Current Medieval Research II
The aim is to identify, from every example of a single repeatable type of decoration - foliage - the more innovative carvers of northern France between 1100 and 1170, and by analysing the evolution of each man's personal style, to date the key buildings. This will provide a coherent foundation for a much-needed chronology of Early Gothic. Current dating is far from uniform, being individual reflections on style not based on any common theme, and often with strong nationalistic overtones. Lacking a firm chronology, histories are inconsistent and contradictory. The outcome of this study should resolve this impasse.
The traditional chronologies of the past century or so are based on two major assumptions
Dating churches 1100-1170 in the Paris Basin
1. that minor works were derived from the major ones such as Saint-Denis and Chartres and
2. that figure sculpture followed an even progression from 'archaic' to 'realistic'. Though this has been questioned in studies on individual buildings [Caviness on Braine (1984), Clark on Notre-Dame (1986), Polk on le Mans (1985), Vermand on Senlis (1987) and myself on Longpont and Soissons (1984a, 1989)], no one has yet attempted a rigorous re-evaluation of Early Gothic architecture.
There being little documentary solidity, most current dating is based on a complex interweaving of stylistic associations. Take one example by Sauerlander of this common approach (1963). He dated the nave porches at Notre-Dame in Châlons in the 1170s from stylistic similarities with the carving on the Porte de Valois. This was dated in the 1170s because it seemed less advanced than the Senlis portal that Salet had placed in the 1190s. Having Châlons in the 1170s meant that the carvings on the ground floor of Saint-Germain-des-Près had to be placed only a couple of years before the finished building was dedicated in 1163, though they could have been twenty or thirty years earlier. For consistency, the Dijon portal was pushed forward into the 1160s in spite of Quarré's well-argued suggestion for 1147± (1957). One can see how 'shifty' and afloat the whole situation is. Wherever one looks, buildings are being dated with an idiosyncratic randomness that is far too soft for comfort. Yet Sauerlander's argument reflects both the importance everyone attaches to finding a chronology and the great efforts that have been made spite of the lack of any common thread.
Consequently, the range of dates proposed by scholars can be very wide. The Bourges lateral porches that were dated to 1172 by Branner (1962) and to the 1150s by Stoddard (1987). A date in the 1130s has never been considered from the assumption that Bourges had to be influenced by Chartres. There is in fact no evidence that Chartres ordained a prototype, other than the undoubted importance of its school of clerical scholars. The concept could have been "in the air", as it were, and discussed at various gatherings well before the first was carved. Even if the ideas had come from Chartres, that cathedral may have had to wait until the fire provided the opportunity for it to be cast in stone.
It is usual to state that the dozen great sculptured porches of the Paris Basin were carved between the mid-1130s (Saint Denis) and the 1190s (Senlis), whereas this project will show they were carved from the later 1120s to the later 1140s, and that this was followed by a hiatus in portal sculpture until those of Braine, Laon and Chartres in the 1180s and 1190s. I find it interesting that this shorter period coincides almost exactly with the presence of Eleanor of Aquitaine in the capital, and though I would not suggest she was responsible, I would follow Greenhill (1977) in arguing that at the least the queen had brought a richer and more decorative southern influence to the austerity of the court in Paris.
Again, Morienval has been credited with the first rib vaults in the Paris area whereas the capitals show precedence should given to a dozen other buildings, all of them small and unpublicised, such as Acy-en-Multien, Athis, Bonnesvalyn, Bourgogne and so on. Notre-Dame in Paris was for years considered the first to start flying buttresses in the 1180s, whereas the foliate carving shows that this vitally important innovation should be dated thirty or more years earlier (James, 1992).
These parameters of traditional scholarship offer no effective means of dispute for their varied and often inconsistent conclusions, as they are all too often affairs of conviction. Some other approach is needed with new evidence and a new perspective that is not caught up with earlier previous opinions.
Over the past 35 years I have come to the conclusion that the manner of carving foliage can provide this evidence, and do so in a coherent and verifiable manner. This is the one common and constant element found everywhere. It offers a greater measure of certainty than historians have had before, with a more secure foundation than the rather Byzantinate arguments used to date.
I have tested the project in a number of pilot studies: with two small exceptions they could not be published as the hundreds of illustrations each required was too much for the limited space in the average journal (James 1990, 1998a). What is needed is a large illustrated monograph.
This has begun. Part A of what is projected to be a nine-volume series, The Creation of Gothic Architecture - an Illustrated Thesaurus: The Ark of God, was published in September 2002. It concentrates on the foliate capitals from 1170 to 1250, being the eighty years after the period of this application, and contains 10,500 photographs of the capitals from 450 buildings.
Establishing the dates for these buildings has been relatively straightforward, if arduous, as the documented churches showed there was a perceptible and regular evolution of foliage from a more spring-like naturalism in the 1170s to high summer and increased realism by the 1240s. As there were enough dateable examples, I was able to define the sculpting mode of each decade, and thus to use the style or fashion of the leaves to point the date.
Part A also includes all 300 contemporary texts relevant to construction, sourced to the originals in all cases, and translated. Assembling all these in the one place has not been attempted before. It has been a gruelling process in itself that has taken many years of work, and is essential as the documents provide the only valid foundation for a chronology.
I established there was a major transformation in foliate carving in the greater Seine valley that was precisely dateable between 1169/1170 and 1180/81. This did not occur in other regions for another twenty years. During this decade the formal and abstract style of foliage of the Romanesque period disappeared totally to be replaced by a more natural style of observation. Thus all buildings without natural-style foliage are before 1170, and those without formal capitals are after 1180. The watershed of the 1170s formed the natural lower boundary for Part A of the The Creation of Gothic Architecture - an Illustrated Thesaurus: The Ark of God.
Part B will contain some 13,000 photographs for the sixty years prior to 1170, along with the thirty or so identified carvers and the chronology that will emerge from this project. Part C, on the evolution of rib vaults from 1090 to 1145, will follow immediately. These two parts are interrelated, as the dating provided by the capitals needs to be consistent with the dating from the rib vault analysis. This includes my current work on Durham cathedral where the first documented rib vaults were built.
It is my intention in Part D, using the techniques of toichology (or, above-ground archaeology) already utilised at Soissons, Chartres, Durham, and so on, to isolate some of the more creative individuals on the major buildings who brought us the Gothic style.
Lastly, in Part E, I wish to reassess the story of how Gothic architecture and carving evolved from Romanesque. It will centre on the Paris Basin and the enormously creative labours of the French, and will include the impact on that region of the Normans and the English before innovation ceased with the civil war, and of the Italians immediately after the great earthquake of 1117.
There is a natural progression in the completion of these volumes. It is hoped that the presentation and the conclusions will support each other to create a holistic view of the period. This will be a new form of history for the middle ages, written primarily through the work of individuals rather than style.
At this point, world scholars will have available a comprehensive resource for the study of the Early Gothic period, such as has never been attempted before. An analytic thesaurus of the hundreds of more important buildings and the thousands of carved capitals is much needed and long overdue. There will, of course, be questions and amendments as time passes, but Australia would be remembered for having made a significant contribution to understanding one of the most important epochs in European history.
There are two key outcomes from this project: to identify individual carvers, and through analysing changes in their workmanship over time to establish a firm chronology for the buildings they worked on. It is anticipated that it will be possible to identify 30 individuals and date the major campaigns of over 350 churches between 1120 and 1170. Considering that there are firm dates for parts of only five of these churches, the potential value of this study would be enormous.
Every scholar who has become immersed in foliate capitals has become aware of the similar designs and cutting styles found in many and often widely-separated places. Usually the similarities are referred to as belonging to un atelier, or un école. This approach is epitomised in the work of the older scholars: Aubert, Grodecki, Lefèvre-Pontalis and Salet, though all four have joined in the attempts to determine which parts of Amiens and Reims were built by the individuals named in their labyrinths.
There is a reluctance among French and some American scholars to ascribe works to individuals in a period that is generally defined as anonymous, though from Harvey and Colvin onwards the British have argued otherwise (1950, 1950). The documentary evidence set out by them, and by Knoop and Jones (1933), Murray (1980, 1982) and Shelby (1977), shows that skilled individuals were honoured and well-paid. In addition many masters put their names to their work: Giselbertus of Autun, Gilabertus of Toulouse, Isembardus in Bernay, Renco in Tournus, Robertus at Chartres, Umbertus at Saint-Benoît, Wiligelmo in Italy and so on. This does show there is no fundamental reason for not being able to distinguish individuals, were one to find them. However, one has to sympathise with the traditional reluctance to personalise the middle ages, for the amount of material needed for any rigorous study is enormous - I would cite the 50,000 pictures needed to encompass the one small region of France covered by this project.
In studies of figure sculpture, there has been a lively interest for over a century in sourcing the origins of the great portals, especially Saint-Denis and Chartres. Over the years Armi (1994), James (1979-81), Lyman (1978), Priest (1923), Stoddard (1987), Vergnolle (1985), Vöge (1894) and Watson (1986) have all attempted to identify individual hands on both the large and small figures. In 1961 Grivot and Zarnecki made a study of a single Burgundian sculptor, Gislebertus, and not only identified his work in a number of churches, but assessed the way in which his style changed over the years.
These studies have shown how difficult it is to separate individuals working on large figure sculpture, mainly from the collective nature of the crews in which a master may have set out the design and possibly finished the head, while many others would have blocked in, carved hands and vestments, and worked on the finishing - a practice that is well-documented in Italian painting. During more than a century of analysis there has been little agreement on who carved what.
On the other hand, the carving of foliage on capitals is a different matter altogether. Most were quick to produce. Small ones could be carved in a day, and in less than week for the average. There is no sense that the masons thought foliage was ideologically important. Capitals were casual works, knocked out by the thousands during a lifetime, with more feeling than thought (Baylé 1990, Johnson 1989). They are, in a sense, products that combine the mood or fashion of the moment with personal idiosyncrasy. Thus each is a personal statement that is much more likely to reflect the mood and artistic concepts of the moment than more monumental works.
Being personal statements we do not have to take into account external factors of which we may know little, such as exercises in authority or local political issues. The foliage can be assessed in its own right and on its own terms. There is an ongoing interest in Romanesque foliate capitals with attempts at identifying the sculptors, especially in Normandy and the centre of France: Armi (1983), Baylé (1988), Cabanot (1987), Camus (1992), Mallet (1984), Schmidt (1981) and Vergnolle (1985). Their method has been to group capitals with similar layouts (seldom more than five or six), either to show there was a common template or that they were the work of one man. Only Armi has studied the foliage with attention to the handwork, mode of finish and design preferences that could locate a singular carver. More importantly, no comparable studies have been made of pre-1170 capitals from the Paris Basin - yet this is just where came the innovations that created the Gothic style.
This projects starts with an enormous advantage: we already possess the necessary photographic database to work at a depth that has not been available to my scholarly predecessors. As they lacked such a resource, they naturally had to concentrate on the more significant or well-known sites. From our photographic database we can provide a wide selection of each man's ouvre: there are a hundred or more examples from each one. This is hugely more than those published by any other scholar. It provides a solid and I believe convincing basis for identification. We can then display every example of a particular design, and as none will be left out assess which are originals and which copies or ones designed by the master by finished by another. This will bring consistency to the project.
The pilot studies and the work of the past thirty years have shown that from such large numbers a man's artistic and technical evolution may be studied as it evolves in time, to mature and to change. There are recognisable progressions in the evolution of a carver's mastery and growing maturity that provide a clear picture of his stylistic evolution. The changes will allow us to divide a working life into phases. Some may be dateable, and those that are not may usually be given an approximate date relative to the others. I personally believe that in this way we will be able to analyse over thirty identifiable masters.
It will then be possible to lock the phases of many masters together to create an overall chronology (as to be explained in E4). This time-line is the ultimate aim of the project. Through it we will be able to date some hundreds of Early Gothic monuments from a common element that exists in nearly every building, running like continuous threads through the cloth of medieval construction.
The process of identification is painstaking and involves the comparison of thousands of photographs and the careful listing of types and details. With careful observation, copies and team-styles may be recognised by examining the details, artistry, surface finish etc, and it will also be possible to distinguish foliage designed and executed by a carver from that worked on by his assistants. For example, the master referenced as Palmier carved over 150 capitals with palm tree motifs. Of these, only ten with the palm-tree motif were carved in a different manner - and all these were from later years when one might expect a great sculptor to have assistants to help him. The study went on to show that the architectural details in these later buildings were very similar, which suggests this carver may, at a later stage in his life, have become the master mason as well - not unlike Giotto in a later age.
Buildings in the early middle ages were constructed in a discontinuous manner. There were many campaigns separated by periods of quiet, and it was common to employ different building crews for each campaign (James 1993). If carvers and their workshops were fixed it would be impossible to order someone's work in relation to another's - a problem recognised elsewhere, as in Tuscan Romanesque (Sheppard 1959). However the pilot studies showed that, in this region carvers were not permanently attached to any construction crew, but were independent and itinerant. It is independence that make this project viable. The short-term presence of identifiable men on many sites provides a connective tissue across the landscape of France that anchors their work in time. In addition, this study will provide an exceptional insight into the working manners, the fashions and artistic attitudes of the medieval mind.
One widely-read scholar, Caroline Bruzelius (1987), has stated the need for such an investigation, for "The study of Gothic architecture is still in its infancy. Hampered by the almost complete absence of documents and lacking secure evidence for the dating of campaigns or the beginning of construction, the historian has had to rely on a rather unsatisfactory patchwork . and even this is treacherous."
As the pilot studies have shown, the outcome will provide a rigorously consistent dating procedure for the foliage that will, from propinquity, date both the major sculptured portals and the buildings they sit in. Since the Paris Basin was the source of the transformation from Romanesque to Gothic, nothing could be more important for our understanding of the innovations that created this eminent and radical style than a soundly-based chronology.
The approach is novel and innovative and provides a new methodology for the study of medieval architecture and sculpture. The capitals from the period 1120-1170 (in contrast to later foliage) are extremely personal in arrangement. There are many types and morphologies. In pilot studies I have successfully applied the methods used by Renaissance scholars to identify the work of individual craftsmen. Such studies require too many photographs for the available journals, and awaits a large publication with tens of thousands of pictures: the three-volume continuation of The Ark of God.
The preliminary stage for this project involved visiting every church in the Paris Basin to define those which contained anything from this period - this was needed as no survey had been attempted before (James 1984b). The survey located over 1,400 buildings in which some part had been built in this period, whereas publications had listed only one third that number. Nearly every church had to be visited more than once to collect detailed photographs of the capitals and to analyse the construction campaigns. This uniquely comprehensive collection is the foundation of the project. It provides the universality required for the process of identification.
The archive of every foliate capital in the Paris Basin that may have been carved in over 700 buildings between 1100 and 1250 now contains over 50,000 photos in 80x60 mm format, labelled by site and location. It is unique in size and speciality. No other archive, in France, England or America, has its equal for this project. I have spent thirty years building this collection, often under difficult circumstances as many capitals are inaccessible without scaffolding, or lie in spaces too dark for normal photography. Many are found in buildings known only to those who live nearby. This stage of the project is now virtually complete.
I have experimented with many approaches, and found that the only effective technique is to identify the individuals, to divide their work into phases marked by increased skill and maturity, and then to integrate these identified masters into a spreadsheet format to determine the chronology. This, then, is the program for the next three years. It is planned as follows:
1. examine all the capitals to isolate thirty or more individual carvers,
2. clearly define the working style of each in words and pictures,
3. divide each person's output into sequential phases and
4. date those which are referred to in the documents.
5. When this is done the lists of work by each individual will be placed alongside one another so that we may estimate the dates of all within the constraints of those few that are documented.
STAGE 1: The process will begin by dividing the capitals into separate designs, such as branched, water-leaf and acanthus. These will be subdivided further into arrangements such as meandering, symmetrical, with people and creatures in them, and so on. These will then be analysed by the cutting techniques, use of drill, pointed or curved foliage, etc. It is simplest to scan the photos with filenames that contain the designation of the master as well as the site and location, so they can be sorted and printed by place and person. Some photos will need digital retouching.
As this process continues we will determine and define those consistent qualities of design and handwork that could be attributed to the work of an individual. In The Template-makers (1989) I used precisely this method to identify a number of architects. One, in particular, could be followed across a number of buildings evolving, step by step, the concepts that led to the traceried window.
On completion our aim will be to illustrate every capital by each master in Part B of The Ark of God. Alongside them will be illustrated all the others from the period. Often this will include close-up details of those features which most clearly define the dossiers. Thus every scholar will have access to all the material we will be using, and allow others to continue the work and make further identifications.
STAGE 2: Each group of capitals will be studied to define the dossier, or working style that bespeaks an individual. This needs to be handled in a rigorous and self-critical manner to avoid blurring the edges of definitions whereby we may lose sight of a particular sculptor or confuse those who worked in similar ways. We know from the pilot studies that this rigour may be maintained as long as it is not hurried.
The scanned photos will be mounted onto 'research pages' for ease of reference, which we expect to recast many times while we firm up on the identities. They will, as they are completed, form the basis for all cross-referencing as the work proceeds.
It would be unlikely that a master would have only one design-type. Sometimes there are unusual ways to finish a leaf or the junction of a branch that can help to broaden definitions. Different designs may be linked where two capitals have been carved onto the same stone, each to its own arrangement. Often a unique leaf-tip will be repeated in very different circumstances, as on a frieze or on an impost, and surrounded by different elements that may then be included in
At best we will be identifying real people, in the fullness and variety of their life, at worst no more than a number of morphological groupings. However, the pilot studies have shown that the best is the more likely, with real people being detected. Indeed, one of the most interesting asides has been the ability to identify a number of masters working around Paris during their apprenticeship, travelling great distances as journeymen, even into Italy and Scotland, and then returning (leaving examples in their wake) to spend the rest of their lives in the Paris Basin. All this adds a convincing verisimilitude to the procedure.
During this stage I will require three months in France to refine the images, take additional pictures and confirm some of the attributions and identifications on site.
STAGE 3: will divide the work of every master into phases. For example, in one of the pilot studies on a master carving palm trees, one phase stood out clearly: at first the lower half of the capital was invariably covered with a row of upstanding leaves while later the lower row was left out leaving much of the surface bare. More subtle changes occurred in the decoration of the branches, the depth of cutting and the under-chiselling that finally enabled the work of this very gifted master working over a twenty-five year period to be divided into eight phases.
It is understood that no transition from one phase to another can be watertight. There will be carry-overs and reversion, there can be jokes and play between members of the team, and any manner of idiosyncratic events that can disturb a neat picture. In the pilot studies I found that the fifth phase did much to sort out the eccentricities.
STAGE 4: Once identified every capital and other foliate carving that can be supported by documents will be dated: there are over 1500 of these in this project. For example, among the eight phases of the man who liked palm trees, his style at Saint-Denis and Chartres is in his second manner, and at Senlis in his fifth. This would indicate that the intermediate phases, the third and fourth, should be dated somewhere between the early 1140s and the early 1150s.
STAGE 5: This is where the most sensitive work is needed. We will now place the phases of each of the carvers in relation to each other. This will be done using a spreadsheet system shown over the page, so we can place all the campaigns of every identified carver in vertical strips ordered by their phases. On it we will mark any known dates. Wherever two or more carvers worked on a site together the horizontal grouping on the chart will be joined, so that if any part of the list moved backwards or forwards in time, all members of that group would all have to move together.
Care is needed to follow the construction sequence in each building. For example, in the narthex of Saint-Denis there are seven levels of capitals. None in a lower level can be placed before the one above, that stands to reason. Thus if there is a change in carving style from one master's work on the western portal compared to another in the upper chapel, we can presume that the modification occurred between 1135±/1140. By treating every distinct building campaign as a separate item we wont mix up the teams when they may have worked at different times, and will be more able to fine-tune the dates.
Step by step as more masters are added and the dossiers become clearer, this procedure will erect a chronological scaffolding that, like tree rings in dendrochronology, should result in a relatively accurate dating for nearly all the more important buildings of the period.
Where there is uncertainty about the dossiers or the phases, the need to maintain the horizontal connections when the masters worked together on the same site would compel revisions or even re-attributions in some cases. In the pilot studies the partial ouvres from eight masters (who together contributed to more than 50 sites over 35 years) demonstrated that the horizontally locked bands and the natural construction sequences could provide such firm controls over the process the dating of some buildings could be determined to the year.
When we expand the procedure to include more masters the dating of many more buildings should become just as precise, and nearly every major building in the area would be included. It is an intricate process that requires time. This extract from one of the charts may give some idea of how the process would operate, with the masters listed in columns, and their phases noted in Roman numerals.
This provides a time-line that is based on one simple form, foliage. The steps in the argument are readily checked, and so may be confirmed or amended by other scholars. It is, in fact, an open and transparent form of analysis that depends only a little on personal bias and mainly on verifiable steps.
The most tangible impact of this research will be to reconsider currently-held views on the transformation of Gothic architecture out of Romanesque:
· The chronology will be the most important contribution. For the first time scholars will have a way to date both architecture and sculpture from one common oft-used feature. By implication this will provide ante quem dates for any wall painting and stained glass contained in these buildings.
· Having a chronology of figurative sculpture independent of the style of the figures themselves will enable scholars to analyse the evolution of sculpture without being caught up in the 'internal' arguments on form. This will be particularly important for the great sculptured portals, which is all that remains of what may have been an enormous amount of work, now lost, in tombs and furniture (both domestic and religious), in reliquaries and so on.
· A chronology for each part of over 400 buildings will allow scholars to identify where and when each of the great inventions were made that transformed solid wall-based structures with small windows and round arches into the Chartres and Reims of the next century with their thin walls, wide windows, pointed and peaked arches, flying buttresses, rib vaults, etc. (James 1993, 1997).
· Up to now it has been considered that inventions occurred in the greater workshops which inspired the smaller. Having a chronology for so many buildings will allow scholars to include the contribution of the smaller rather than having to concentrate only on the larger.
· By identifying individual carvers we will be able to understand gothic history through the individuals rather than the more remote mannerisms of style.
· By dating the carvers and locating where they worked, and of noting which men laboured together, we will gain a deeper insight into the working methods and contractual arrangements of the times.
· This will affect the history of Gothic architecture and sculpture. Some of the changes affect only the details, but some will impact on the very fundamentals of current historical views.
This is intended to be a complete resource and is sorely needed by historians world-wide. By displaying every capital conclusions may be verified. There is also the great benefit of being able to train others in the joys and intricacies of medieval research, in learning how to handle large amounts of data, and in using this novel methodology for assembling data and analysis into coherent and meaningful arrays.
In this project Australia has the possibility of providing something both rare and remarkable, a human history of a period that has, unlike the Renaissance, been usually seen as an anonymous undertaking.
John James bibliography