Carl A. Carlson III, Department of Art
Introduction
While Romanesque churches are characterized by a longitudinal barrel vault over the nave, some unusual examples of vaulting from the period. Like an endangered species, St-Philibert of Tournus stands as a monument to these diverse examples. In place of a longitudinal barrel vault, the nave is covered by a series of transversal barrel vaults (fig. 1 & 2). Not only does this research paper seek to place this phoneme in Western architecture and explain its limited repercussions, but it also addresses its unresolved origin.
During a time when church architecture often followed regional standards, St-Philibert of Tournus provides an interesting example of deviance from the norm. Before the mid 11th century, most existing churches such as Tournus had a flat, timber roof. It was at this time that churches were being built with barrel vaults, and timber roofs were being replaced in preexisting churches. The major complication to adding a barrel vault over a previously vaultless nave, was that walls were often too thin to support the heavy weight associated with stone vaults. To combat the problem, walls were often thickened or buttresses added to support the weight. To our great fortune, the architects at Tournus found an alternative solution.
Features of Tournus’ system of nave vaulting are the following. The design of the transversal vault directs the weight of the vault onto massive columns, a tower at the crossing of the nave and transept, and the huge Carolingian narthex preceding the nave, thus eliminating the need for modifying the walls (fig. 3). Because the nave walls were non-load bearing, it was possible for it’s windows to become quite large compared to traditionally vaulted churches of the same period (fig. 1). The latter proved to be quite important in that at this time there was great demand that churches have more light. Non-load bearing walls and large windows are characteristics credited to Gothic architecture a century later.
Survey
The initial goal of this research project was to determine why this system of nave vaulting was not adapted in more churches while it’s advantages were so great. Using funds received from the BYU Office of Research and Creative Activities, a survey was conducted on 26 Romanesque churches in the region surrounding Tournus. The goal of this survey was to establish whether or not these churches in the region would have been suited for a similar transversal nave vault and if such a vault would have been advantageous.
The survey began with a thorough examination of both St-Philibert of Tournus and Mont-S!-Vincent (the only other church with transversal barrel vaults over the nave), located about 45 miles north-west of Tournus. The two churches were visited and compared to one another in order to build a criteria from which to evaluate possible candidates. While 23 of the 26 churches had cross towers which could have offered buttress support to one end of a transversal vault, only one had the necessary structural strength at the front of the church to support the thrust of a transversal vault.
Discussion
Having preexisting forward structures to the nave and cross towers, Tournus (fig. 4) and Mont-St-Vincent already had the buttress strength needed to support the heavy and complex system of vaulting. In their situation, the incorporation of a transversal vault became an efficient way to provide a stone vault over the nave. Though the system of vaulting was more time consuming and costly than a longitudinal barrel vault, this was offset by the time and money saved from not modifying the walls. When modifying a church to accept a vault of stone, the choice of building a forward structure to the nave as opposed to modifying preexisting walls, would have proved extremely costly and time consuming.
Conclusion
It is thus quickly deduced that for a transversal barrel-vault to be advantageous over a longitudinal barrel-vault, a church must already have preexisting buttress support at it’s cross section and front end in order to justify the added expense and time associated with the construction of a transversal vault. The use of the transversal vault was further diminished by the degree to which each region focused on their own school of architecture. In addition, though superior to the longitudinal vault by means of permitting the walls to have larger windows, the transversal vault was less aesthetic to French ideals. In contrast to the ideals of visual continuity and harmony found in a longitudinal barrel vault, the sheer structure of a transversal vault divides the nave at each bay, creating the effect of a series of individual box-like spaces.
Not only was this effect unappealing to the architect’s eye, but it also took away from harmonics. It is certain that other churches may had been ideally suited for this type of vault, and even that their architects may have been aware of Tournus. However, the inquietude that such a heavy vault would prove stable could have been enough to keep an architect or builder attached to more traditional and sure methods of covering a nave. More information may lie in the unpronounced origin of the vault.
Additional Discussion
During the 1920s and early 1930s over half a dozen scholars and writers wrote about the unusual vault and each gave a possible theory concerning its origin. Though several of the scholars briefly commented on one or two of the other theories, none were refuted. Only one received any real doubt. To the present day, few developments have been made and very little has been produced on the subject.
Many suggestions have been made as to where the architects of To urn us found inspiration for the transversal vault. Roman aqueducts, theaters, triumphal arches and draw bridges were to be found close by. On the outer reaches of their known world, Persian palaces, such as Tag-Iwan, provided similar vaults. In Spain, yet another interesting structure reproduces exactly the Persian mode of vaulting. The head abbot who oversaw the construction ofTournus’ vault (Pierre 1st), may have seen this building during a journey. Or perhaps the transversal vault was a shear product of the ingenuity of it’s designers. To properly ascertain a single origin of the vault is a project that will require much more study and research, the results of which could shed new light on the history of architecture.
References
- Cure, L’Abbe. Saint Philibert De Tournus. Paris, 1903.
- Dickson, Marcel et Christiane. Les Eglises Romanes de L’ Ancien Diocese de Chalon. Macon, 1935.
- Enlart, C. Bourgogne Romane.
- Henriet, Jacques. “Saint Philibert de Tournus: L’Oeuvre du Second Maitre, La Galilee et la Nef.” Bulletin Monumental Societe Francais D’Archeologie. 1992 103-56.
- Laroche L. “Les Berceau Transversaux De Saint-Philibert De Tournus.” Societe Des Amis Des Arts Et Des Science De Tournus. 1941, 134-44.
- “La Genese De Saint-Philibert Macon Et Let Premiere Predications Apostolique.” Societe Des Amis Des Arts Et Des Science De Tournus. 1961, 11-21.
- Lasteyrie, R. L’ Architecture Religieuse En France A L’Evogue Roman.
- Masson, H. “Saint-Philibert de Tournus: E-.sai Sur La Construction de L’Eglise.” Societe Des Amis Et Des Arts Et Des Science De Tournus. 1937. Oursel, C. L’Art Roman De Bourgogne. Dljon, 1928.
- Saint-Paul, Anthyrne. Histoire Monumental de la France.
- Virey, Jean. Les Eglises Romans DeL’ Ancien Diocese de Macon. Cluny Et Sa ReOion. Macon. Protat. Paris, 1935.