THE ‘HORIZONTAL RIB’ OF THE CRAZY VAULT INVENTED BY GEOFFREY DE NOYERS AROUND 1190

WILBERT WALTA, Independent Scholar

Abstract

Vaults were an important aspect of the many impressive Gothic cathedrals built throughout Europe during the Middle Ages. Extremely complex structures, these vaults differed from city to city. Master masons had reached dizzying levels of expertise in the Middle Ages. They were responsible not only for the aesthetic aspects of the vaults but also for their stability – an extremely demanding role, with strength calculations to determine the future behaviour of the design still several centuries away. So, how were the master masons able to construct all these different types of vaults? What was their guiding concept? Did they each have their own techniques that were unique to a particular city, or was there an overarching European model?

Online image banks containing extensive photographic documentation of the interior and exterior of more than 3,000 European Romanesque and Gothic churches may provide an answer. It appears that all the late-Gothic examples were designed based on a separation of functions between ribs and compartments and that a special type of rib referred to here as the ‘horizontal rib’ was used. This concept was applied for the first time by Geoffrey de Noyers around 1190 in his design of Lincoln Cathedral’s ‘Crazy Vault’.

The research question can thus be framed as follows: What was the significance of Geoffrey de Noyers for European vault building in the Middle Ages? The knowledge level of the master masons and guilds operating at that time can be deduced by examining the cathedrals and churches built before 1190. Taking this as a benchmark, it is then possible to consider the ways in which Geoffrey de Noyers developed and expanded upon this knowledge.

Two structural elements were well known in the Middle Ages: the arch and the groin vault. Invented by the Egyptians around 2000 BC, the arch is a construction based on the principle of stone stacking. Stone arches are extremely robust compared to wooden structures with the same span. For constructions that had to endure the test of time, a stone arch was the ideal solution.

The groin vault, which also occupies an important place in Gothic architecture, was invented by the Greeks in the third century BC. A complex construction, it consists of two diagonal arches with other superimposed arches. This was a standard construction during the Roman Empire. Many Italian churches featured groin vaults by the Middle Ages, including several improvements on the earlier design. The most important of these was the simplification of the wooden construction of the mould in which the vault was installed. The diagonal arches were no longer the intersection of two barrel vaults but rather took the form of a semicircle. At the beginning of the 12th century, French groin vaulting was further improved by the addition of ribs. The vaulted fields were now slightly curved so that they could be bricked with the free hand, which greatly reduced building costs.

This was the existing knowledge base employed by master masons and guilds around 1190, the period in which Geoffrey de Noyers designed the asymmetrical vault in Lincoln Cathedral choir – the so-called ‘Crazy Vault’. A departure from the standard groin vault, this new design raised questions as to its stability, leading de Noyers to conclude that this was only possible using ribs to support the compartments; and this in turn would require the addition of an extra ‘horizontal’ rib. His concept – a bearing cage of well-designed ribs – was a revolution that freed the master masons from the limitations that had restricted them up to that point.

The concept was ingenious in its simplicity, with far-reaching consequences. In England, it led to the development of the fan vault, and later to the innovative rib cage constructions that became a feature of English vault building in the Middle Ages. The central role that England played in the development of the medieval European vault was described in the 19thcentury by Viollet-le-Duc, a well-known expert of the time. As a Frenchman, however, he had little appreciation for English expertise – a fact that may have contributed to the work of Geoffrey de Noyers being underestimated. However, it is high time that we re-consider his important contribution to medieval European vaulting.

Keywords

Geoffrey, Noyers, gothic, concept, arch, groin vaults, construction, funicular, quadripartite, tripartite, crazy vault, fan vault, master mason, guild, function separation, freehand bricklaying, horizontal rib, Viollet-le-Duc, Robert Hooke

Introduction

The Middle Ages were known for their impressive Gothic cathedrals built in large European cities, of which the vaults were a key element. The observer looking up from the cathedral floor sees – unconsciously or consciously – an exceptionally complex construction. This construction is exclusively composed of stones and masonry mortar. The vault in one bay of an average cathedral weighs around 20 tonnes. This mass of stones and mortar hovers, so to speak, about 20 metres or more above the floor of the cathedral. The design of the vault resists gravity through an ingenious structure of loose stones, which are stacked in such a way that they clamp onto one another, stabilising the construction.

By the end of the Middle Ages, some of these vaults gave the impression of mocking gravity. Vaults of completely different design, meanwhile, were constructed in European cities. On this basis, it could be assumed that the art of stacking stones is such a way is not particularly complicated. However, this is far from the case. Stones can be stacked in countless ways, but only a few specific methods result in the construction of a stable vault. These stacks are so specific in nature that each individual technique can rightly be called an ‘invention’.

Calculating the strength of vaults (now a legal obligation for any building) was not possible in the Middle Ages. The scientific basis for this calculation only came about much later, with Isaac Newton (1643–1726). In addition, vaults are statically indeterminate. In layman’s terms, this means that vault calculations can only be made if the material properties of the stones and mortar are also known. This was not the case in the Middle Ages. All of the modern tools that predict future behaviour and mitigate against potential problems were hence unavailable in the construction of these vaults.

Medieval master masons, the leading members of a guild, were both designers and constructors. They were responsible for both the beautiful appearance and the strength and stability of their vaults. Training and experience gave them the knowledge of what was technically possible or not. The most creative master masons designed new forms of vaults using inventive, novel ways of stacking stones. However, the knowledge and expertise of the guild members who actually built the vaults was also an important factor in the process.

By the end of the Middle Ages, the vaults in different European cities were somewhat diverse (1). This raises the question: How were they constructed? Were they all independent designs –  special stacks of stones developed separately by different master masons from different regions –  or was there a common ‘invention’ linking these ostensibly disparate forms?

Developments in vault building mainly occurred in the Gothic era, beginning with the vault of the deambulatory of the Basilica of St Denis built in 1140 /see Viollet-le-Duc(2). It is therefore logical to limit the scope of this research to this early Gothic period.

Modern image banks such as http://www.gotik-romanik.de (1) contain pictures of more than 3,000 Romanesque and Gothic medieval churches. A comparative phenomenological analysis of innovative or novel vaults reveals that almost all were built of ribs and compartments, and that a special role was reserved for a rib known here as the ‘horizontal rib’. From this phenomenological survey, it emerged that:

• All late Gothic vaults are designed based on a separation of functions between ribs and compartments. The ribs that must meet high requirements feature compartments with much lower requirements.
• The horizontal rib is used in these vaults. This rib has a special function and special properties.

This is the basic concept underlying almost all late Gothic vaults, giving rise to the question: When was this concept developed, and by whom? To which the answer: In around 1190, in the design of Lincoln Cathedral’s ‘Crazy Vault’ (3) by Geoffrey de Noyers.

This leads to a second question, which will be the main focus of this paper: What role did Geoffrey de Noyers play in the development Middle Ages vault-building techniques?

Geoffrey de Noyers was a master mason, trained by a guild of masons. In investigating his contribution to the craft, we much first establish what knowledge the guilds had at that time. Which structures and which stone configurations were they aware of? Only then can we determine the level of Geoffrey de Noyers’ technical invention and innovative importance.

Basic construction of the guilds: the Egyptian arch

The level of knowledge of the guilds can be deduced by observing changes in the vaults constructed over time. The image bank shows that numerous Romanesque churches were built throughout Europe during the 11th century. In Italy, France, England and Germany, a number of these churches featured barrel vaults, which are semi-circular in cross-section. In the 12th century, arches were semi-circular, and barrel vaults were the traditional, standard forms of construction based on this structure. Arches were invented around 2,000 BC by the Egyptians /El Naggar, 2001 /(4), and knowledge of arch-building techniques later spread across Europe. The importance of this invention cannot be underestimated. Cultures that had no contact with  Egyptian techniques (for example, the Aztecs) were unfamiliar with either arches or barrel vaults.

The arch – smart stacking of stones

The arch is a complex but basic form designed on the simple principle of a stacked construction. The parts are loosely stacked in such a way that the stones forming the arch keep each other in balance and clamp to one another (see figure 1).

Figure 1. The arch is a loose stack of stones that clamp together under their own weight. Stability is achieved when the length of the arc is greater than the hole into which it can fall (the span).

The success and popularity of the arch is mainly down its one defining feature: its enormous strength. Under certain conditions, a stone arch can support virtually any weight (5). In the Middle Ages, a stone arch was by far the strongest construction, greatly surpassing a timber truss for the same span. When a structure was built to endure for thousands of years the arch was the only option – a situation that remained the case until iron profiles came into use on an industrial scale in the 19th century.

A stable stone arch had to meet certain conditions, which at the time were shrouded in mystery. Every master mason understood the conditions needed for stability, but none knew exactly how or why this was the case. The master masons used the rule of thumb set out by Gaetani, Giorgio, Paulo B. and Marcari: the thickness of an arch is 10% of the radius (6). This thickness is also technically justified, according to modern standards developed by Heyman (5). However, theoretical knowledge in the Middle Ages did not extend any further than this rule of thumb.

It was only in 1670 that Englishman Robert Hooke provided further insight. At a lecture to the Royal Society, Hooke stated that the ideal arch had the shape of a reverse chain. A chain, like an (unloaded) stone arch, is equally heavy along its full length. If a chain is loosely suspended between two points, the chain takes on a certain shape. If that shape is turned upside down, it gives the shape of an ‘ideal curve’ or funicular. The shape of this ideal curve for an arch is indicated in figure 1 (dashed line).

Hooke argued that a stone arch would stand firm if the ‘ideal arc’ or funicular fit within its contour. It can be deduced that the stone arch shown in Figure 1 is not stable because the ideal arc (dashed line) runs outside the contour. A solution to this problem in the form of an infill (figure 1) had already been discovered long before the Middle Ages: the space between the sidewall and the arch was filled with a sand/cement mixture that could absorb the forces crossing the contour. Infills of this type can be seen in every medieval vault.

Figure 1 illustrates two other aspects that play an essential role in the development of vault construction. As mentioned above, Hooke stated that a stable arch requires that the ideal arc must fit in its contour. If we do not take the whole arch but only a part of it – for example, the middle part (indicated by: width <%... figure 1) – then the arch for that part turns out to be much thicker than required. The ideal arc or funicular that fits in this part of the arch is only 1% of the radius, while the arch shown in figure 1 is 10% of the radius. The construction of short, slightly curved compartments (discussed later) is based on this property.

The second aspect is considerably more complicated: How can one ensure that the ideal arc does not cross the contour but stays within it over the full length of the arch? The solution was already known at the beginning of the Roman era; however, a mathematical explanation was not forthcoming until recently. Tsutsui, Mizuta and Sakata (7) have shown that an ideal arc exists when it follows the intrados – the lower or inner curve of an arch. In this case, however, the arch is not evenly loaded over the entire length. Rather, the load increases from the centre to the outside portion, according to a specific mathematical formula. 

Practically formulated, an arch becomes stronger to the extent that the load increases from the centre to the corner. While this may seem an obvious property, it is extremely important for groin vaults. Due to the rectangular shape of a groin vault, the load on the diagonal (internal) rib increases from the inside out. Due to its structure alone, this feature makes a groin vault more stable than a barrel vault.

Greek groin vault

The groin vault, as a concept, plays an essential role in the Gothic vault. The basic groin vault consists of two internal diagonal arches on which other arches are superimposed.One of these diagonal arches is indicated with a dark band in figure 2 (left image). The arch is supported on both sides by the internal diagonal arches. A groin vault is a complicated structure due to the superposition of these arches.

Figure 2. Left: groin vault invented by the Greeks in the 3rd century BC. Right: groin vault as applied by the Romans during construction of the Colosseum in 90 AD.

The groin vault was invented by the Greeks in the 3rd century BC, almost 2,000 years after the invention of the arch, as shown by Boyd (8). The Romans were the first to apply the groin vault on a large scale. They also distributed the groin vault throughout Europe and the Middle East. This style of vault does not occur outside these regions. The storey floors of the Colosseum in Rome consist largely of groin vaults, according to the principle demonstrated in the right-hand image in Figure 2. (9) 

Italian improvements to the groin vault

A groin vault, like the arch, is built on a mould – a wooden structure exactly at the intrados of the vault on which the stones are placed. The mould for the original Roman groin vault is complicated and therefore expensive. A simplification invented in Italy at the beginning of the 10th century involved replacing the diagonal arches of the Roman vault, which were by definition elliptical, with circular diagonal arches (10, 11). The result of this adjustment is shown in figure 3.

Figure 3. Italian groin vault shortly after 1000 AD. The diagonal arches have been replaced by semicircles (arrow).

As seen in figure 3, the direct application of this method does not produce a beautiful vault. A lowering in the vault caused by the straight planks of the supporting mould is visible near the top. Viewed from the floor, that lowering appears as a sag. The master masons found two solutions to eliminate this unwanted sag. First, the diagonal arches of the groin vault were placed lower than the transverse side arches of the vault, with the result that the sag disappears. This solid and relatively simple solution is seen in the crossing of Como Carpoforo built in 1050 (12). The better, and also structurally more complicated solution,  is the cambering of the vault, giving the compartment a convex shape. How the master masons managed to achieve this cambering remains a mystery. Possibly, the wooden mould was filled with sand to achieve a rounded top to which the stones could eventually be applied. Examples of cambered groin vaults are the side aisles of Pavia Lomello year 1025 (12), Milan San Celso 1050 (10) and Speyer Dom 1085 (1).

Gothic inventions of the 12th century

Figure 4. The longitudinal section of the Sante Saviano church in Piacenza. The groin vaults were ribbed only for a short time. Source image: Porter.

Figure 5. The Gothic invention of bricklaying vaulting using the ‘free hand’. The continuous lines represent arches for which moulds were made. Compartments between the ribs were built layer by layer in the free space with the help of a single, slightly rounded plank.

Innovations once again from Northern Italy at the end of the 11th century. Heavy ribs were introduced that should have helped reinforce the diagonals of the groin vault. However, this turned out to be unsuccessful (12). Only a few vaults of this type remain, with most lost due to their constructional shortcomings. Italian experimentation with ribs was finanlly terminated, as seen from the longitudinal section of the Sante Saviono church in Piacenza built in 1100 figure 4 (12). At that time, churches were built from east to west, meaning that the masons started with the choir area and then moved onto the nave. As shown in figure 4, the first (left) groin vault of the nave (next to the choir) still contained ribs (arrows). For the second groin vault, the groin ribs are slightly smaller, disappearing completely in the right vault.

The negative Italian experiences did not discourage master masons on the other side of the Alps, in France, from experimenting with ribs. As Porter describes (13) they managed to use these structures with some success. This was mainly due to three improvements. The first Italian ribs were designed as an arch with a relatively large, square cross-section. As a result, they were too strong and too heavy for the rest of the vault. The French replaced these heavy square ribs with light, flexible arches built from relatively thin, light stones. These ribs had no constructive significance but were primarily aimed at reducing costs. The sharp corners on the diagonals of the groin vault were difficult to make, and could be avoided by applying a rib. The rib covered the corner, making the construction much easier to complete. Groin vaults with light flexible ribs thus soon became the new European standard.

The new French ribs did not change the way a groin vault was built. The traditional, 1,000-year-old method was still in use at the beginning of the 12th century. A negative form mould (the centering) made of wood was placed at the place of the vault. The space for the ribs was also saved in the mould. From this, we can conclude that the ribs were not considered a separate part of the vault. The entire vault was seen as a single, uniform construction, with the vault installed on the mould. When the vault was completed and properly set, the mould was removed and the vault could be completed. As stated by James (14), the vault of Durham Cathedral was built this way at the beginning of the 12th century

The second Gothic invention was the replacement of the semi-circle – the standard form for arcs in Romanesque churches. The new pointed arc consisted of two identical circlular segments. The pointed arc made it much easier to vault non-square spaces. The advantages of the pointed arc were generally understood and described in detail by Viollet-le-Duc (2).

Figure 6. Deambulatory of the Basilica of St Denis in Paris, built in 1140. Viollet-le-Duc designated this vault as the starting point for the Gothic style. For two centuries, French cathedrals featured vaults of this type (quadripartite). Photo:Author.

The third Gothic innovation was the freehand bricklaying of vaults (figure 5). Often, Romanesque compartments were already cambered. A mason in the Paris region discovered that it was not necessary to make a complete mould for the entire vault. Rather, it was sufficient to manufacture moulds for the groin, transversal and side ribs. For the brickwork of compartments, a simple support between these rib moulds was sufficient for just one layer of stones. Once that layer had been completed, its arc form would stand in place. The support could then be used for the next layer of bricks. This method of freehand bricklaying saved wood, with a complete set of wooden moulds no longer required. 

In a Gothic vault, only the ribs are supported with moulds. The combined system of a wooden mould and stone ribs was already quite strong in the building phase, and could easily support the load of the freehand-built compartments. Figure 6 demonstrates the standard form of the (French) Gothic groin vault in around 1140. All churches and cathedrals built in the next two centuries in France featured this type of vault, as Viollet-le-Duc remarks in his Dictionnaire (2).

The invention of Geoffrey de Noyers

Figure 7. Geoffrey de Noyers’ Crazy Vault was designed around 1190. The base is formed by a tripartite vault with three ribs that run from boss 1 to A, B and C, respectively. A bay has two tripartite vaults. The horizontal rib (indicated by the letter R) closes the gap between the two bosses. Source photo: Wikipedia.

In the year 1186, Hugh of Avalon, a French nobleman, was appointed Bishop of the diocese of Lincoln, which at that time included a large part of England. The Bishop wanted the choir area to feature a vault. To complete this assignment, he chose one of the most prominent master masons of the day, Geoffrey de Noyers (15), who had worked for many years at Canterbury Cathedral under the direction of William of Sens, the most famous master mason of the 12th century. William had designed and built the cathedral at Sens in France –  the first construction utilising the full range of the technical innovations underpinning the Gothic style. The cathedral in Sens is a huge space. Instead of thick, heavy walls, it has slender pillars supporting the vaults. 

Large windows allow a huge influx of daylight. Romanesque churches built prior to this had thick walls and small windows that made their interiors somewhat dark. The cathedral at Sens was a European wonder of the world. After this, nearly all European churches in the Middle Ages were built in the new Gothic style.

Geoffrey de Noyers was well aware of the latest developments in vault construction 15 and the French basic form for the vault in a bay (Figure 6). Yet he went his own way and designed the ‘Crazy Vault’ (Figure 7). Instead of the standard French symmetrical quadripartite vault, he chose the tripartite vault as his template –  a form seen in the deambulatories of certain cathedrals or their crypts (e.g., the cathedral at Aachen, built at the beginning of the 9th century).

Figure 8. Left: one tripartite vault from the previous figure. Right: This vault has been turned slightly to show how the ribs support the compartments. Rib A supports the parts marked A; rib B supports the parts marked B; and rib C supports the parts marked C.

Geoffrey de Noyers used this type of vault for a rectangular bay. Instead of one quadripartite groin-rib vault per bay, he needed two tripartite groin rib vaults for each. This choice led to asymmetrical tripartite vaults (figure 7; tripartite vault ABC). The stability of French Gothic vaults was due to their symmetry: the left and right halves are mirror images that balance each another. This symmetry does not apply for the asymmetrical tripartite structure used in the Crazy Vault (figure 8), where the weight of the left part is clearly greater than the weight of the right part.

Constructive analysis of Geoffrey de Noyers’ invention

Figure 9. Left: Forces at boss 1, the top of the vault. Right: the forces are in equilibrium when they form a closed polygon. The forces A, B and C are thus unbalanced. Balance can only be achieved by adding the horizontal rib R

At the beginning of the 12th century, ribs were not seen as a separate aspect of the vault. Ribs and compartments were made in the same mould. At the start of the Gothic period, however, moulds were made only for the diagonal ribs, the transversal ribs and the side ribs. The compartments were bricked with a free hand on a light, movable plank. Geoffrey de Noyers concluded that the ribs were the actual carriers on which the compartments rested. 

Figure 10. Geoffrey de Noyers stabilised the two unstable tripartite vaults with his revolutionary invention of the 'horizontal rib'.

This seems a clear conclusion, but what exactly does it mean? Which rib supports which part of the vault? A precise answer to this question is extremely difficult to achieve. However,  a relatively simple approach provides a reasonably accurate indication. Gothic vaults are slightly curved (see figure 5, compartment). The ends of the compartments rest on the ribs. As an initial approximation, it can be assumed that a rib carries half the compartments resting on it. Compartments can be divided into two, allowing us to determine which part rests on each rib. The thickness of the compartments (above the infill) is the same throughout, so the surface of the parts of the compartments resting on a rib gives the load on a particular rib. 

This is shown schematically in figure 8. The classification follows the approach according to figure 4.16 from Heyman’s Stone skeleton (5). The parts of the vault resting on rib C are marked with a C, on rib B the parts are marked with B, and likewise for rib A. The uniform grey part does not rest on ribs A, B or C but on the transversal rib, which runs in a straight line from A to C, at the border of the bay.

The forces that the ribs exert on top of one another can be estimated with the help of these assumptions. The problem of asymmetry between the left half and the right half of the vault is thus shifted to the horizontal forces that ribs A, B and C exert on each other at boss 1. The horizontal forces in ribs A, B and C are not balanced. Using his experience and technical insight, Geoffrey de Noyers was able to establish the imbalance at the top. 

The forces in these ribs can be determined using Grafostatica, invented by K. Culmann in 1866 (16). The forces in the top of the vault are shown graphically in figure 9 (left image). Their magnitude is based on the weight distribution given in figure 8. These forces were calculated using GeoGebra software. The computed forces can also be placed behind each other while maintaining their orientation (vector addition). If these forces form a closed polygon, then they are in balance. In the right-hand image of figure 9, these forces are placed behind each other. As expected, the ribs are unbalanced because the forces in the ribs B, C and A do not form a closed polygon. The polygon can only be closed by force R, which is provided by the horizontal rib R in the axis of the vault. There is no other solution due to the directions of the forces A, B and C. Rib R has a constructive rather than a decorative function, and is indispensable for the balance between the ribs.

Rib R thus has a special character. This rib provides the balance between ribs A, B and C, and its role in supporting the compartments is secondary. For this reason, it is referred to here as a ‘horizontal rib’. It fulfils an essential function in the rib cage construction that carries the vault in a bay (see figure 10), and was used for the first time by Geoffrey de Noyers.

The significance of Geoffrey de Noyers’ invention

Figure 11 The vault of the chapter house of Lincoln Cathedral was designed by a student of Geoffrey de Noyers in 1220. Here, too, the base is a tripartite vault. Different ribs have the same designations (A, B and C) as in the previous figure. The horizontal rib is also indicated. These ribs R form a regular polygon and can therefore absorb forces. Source photo:Wikipedia.

Figure 12. The Jacobins church in Toulouse. A fan vault was chosen for the choir, as in the Chapter House in Lincoln. Unusually, the rest of the church was designed around a row of columns running through its middle. Source photo: Wikipedia.

The many possessions of the wealthy diocese of Lincoln were administered by a chapter that assembled in the Chapter House (figure 11) built in 1220. It is assumed that this was designed a few decades after the Crazy Vault by a student of Geoffrey de Noyers. The Chapter House vault (figure 11) feature a creative application of Geoffrey de Noyers’ innovation. The vault is a regular polygon with 10 sides. The same basic element (tripartite vault: A, B and C) used in the choir was employed with one significant difference: the ribs are now positioned alongside rather than opposite one another, creating a round vault. The horizontal ribs form a ring, absorbing the forces created by the asymmetry.

When looking at the construction it can be stated that this is actually a fan vault. If the central part is vertically divided into four equal parts, each part forms a quarter of a fan vault. For what concerns the construction this is the first fan vault ever built, and it made a deep impression in Europe. Around 1275, a whole church – the Jacobins church in Toulouse, France – was designed and built around this concept (17). Basically a fan vault was chosen for the choir, the most important feature of the church (figure 12). This concept required that a pillar was placed in the middle of the choir. The standard layout for churches (middle nave and side aisles) was replaced here by a nave with pillars down its centre. This church is the only exception to the standard layout seen in the 13th century. It is also the first of only a few fan vaults on French soil, and almost the only 13th- century vault in which the French standard groin rib vault was not applied.

Figure 13. Vault in the nave of Lincoln Cathedral. The almost horizontal rib R connects bosses b1 and b2 and supports ribs B-b2 and C-b2. Source photo: Wikipedia.

The vault in the nave of the Lincoln Cathedral was also built by a student (15) of Geoffrey de Noyers in 1220. The base of this vault is the standard French quadripartite vault. Its special feature is the tripartite vault installed in each of its four compartments. Figure 13 shows one compartment from this vault. The ribs are indicated by C-b2, B-b2 and R between b2 and b1. This is not an ordinary tripartite vault but rather a tripartite vault in which one of the ribs is replaced by an almost horizontal rib (R).

Rib R does not follow (or perhaps we should say: ‘does not have to follow’) the curvature of ribs B-b2 or C-b2. It can fulfil its underpinning function as a straight rib, as long as the distance b2-b1 is not too great. This form illustrates the innovative character of the horizontal rib. The keystone, boss b2, is no longer a keystone in a regular tripartite vault but rather a node in a three-dimensional rib cage. This is the essence of Geoffrey de Noyers’ horizontal rib, and the starting point for the great strides that would be made in European vault construction thereafter.

Figure 14. Vault with horizontal ribs and bosses in the aisle of Wells Cathedral. Photo: Author.

More than a hundred years later, in 1324 (18), the side aisles of Wells Cathedral featured vaults (Figure 14) that were a movement away from the French standard form to a freer three-dimensional design. Figure 14 shows a bay of the northern side aisle. The master mason’s methods and thought process can be seen as an extension of the design of the Lincoln Cathedral nave (figure 13). The construction of the tripartite vault in the nave consists of C-b2, B-b2 and R (stable). In this construction, not just one of the three ribs (as in figure 13) but also a second rib (e.g., rib C-b2) is replaced by a second horizontal rib, which is also stable. This laid the foundation for a spatial cage construction –  an innovation that would have a significant impact of the future of European vault construction.

Europe

Figure 15. Four of the most beautiful and complicated medieval vaults. (1) Cell vault in the Albrechtsburg in Meissen, Germany; (2) Vault in the nave of the Hieronymites Monastery; (3) Vault in the Henry VII Chapel in Westminster, London; and (4) Vladislav Hall in Prague. Source/all photos: Wikipedia.

Around 1500, the European master masons had reached unparalleled technical abilities. In London, work was done on the vault of the Henry VII Chapel of Westminster Abbey, completed in 1510 (figure 15, no. 3). This is a fan vault of a type mainly seen in England. In Prague, the vault of Vladislav Hall (figure 15, no. 4) was completed in 1495. This is a sling rib vault of a type mainly seen in Eastern Europe. In Meissen, Germany, in 1471, the Albrechtsburg featured cell vaults (figure 15, no. 1) of a type found only in Eastern Europe. The vault in the nave of the church of the Monastery of Hieronymites near Lisbon (figure 15, no. 2) was a combination vault completed in 1528 (19). All these vaults feature three-dimensional rib cages, and can be seen as extensions of the basic concept developed by Geoffrey de Noyers.

The judgement of Viollet-le-Duc

The impressive medieval innovations in English vaulting, initiated by the invention of Geoffrey de Noyers, were already noted in the 19th century by the Frenchman Viollet-le-Duc, a recognized expert. As a Frenchman, his appreciation for these innovations was limited. He preferred the standard quadripartite vault, which was used in French cathedrals until the 14th century with very few changes. The French were justly proud of the quadripartite vault they had developed (2), and remained loyal to it for a long time. The essence of Geoffrey de Noyers’ innovation is his analysis of the Gothic vault. Unlike the master masons before him or the French masters after him, he did not accept the quadripartite vault as an indivisible unit. Rather, he divided the vault into ribs and compartments, each with its own function. It was a decisive invention that created new-found freedom in vault construction, and which had a significant impact on European vault design over the centuries that followed. Sadly, the brilliance of his invention never received the attention it deserved, perhaps due to Viollet-le-Duc’s instinctively negative reaction to English innovation.

Conclusion

Geoffrey de Noyers can be considered the inventor of the horizontal rib – an invention that underpins some of the most beautiful vaults of the Middle Ages. However, his importance and influence is given only incidental mention in the literature, which in this author’s view is wrong. The enduring impact of his technical innovation and inventiveness deserves greater scholarly recognition and merits further re-evaluation.

wilbert_walta@icloud.com

REFERENCES

1.        Dr.-Ing. Heinz Theuerkauf. Gotik-Romanik.de von Heinz Theuerkauf [in German]. Available at <http://www.gotik-romanik.de/>.

2.        Viollet-le-Duc E. Dictionnaire Raisonné de l’Architecture Française du XIe au XVIe Siècle. Morel, Paris 6, (1854),538.

3.        Frankl P. The 'Crazy' vaults of Lincoln cathedral. Art Bull. (1953). doi:10.1080/00043079.1953.11408169

4.        El Naggar S. Les voutes dans l’Egypte pharaonique [in French]. (2001). 74–78. 

5.       Heyman J. The Stone Skeleton : Structural Engineering of Masonry Architecture. Cambridge University Press, 1997.

6.        Gaetani A, Monti G, Lourenço PB, Marcari G. Design and analysis of cross vaults along history. Int J Archit Herit.(2016). doi:10.1080/15583058.2015.1132020

7.        Tsutsui M, Mizuta Y, Sakata T. Line of thrust and theoretical load of masonry arch bridge. ARCH´10 – 6th Int Conf Arch Bridg. (2010). 0–5.

8.        Boyd TD. The Arch and the Vault in Greek Architecture. Archaeological Institute of America: Boston, MA. (2012). 

9.        The-Colosseum.net: Figures. Available at: http://www.the-colosseum.net/architecture/numeri_en.htm

10.      de Dartein F. Étude sur l’architecture lombarde et sur les origines de l’architecture ...[in French]. (1865).

11.      Porter A. The construction of Lombard and Gothic vaults. (Yale University Press, 1911). doi:10.1016/S0140-6736(13)62568-4

12.      Porter AK. Lombard architecture. (1967).

13.      Porter AK. Medieval architecture. (The Baker and Taylor, 1909). Available at: https://archive.org/details/medievalarchite02portgoog/page/n11

14.      James J. The rib vaults of Durham Cathedral. Int Cent Mediev Art. 1983;22:135–145.

15.      Hendrix JS. The splendor of English Gothic architecture. (2013).

16.      Culmann K. Die graphische statik [in German]. Meyer & Zeller, Zurich (1866). doi:10.1016/S1003-6326(10)60534-1

17.      Stephenson D. Heavenly Vaults: From Romanesque to Gothic in European Architecture. Development 134,Princeton Architectural Press (2009).

18.      Webb N, Buchanan A. Tracing the past: a digital analysis of Wells cathedral choir aisle vaults. Digit Appl Archaeol Cult Herit. (2017). doi:10.1016/j.daach.2017.01.001

19.      Genin SM. Form, design and construction of ribbed vaults. João de Castilhos innovations in the Jeronimos Monastery, Lisbon (1470–1552). Int J Construct Hist Soc. 2018;33(1):27–48.