The cathedral as a stone argument

A Gothic cathedral is usually described as a prayer in stone. It is more useful to describe it as a balance sheet with a roof. The building records, in the same object, who had the money, who had the labour, who had the quarry, and which neighbouring town had to be outdone — and it records them in a form that cannot be edited afterwards, because stone does not take corrections.

That is why the engineering matters more than the decoration. The decorative programme is the part we can read easily; the structure is the part that says what the town was actually able to do.

Light needs height, and height needs three tricks

The Gothic is not a style that happens to be tall. It is a set of three structural moves that make height possible, and the theology rides on them. Abbot Suger, rebuilding the choir of Saint-Denis outside Paris in the early 1140s, wanted light: his own account of the building is an argument that divine light enters through the windows and that the material of the church should therefore be as transparent as it can be made.

The three moves are the pointed arch, the rib vault and the flying buttress. The pointed arch is geometry: a round arch pushes outward in proportion to how wide it is, but a pointed arch can be made any width to any height, so the builder can keep the span narrow and send the thrust down steeply. The rib vault takes the load off the whole wall and puts it on a few diagonal lines meeting at the piers — which means the wall between the piers is no longer carrying the roof, so it can be opened. The flying buttress takes the remaining outward push and hands it to a pier standing outside the building, so the upper wall does not have to be thick enough to resist it.

Put the three together and the wall becomes a frame. That is the innovation: not that the building is tall, but that the wall stops being load-bearing and becomes a place to hang glass. Everything else — the tracery, the sculpture, the colour — is carried on that one structural fact.

Who paid, and what they thought they were buying

Chartres is the clearest case because the fire is documented. In 1194 the Romanesque cathedral burned; the reliquary of the Virgin's veil survived, and the town's income depended on the pilgrims who came to see it. The rebuild began almost at once and the nave was standing by 1220 — twenty-six years, which for a building of that size is fast. Speed in medieval construction is a financial decision, not an aesthetic one: it means paying more men at once, in more seasons, with more scaffolding up at the same time.

Then look at the windows. Chartres has 172 bays of glass, and a large number of them were given by guilds and corporations rather than by the cathedral. The donor appears in the lowest panel, at eye height: the bakers with their bread, the water carriers with their barrels, the furriers, the moneychangers. A stained-glass window in this period is not primarily a picture for the illiterate; it is a receipt, displayed in the most public surface the town had.

Above the donors sat an institution the historians call the fabrica: the building fund, with its own accounts, its own officers and sometimes its own conflicts with the bishop. A cathedral was usually not one client but three — bishop, chapter and town — who wanted the same building for different reasons. The bishop wanted a seat, the chapter wanted a revenue stream from pilgrims, the town wanted the prestige and the trade. The building is where those three interests had to be reconciled in stone, because stone is expensive enough that nobody could build it alone.

The site is a supply chain before it is a design

A cathedral is a logistics problem with a roof. Chartres stone came from Berchères, about eight kilometres away, and the transport of dressed stone over that distance was itself a permanent works department. Add lime kilns for mortar, forests for scaffolding and charcoal, iron for cramps, lead for the roof, and glass from workshops that might be a hundred kilometres away.

The workforce was small and highly skilled rather than vast and unskilled: a permanent lodge of masons and carpenters, swollen by seasonal labour in summer, idle or reduced in winter when the mortar would not set. The master of works was not an architect in the modern sense — he did not produce a finished design and hand it over. He held the templates, decided the next bay, and adjusted as the building went up.

You can see the adjustment at Notre-Dame de Paris. The cathedral was begun in 1163, and its flying buttresses were added after 1180, when the clerestory was raised: the building was made taller than originally planned and the buttresses were the correction. A cathedral is therefore a record of decisions rather than the execution of a plan — which is exactly why it is good evidence about the town that built it.

The race, and the year it ended

Once the technique existed, the competition was obvious and it ran for about a century. Laon's vault is around 24 metres. Notre-Dame is 33. Chartres 37. Reims 38. Amiens, finished in 1269, is 42 — and Amiens is the one usually held up as the balanced solution, the point where height and surface are still in proportion.

Then Beauvais, where the choir reaches 48.5 metres. It was finished around 1272 and part of the vault came down in 1284. The cause is still argued: the piers at the crossing were too few for the span, or the vault was built too thin for its height, or the whole thing was simply pushed past what the material would carry. The precise cause matters less than the result, which is that the sequence stopped. After 1300, French cathedral money went into width, window area and ornament, not into another ten metres.

That is the argument the stone is making. A cathedral is a claim about heaven, and a claim about heaven has to be built out of limestone that has a compressive strength and a quarry that has a delivery schedule. The claim can be sincere and still fail at 48.5 metres. Beauvais is the building where the two ledgers — the spiritual and the structural — are legible on the same page.

What the town got back

Pilgrims spent money. That is the plainest return: a cathedral with a relic drew travellers, travellers needed beds, food, candles and souvenirs, and the town's fairs were timed around the feasts. The building was infrastructure for a pilgrimage economy, in the same way a harbour wall is infrastructure for a fishing economy.

It also bought position. Amiens, Beauvais and Rouen are close enough to have been rivals, and their cathedrals are arguments addressed to each other. A town that could not out-build its neighbour could at least out-glass it, which is cheaper per metre and shows from further away.

And the building bought a claim to be a place rather than a dot on a road. The towers are the receipt for that claim, which is why the unfinished ones are so informative: Chartres has two towers that do not match because they were built decades apart and the money ran differently; Strasbourg has one spire because the second was never afforded. The skyline is a budget statement.

How to read one in ten minutes

Three tests, in order. First, the buttresses: a single tier means one height campaign, a double tier means the building was raised and reinforced later — the second tier is a correction, and corrections are the most honest part of any building. Second, the bottom of the windows: if there are small figures with tools at the base of the glass, you are looking at who paid. Third, the towers: matched or mismatched, finished or truncated.

Then one ratio, which is the whole argument in a single glance: how much of the upper wall is glass, and how much is stone. In a Romanesque church the wall is mostly wall. In a Gothic cathedral the wall has become a frame holding glass, and the frame is being held up from the outside by a set of props that the town paid for and that nobody was ever meant to see as structure. The cathedral is a machine that was dressed as a prayer, and the dressing is very good — which is why we still describe it as one.

This is entry 01 of Faith & Reason in About France. The pillar's argument — that the country of cathedrals and the country of laïcité are the same argument at two ends — is set out on the pillar page.