The Medieval Quest For the Perfect Vault |
Remind yourself of the components and functions of Classical Beauty, and when it became established.
Now—Classical Beauty was already functioning in Early Christian churches like Old St Peter's. We'll look at St Paul's Without the Walls, which is almost identical. The nave and transepts were the same length (1:1), the width of the church was twice the width of the nave (2:1), and the height of the nave made the Golden Ratio with its width. The wooden roof put little strain on the building, so it could have thin walls and large windows to fill it with light, and all that was left to do was to keep it in good repair and cover it with brightly colored paint, gold and colored marble. Thus Old St Peter's, and churches like it, were Classically Beautiful, and could be relied upon to manifest God.
Around the year 1000 AD wooden roofs suddenly began to be replaced with stone vaults and many new churches were built which were designed from the beginning to be vaulted. Raoul Glaber states that at this period, "there occurred, throughout the world, especially in Italy and Gaul, a rebuilding of church basilicas ... each Christian people strove against the others to erect nobler ones". So we can infer that these vaulted churches were considered "nobler" (i.e. more Beautiful) than the older, wooden-roofed ones. When we also see that the first vaults were designed to an Ideal geometric shape: the semi-cylinder (barrel-vault), we can be fairly confident that we are looking at a well-organized attempt to create a new and better interpretation of the Classically Beautiful church.
Now let's consider the structural implications: a vault is always trying to fall down (diagram a), but so long as it is curved and the blocks are wider at the top than at the bottom, it cannot do so (diagram b). So its downward energy (thrust) gets diverted sideways. This threatens to make the supporting walls burst, so they have to be very thick and solid. However far outwards the line of thrust goes, that is how wide the supporting walls need to be - for the full length of the vault!
In fact, the situation not quite as bad as this because once the sideways thrust has gone into the walls, it meets a new downwards thrust caused by the weight of the masonry in the walls plus anything on top of them (like statues, towers, etc). These two thrusts now join forces, pushing each other slightly out of alignment - so the downward thrust goes slightly side-ways, and the sideways thrust goes more downwards (diagram b again). This means that the walls can be a bit narrower than we might have predicted, and the windows can be a little larger. But not much. Typically, in Romanesque churches, a 30' wide vault is supported on 10' thick walls. But the fact that we can physically "push" the line of thrust around by making sure it meets with other thrusts creates new opportunities.
Suppose you pass one barrel vault at right-angles through another one. That will give you a square space where they cross, and the vaults will meet at a great X-shaped ridge (the arris). This is called a groin vault (diagram c). The thrusts running sideways out of each vault meet at the arris and flow together. But the arris is also curved, so it has a thrust of its own—and it's a bigger curve and therefore has a stronger thrust than the two which meet it, and it's running in a different direction, so it can force the other two thrusts backwards and downwards. These three thrusts now flow together in a much narrower area and you can put quite slender supports (piers) at the corners of the arris. You can put a series of these together and support them on piers so you don't need walls, as in the nave. Or you can wall in the space between the piers if you like, or put in big windows. Should be good?
But you have now created a new problem. The vault is sound, but the arris is a bad shape. It is not geometrically Ideal. If you're lucky, it will be a half-ellipse (which will not be Ideal until Raphael cracks its geometry in the sixteenth century). More likely it will be wonky, with a crooked edge and a dip at the centre. So the groin vault is not Classically Beautiful and you will almost never find it over the nave of a Medieval church where people would be exposed to it. But you will find it in the aisles, probably because it allowed for better windows.
Well, you can get round this problem, but you have to stop thinking about the vault and start thinking about the arris instead. If you build the arris first, you can make it a perfect semi-circle. Then you can fill in the vault round it. What you have done is make the arris into the skeleton of a dome which is an Ideal shape (diagram d). What happens to the shape of the vault now depends on whether you decide to make it dome-shaped, in which case the top of the nave undulates like a series of bubbles; or whether you decide to keep the top of the nave level (which is also Ideal), in which case the sides of the vault will rise to a point (diagram e). In either case, this kind of vault is called a domed-up groin vault—or if it is emphasized with decorative ribs it is called a rib vault instead—and you will find it with or without ribs over the naves of many Medieval churches.
Remember: if there are no ribs, look at the shape of the arris. If it is semi-circular, you are looking at a domed-up groin vault. If it is wonky, you are looking at a (plain) groin vault. If the vault has not been plastered, look also at the masonry pattern. On a groin vault, the masonry courses will spring perpendicular from the side walls and run in parallel lines to meet the arris, where they make a chevron pattern. On a domed-up groin, they will spring perpendicular from the arris and meet at the top of the vault to make a chevron pattern there.
Remember also that the pointed arch is a side effect of the Classically Beautiful domed-up groin or rib vault. It is not the cause: it is the result.
To emphasize how important this project was, consider this. The vaults started being built c1000. What doesn't get told, is that they kept falling down! Some vaults had to be built five times before they could be made to stay up. This was a very big problem! It took nearly a hundred years and many failures to solve it, and the solution is pure genius. In my view, that is the reason for the ribs. They are purely decorative, but they really draw your attention to the Classical Beauty of the vault.
Norwich Cathedral makes a delightful post-script to the problem of the vault. It is the only cathedral in England with an intact Norman apse with ambulatory. The ambulatory has an ordinary groin vault (look at the arrises) and was begun in 1096, at much the same time as Durham Cathedral. Durham has the earliest surviving rib-vaults in England (indeed, they may have been invented there) and they were completed in 1104. The ambulatory at Norwich was painted in 1111. At that time, the wall was plastered and painted with Ideal masonry: perfect double cube blocks spring perpendicularly from the arris and meet in a chevron at the top of the vault, and the arris is emphasized with a painted rib! A mason can dream ...
Having got the buildings right, the other big issue is the windows. Check out what they did before there was stained glass.
Or find out how Gothic architecture echoed the power of stained glass.
Or go back to Why Stained Glass?
Why Use Stained Glass? | Classical Beauty | Before Stained Glass | Seeking the Perfect Vault | The Vault as Illusion |