need only consider the importance of towns in the history of Romanesque architecture in Italy, or the fact that in the German Empire the Romanesque cathedrals of Mainz, Speyer, and Trier have a status and historical importance rivaled by few if any monasteries. The false picture of the preponderance of monasteries in the Romanesque period is due in part to the large numbers of Romanesque cathedrals demolished to make way for their Gothic successors.
If we are justified in associating the “short” Romanesque with the origins, in the Carolingian period, of the culture which became that of Europe, then this new definition is clearly more than an academic refinement. It can be described as the first “European” style because of the number of identifiable architectural links it displays across Western and Central Europe, links explicable by the effects of political and religious power. Thus the belief on the part of the German emperors that they had inherited the mantle of the Roman Empire ensured a constant interchange across the Alps with Italy, while their prestige dispersed the styles of their patronage to Scandinavia and Eastern Europe; pilgrimages to Rome and Jerusalem created opportunities for exchange down the length of Italy, while that to Santiago formed a school of architects and sculptors extending north across the Pyrenees into France and Italy; Norman control of Normandy, England, and Sicily produced an extraordinary series of connections; monastic reform movements were even more wide-ranging, and the bureaucratic requirements of the bishops of Rome entailed a constant flow of dignitaries and emissaries across Western Christendom on even the most trivial of matters. If we are uncertain of the extent to which contemporaries were aware of this Romanesque style, that it was recognized at least by the early fifteenth century, by whatever name, is made graphically clear by the Master of Flémalle’s Betrothal of the Virgin in the Prado. In this, the betrothal of Mary and Joseph, representing the start of the new dispensation, takes place in the portal of a Gothic church which is under construction, while the Old Law is represented by a complete building standing behind, in a style which is clearly Romanesque.
My warm thanks go to Judson Emerick, Kristen Collins, Michael Herren, Alison Langmead, Robert Maxwell, Elisabeth Monroe, Richard Plant, and David Thompson for comments and suggestions (and Richard Plant also for his help with the bibliography), to Chris Kennish for so expertly producing the maps, to Conrad Rudolph for his incisive comments, his editing, and his hospitality, to the British Academy for a research grant, and to the Getty Research Institute for their facilities and support.
1 OED: Romanesque. For Gunn, and for his French contemporary Charles de Gerville, see Bizzarro, Romanesque Architectural Criticism, p. 142. As it is put by de Caumont, Abécédaire, p. 7: “la période de six siècles (du Ve au XIIe) à laquelle je donne le nom de romane.” The oldest use of the term “Romanesque” for an architectural style of which I am aware occurs in Corrozzet, Les Antiquitez, p. 22: “Il y a de present autres excellents bastiments faits à la Romanesque, à la Grecque et à la Moderne…,” but it is not clear that he means anything by it other than “Roman” (quoted in Thompson, Renaissance Paris, p. 11).
2 de Lasteyrie, L’architecture réligieuse, ch. 8; Focillon, Art of the West, ch. 1; Vergnolle, L’art roman, ch. 1.
3 Pirenne, Medieval Cities; McCormick, Origins of the European Economy, pp. 576–7 and passim. There is also a view that the end of antiquity should be placed in the tenth century.
4 Dawson, Making of Europe.
5 Reuter, “Introduction.”
6 Gimpel, Medieval Machine, 69. On quarrying, see Ward-Perkins, “Quarrying in Antiquity”; Gem, “Canterbury”; Parsons, Stone Quarrying; and Vergnolle, “La Pierre de taille.”
7 Horn, “On the Origins” (on wooden Romanesque buildings, see also n.22). For the Roman material see Ward-Perkins, Roman Imperial Architecture; and, for its Byzantine continuation, see Krautheimer, Early Christian and Byzantine Architecture.
8 [On Romanesque sculpture, see chapters 15 and 16 by Hourihane and Maxwell, respectively, in this volume (ed.).]
9 de Lasteyrie, L’architecture réligieuse, pp. 227–8. Examples of books covering 800 to 1200 include Swarzenski, Monuments of Romanesque Art, and Lasko, Ars Sacra, while Kubach and Verbeek, Romanische Kirchen, have Aachen Palace Chapel as the first building in the chronologically arranged illustrations; Conant, Carolingian and Romanesque Architecture, uses both Carolingian and Romanesque in the title, but chapter 1 is entitled “Carolingian Romanesque.” The prefixes “pre-” and “proto-” express this ambivalent attitude to the ninth and tenth centuries. On the definition of Romanesque, see also Trachtenburg, “Suger’s miracles.”
10 In this regard see Bizzarro, Romanesque Architectural Criticism, p. 13 and passim, and the helpful bibliography in Stalley, Early Medieval Architecture, pp. 247–55.
11 For overviews, see Clapham, Romanesque Architecture; Conant, Carolingian and Romanesque Architecture; Kubach, Romanesque Architecture; Stalley, Early Medieval Architecture. The series La Nuit des Temps, begun in 1954, now covers the whole of Europe. For documents see Davis-Weyer, Early Medieval Art, and for bibliography, see Davies, Romanesque Architecture.
12 Krautheimer, “Carolingian Revival”; Horn and Born, Plan of St Gall; Heitz, L’architecture religieuse; McClendon, “Carolingian Art.”
13 Grodecki, L’Architecture ottonienne.
14 Gall, Cathedrals; Oswald et al., Vorromanische; Kubach and Verbeek, Romanische Kirchen; Kubach and Haas, Dom zu Speyer; Kidson, “The Mariakerk at Utrecht.”
15 Puig i Cadafalq, Premier Art Roman; Whitehill, Spanish Romanesque; Armi, “Orders” and “Corbel Table.”
16 Thümmler, “Baukunst”; Kidson, “The Mariakerk at Utrecht”; Bony, “Projet premier.” Willis, “Construction of Vaults.” is still useful. D’Onofrio, “Precisazioni sul deambulatorio.”
17 Thümmler, “Baukunst”; Brucher, Sakrale Baukunst Italiens; Wharton Epstein, “Date and Significance.”
18 Vergnolle, L’Art roman; de Lasteyrie, L’Architecture réligieuse, is still a benchmark. The literature on individual areas and buildings is copious; see, e.g., McNeill and Prigent, Anjou.
19 Baylé, L’Architecture normande.
20 For Tournus, see Henriet, “Saint-Philibert,” and Armi, “Nave of St Philibert”; and for St Bénigne, see Malone, “Rotunda of Sancta Maria.”
21 Borg, Architectural Sculpture; on Cluny III, see Stratford, “Documentary Evidence.” For regions and schools, see Bizzarro, Romanesque Architectural Criticism, pp. 13, 145, 153.
22 Whitehill, Spanish Romanesque; Durliat, L’Art roman en Espagne. Even more than with most issues in the period, the French-Spanish problem involves sculpture.
23 Gem, “L’Architecture pre-romane”; Fernie, Architecture of Norman England; and the Zodiaque volumes for Scotland, and Ireland.
24 See relevant volume in the Nuit des Temps series on Scandinavia, and for wooden buildings, see Bergendahl Hohler, Norwegian Stave Church, including the proof of their dependence on buildings in the stone tradition.
25 For monasteries, see Braunfels, Monasteries. For iconography, see Krautheimer, “An Introduction”; Crossley, “Medieval Architecture”; Bresc-Bautier, “Les Imitations.” For planning, see Wu, Ad Quadratum. See also the extensive literature on parts of buildings, such as Sapin, Avant-Nefs; for vaults, see note 16. [On architectural layout, see chapter 25 by Zenner in this volume (ed.).]
26 Heslop, “Orford Castle”; Mesqui, Châteaux; Coulson, Castles. The comment is on Tuulse, Castles of the Western World, and is from Hohler, “Bibliography,” p. 20.
27 Voronin, Yuryev-Polskoi. For Armenia and Georgia, see Krautheimer, Early Christian and Byzantine Architecture, pp. 338–48.
28 Gall, Gotische Baukunst. [On Gothic architecture, see chapter 18 by Murray in this volume (ed.).]
29 For the Cistercians, see Dimier on France in the Nuits des Temps series; Stalley, Cistercian Monasteries, on Ireland; and Fergusson and Harrison, Rievaulx Abbey, on a major building. On the Gothic question, see Wilson, “Cistercians as ‘Missionaries’.” Thurlby, “Roger of Pont l’Évêque,” on the other hand, argues that it was the Archbishop of York (1154–81) who was chiefly responsible for the introduction of Gothic – at least into the north of England.
C. E. Armi, “The Corbel Table,” Gesta 39 (2000), pp. 89–116.
——, “The nave of St Philibert de Tournus,” Journal of the Society of Architectural Historians, 60/1 (2001) 46–67.
——, “Orders and Continuous Orders in Romanesque Architecture,” Journal of the Society of Architectural Historians 34 (1975), pp. 173–88.
M. Baylé, L’Architecture normande au Moyen Age, 2 vols. (Caen, 1997).
E. Bergendahl Hohler, Norwegian Stave Church Sculpture, 2 vols. (Oslo, 1999).
T. W. Bizzarro, Romanesque Architectural Criticism, a Prehistory (Cambridge, 1992).
J. Bony, “Le Projet premier de Durham voûtement partiel ou voûtement total?” in Urbanisme et architecture études écrites et publiées en honneur de Pierre Lavedan (Paris, 1954), pp. 41–9.
A. Borg, Architectural Sculpture in Romanesque Provence (Oxford, 1972).
W. Braunfels, Monasteries of Western Europe: The Architecture of the Orders (London, 1972).
G. Bresc-Bautier, “Les Imitations du St-Sépulchre de Jerusalem (IXe–XVe siecles). Archéologie d”une dévotion,” Revue d’histoire de la spiritualité 50 (1974), pp. 319–42.
G. Brucher, Die Sakrale Baukunst Italiens im 11. und 12. Jahrhundert (Cologne, 1987).
Arcisse de Caumont, Abécédaire ou rudiments d’archéologie (Paris, 1851).
A. W. Clapham, Romanesque Architecture in Western Europe (Oxford, 1936).
K. Conant, Carolingian and Romanesque Architecture 800 to 1200 (Harmondsworth, 1974).
Gilles Corrozzet, Les Antiquitez, Croniques et Singularitez de Paris (Paris, 1586).
C. Coulson, Castles in Medieval Society (Oxford, 2002).
P. Crossley, “Medieval Architecture and Meaning: The Limits of Iconography,” Burlington Magazine 130 (1988), pp. 116–21.
M. Davies, Romanesque Architecture: A Bibliography (New York, 1993).
C. Davis-Weyer, ed., Early Medieval Art, 300–1150 (Englewood Cliffs, 1971).
C. Dawson, The Making of Europe (1932)(New York, 1956).
M. Durliat, L’Art roman en Espagne (Paris, c.1962).
——, “La Catalogne et le première art roman,” Bulletin Monumental 147 (1989), pp. 209–38.
P. Fergusson and S. Harrison, Rievaulx Abbey: Community, Architecture, Memory (Yale, 1999).
E. Fernie, The Architecture of Norman England (Oxford, 2000).
H. Focillon, The Art of the West, I (1938; repr. London, 1963).
E. Gall, Cathedrals and Abbey Churches of the Rhine (London, 1963).
——, Gotische Baukunst in Frankreich und Deutschland, I (Brunswick, 1955).
R. Gem, “L’Architecture pre-romane et romane en Angleterre,” Bulletin Monumental 142 (1984), pp. 233–72.
——, “Canterbury and the Cushion Capital. A Commentary on Passages from Goscelin’s De Miraculi Sancti Augustini,” Romanesque and Gothic, Essays for George Zarnecki, 2 vols., ed. N. Stratford (Woodbridge, 1987), pp. 83–102.
J. Gimpel, The Medieval Machine (London, 1979).
L. Grodecki, L’Architecture ottonienne (Paris, 1958).
C. Heitz, L’Architecture religieuse carolingienne (Paris, 1980).
J. Henriet, “Saint-Philibert de Tournus,” Bulletin Monumental 148 (1990), pp. 229–316, and 150 (1992), pp. 101–64.
T. A. Heslop, “Orford Castle, Nostalgia and Sophisticated Living,” Architectural History 34 (1991), pp. 36–58.
C. Hohler, “A Bibliography for the Study of Romanesque and Early Gothic Art up to c.1250,” typescript, Courtauld Institute of Art, 1962.
W. Horn, “On the Origins of the Mediaeval Bay System,” Journal of the Society of Architectural Historians XVII (1958), pp. 2–23.
W. Horn and E. Born, The Plan of St Gall, 3 vols. (Berkeley, 1979).
P. Kidson, “The Mariakerk at Utrecht, Speyer, and Italy,” in E. de Bièvre, ed., Utrecht Britain and the Continent, Archaeology, Art and Architecture, British Archaeological Association Conference Transactions, 18 (Leeds, 1996), pp. 123–36.
R. Krautheimer, “The Carolingian Revival of Early Christian architecture,” Art Bulletin 24 (1942), pp. 1–38.
——, Early Christian and Byzantine Architecture, 4th edn. (with S. Ćurčić) (Yale, 1986).
——, An Introduction to an “Iconography of Medieval Architecture,” Journal of the Warburg and Courtauld Institutes 5 (1942), pp. 1–33.
H. Kubach, Romanesque Architecture (1979; repr. London, 1988).
H. Kubach and W. Haas, Der Dom zu Speyer, 3 vols. (Munich, 1972).
H. Kubach and A. Verbeek, Romanische Kirchen an Rhein und Maas (Neuss, 1971).
P. Lasko, Ars Sacra 800–1200 (Harmondsworth, 1972).
R. de Lasteyrie, L’Architecture réligieuse en France á l’époque romane (1911; repr. Paris, 1929).
C. B. McClendon, “Carolingian Art, II, Architecture,” Dictionary of Art (London, 1996).
M. McCormick, Origins of the European Economy: Communications and Commerce AD 300–900 (Cambridge, 2001).
J. McNeill and D. Prigent, eds., Anjou: Medieval Art, Architecture and Archaeology, British Archaeological Association Conference Transactions, 126 (Leeds, 2003).
C. Malone, “The Rotunda of Sancta Maria in Dijon as ‘Ostwerk’,” Speculum 75 (2000), pp. 285–317.
J. Mesqui, Châteaux et enceintes de la France médiévale, 2 vols. (vol. 1: Paris, 1991).
M. d’Onofrio, “Precisazioni sul deambulatorio della cattedrale di Aversa,” Arte Medievale, 2nd ser., VII (1993), pp. 65–79.
F. Oswald, L. Schaefer, and H. Sennhauser, Vorromanische Kirchenbauten: Katalog der Denkmäler bis zum Ausgang der Ottonen, 3 vols. (Munich, 1966–90).
D. Parsons, Stone Quarrying and Building in England AD 43–1525 (Chichester, 1990).
H. Pirenne, Medieval Cities: Their Origins and the Revival of Trade (Garden City, 1925; repr. New York, 1956).
A. K. Porter, Lombard Architecture (1917; repr. New York, 1967).
J. Puig i Cadafalq, Le Premier Art roman (Paris, 1929).
T. Reuter, “Introduction: Reading the Tenth Century,” in Reuter, ed., The New Cambridge Medieval History, II, c.900–c.1024 (Cambridge, 1999), pp. 1–26.
C. Sapin, ed., Avant-nefs et espaces d’accueil dans l’église entre le IVe et le XIIe siècle (Auxerre, 2002).
R. Stalley, The Cistercian Monasteries of Ireland (London and Yale, 1987).
——, Early Medieval Architecture (Oxford, 1999).
N. Stratford, “The Documentary Evidence for the Building of Cluny III” and “La Sculpture des partis orientales de l’église Cluny III d’après les fouilles de K. J. Conant,” in Studies in Burgundian Romanesque Sculpture, 2 vols. (London, 1998), pp. 41–59 and 61–3.
H. Swarzenski, Monuments of Romanesque Art (London, 1954).
D. Thompson, Renaissance Paris (London, 1984).
H. Thümmler, “Die Baukunst des 11. Jahrhunderts in Italien,” Römisches Jahrbuch für Kunstgeschichte 3 (1939), pp. 141–224.
——, “Carolingian Art,” Encyclopaedia of World Art, vol. III (New York, London and Toronto, 1960), pp. 82–103.
M. Thurlby, “Roger of Pont l’Évêque, Archbishop of York (1154–81) and French Sources for the Beginnings of Gothic Architecture in Northern Britain,” in John Mitchell, ed., England and the Continent in the Middle Ages: Studies in Memory of Andrew Martindale (Stamford, 2000), pp. 35–47.
M. Trachtenburg, “Suger’s miracles, Branner’s Bourges: Reflections on ‘Gothic Architecture’ as Medieval Modernism,” Gesta 39 (2000), pp. 183–205.
A. Tuulse, Castles of the Western World (Vienna and Munich, 1958; repr. Dover, 2002).
E. Vergnolle, L’Art roman en France (Paris, 1994).
——, “La Pierre de taille dans l’architecture religieuse de la première moitié du onzième siècle,” Bulletin Monumental 154 (1996), pp. 229–34.
N. Voronin, Yuryev-Polskoi (Moscow, 1971).
J. B. Ward-Perkins, “Quarrying in Antiquity: Technology, Tradition and Social Change,” Proceedings of the British Academy 57 (1971), pp. 137–58.
——, Roman Imperial Architecture (Harmondsworth, 1981).
A. Wharton Epstein, “The Date and Significance of the Cathedral of Canosa in Apulia, South Italy,” Dumbarton Oaks Papers 37 (1983), pp. 79–89.
W. Whitehill, Spanish Romanesque Architecture of the 11th Century (Oxford, 1941).
R. Willis, “On the Construction of Vaults in the Middle Ages,” Architectural History of Some English Cathedrals: A Collection of Papers… 1842–63 (Chicheley, 1973), appendix.
C. Wilson, “The Cistercians as ‘Missionaries of Gothic’,” in C. Norton and D. Park, eds., Cistercian Art and Architecture in the British Isles (Cambridge, 1986), pp. 88–116.
N. Wu, ed., Ad Quadratum: The Practical Application of Geometry in Medieval Architecture (Aldershot, 2002).
Encyclopaedias and series
Encyclopaedia of World Art (New York, London and Toronto, 1960).
Macmillan Dictionary of Art (London, 1996).
La Nuit des Temps (La-Pierre-qui-Vire: Zodiaque, from 1954).