Flying Buttress: Structure and History

Flying buttress definition

An attached buttress thickens a wall along all or most of its height. A flying buttress leaves a gap: its arch or strut rises above an aisle, chapel or roof and meets the upper wall near the level where support is needed. The open interval makes it possible to brace a tall central vessel without filling the surrounding lower spaces with masonry.

Specialist terminology is narrower than everyday usage. Structural studies may call only the bridging arch the flying buttress or flyer, reserving culée for the masonry block at its outer end. General descriptions often use “flying buttress” for the whole assembly of flyer, pier and pinnacle. ^[6] Getty’s controlled vocabulary likewise identifies the term as a distinct architectural support. ^[14] When reading a survey, check which meaning is intended.

Flying does not mean suspended. The stones remain in compression along an arched or sloping path. Nor does a flyer hold a cathedral roof up like a column. Its principal role is lateral restraint at a high level, though individual systems can also brace wind effects, carry gutters or participate in a wider network of walls and roofs.

Parts of the system

At the inner end, the head of the flyer bears against the clerestory wall near a vault springing or another critical level. Its curved underside is the intrados; the upper weathering or coping protects masonry and may form a water channel. The flyer descends to a culée above the outer pier.

The outer pier provides mass and directs force toward its foundation. A pinnacle can increase downward load over the pier and improve the stability of the compression path, while also marking the vertical rhythm of the elevation. Pinnacles are therefore neither automatically structural nor merely ornamental: their contribution depends on position, weight and connection. ^{[4] [6]}

Some cathedrals use one flight from high wall to outer pier. Others employ two levels or two stages with an intermediate support. At Notre-Dame de Paris, earlier upper double-flight arrangements were replaced by long single-flight flyers spanning the galleries; later choir work produced another visible campaign. ^[4] Such changes warn against treating an apparently unified exterior as one medieval design moment.

Flying Buttress Evidence: What to Trace

A picturesque exterior is not enough: follow each component, its contact and its construction history.

How the forces travel

A masonry vault pushes downward and laterally at its supports. Interior piers accept much of the vertical load, but the upper wall must also resist outward thrust. A flyer meets that wall, receives compression and sends it diagonally across the aisle to the outer pier. The pier then carries the resultant toward the ground. ^{[6] [9]}

This is not a rigid arrow frozen in stone. Masonry has strong compressive capacity but little tensile capacity. Small movements can open joints and shift the line of thrust through a different sequence of blocks. Stability requires that the compression path remain inside an adequate thickness of masonry. Geometry, self-weight and contact between units matter as much as the visible size of the arch. ^[6]

The paired wall, flyer and pier also respond to wind on the roof and clerestory, thermal movement, settlement and changes in adjacent bays. A timber roof may brace or load the upper structure; a missing roof can remove restraint. Real cathedrals therefore behave as three-dimensional, phased buildings rather than as isolated textbook sections. ^{[6] [10]}

Why flyers permit larger windows

Earlier vaulted churches could resist lateral thrust through thick walls, galleries, adjacent vaults and attached buttresses. The flying system moves concentrated support outside the wall plane at selected bays. Between those points, the clerestory can become thinner and more extensively glazed.

Chartres demonstrates the architectural result: ribbed vaults, piers and exterior support work together so that stained glass replaces much of the enclosing wall. ^[2] At Amiens, flyers are visible through clear clerestory glazing as part of the skeletal structure outside. ^[3] The glass does not carry the vault, and the flyer does not act alone; opening the wall depends on coordinated ribs, wall piers and buttressing.

Chartres Cathedral’s north nave elevation, photographed in 2007. Andreas F. Borchert, Chartres Cathedral Buttresses N 2007 08 31; CC BY-SA 4.0. JPEG prepared; no crop or retouch.

The popular formula “pointed arch plus rib vault plus flying buttress” is useful orientation but poor diagnosis. A pointed profile does not eliminate thrust. Ribs do not collect every force into perfect lines. Flyers may be redundant, altered or positioned for loads no longer present. Read the whole bay before assigning a single cause.

The disputed beginnings

The Metropolitan Museum places the appearance of flying buttresses in the 1170s within northern France’s wider Gothic experiments. ^[1] That is a sound broad chronology, not a secure birthday. Many early structures have been rebuilt, and documentary evidence is sparse.

John James argues from masonry and written records that flying buttresses were built or intended before 1170 at Sens Cathedral, Saint-Lomer at Blois, Voulton and Saint-Remi at Reims. He identifies a further group where the evidence is possible but equivocal. ^[7] The claim shifts attention from a named inventor to connected workshops, construction sequences and fabric.

Cambridge research reinforces the caution. Formal similarity does not prove that two flyers share a date, because medieval-looking masonry may be an addition or restoration. Archaeology is essential, yet even physical evidence can be compromised by replacement. ^[6] The most responsible answer is therefore plural: flyers emerged through twelfth-century experimentation, with individual dates established building by building.

Medieval words and working evidence

Villard de Honnecourt’s thirteenth-century portfolio includes a drawing associated with the flying buttresses of Reims. The image is invaluable evidence of attention to the form, but it is not a measured engineering specification and does not exactly record every surviving stone. ^[13]

A 1316 expertise at Chartres uses the medieval expression arz boterés and reports that the flyers supporting the vaults needed jointing and prompt repair. ^[5] This document does more than supply a name. It shows that maintenance and structural concern accompanied Gothic masonry within living memory of construction.

Drawings, account rolls, photographs and tool marks answer different questions. A sketch may reveal a concept; an account can date purchased lime; bonded joints can show sequence; a photograph records one later state. Agreement among these strands is stronger than any picturesque image by itself.

Notre-Dame de Paris: support and change

Notre-Dame’s exterior demonstrates how structure, drainage and architectural revision overlap. Long single-flight flyers reach across the galleries to the upper walls. Gutters and gargoyles use the same elevated routes to carry rainwater away from vulnerable masonry. ^[4]

The present arrangement is not simply the first scheme completed in the twelfth century. High windows were enlarged, terraces altered and buttressing modified during later campaigns. Choir flyers associated with Pierre de Chelles, Jean Ravy and successors belong to further work in the late thirteenth and fourteenth centuries. ^[4]

A Library of Congress stereograph preserves a historic view of the apse and its radiating supports. ^[13] It is useful for studying the exterior as it appeared to a camera, but it cannot alone distinguish medieval blocks from nineteenth-century restoration. Comparison with measured fabric records is essential.

Nave-side flying buttresses at Notre-Dame de Paris, photographed in 2007. Jean Lemoine, www.flickr.com/people/7761867@N06, Notre Dame buttress; CC BY-SA 2.0. JPEG prepared; no crop or retouch.

Chartres, Bourges and varied solutions

At Chartres, regularly spaced flyers meet the high nave wall while substantial outer piers define the aisle perimeter. Their visibility expresses the structural bay on the exterior. The 1316 report also proves that even a celebrated system required inspection and repair. ^{[2] [5]}

Bourges offers a different spatial problem. Double aisles step upward around a central vessel, and flying supports cross this layered section. The example matters because Gothic builders did not apply one standard diagram to every plan. Heights, aisle roofs, vault springings and available abutments changed the geometry.

Bourges Cathedral’s stepped chevet and double-flight buttresses, photographed in 2007. KoS, Chevet of Bourges Cathedral; Public domain. JPEG prepared; no crop or retouch.

Some flyers have upper and lower arches. An upper member may assist with wind restraint or roof-level behaviour rather than receive the main vault thrust. Its purpose must be tested against the level where it meets the wall, not assumed from parallel appearance. ^[6]

Amiens: refinement and reinforcement

Amiens combines solid nave flyers with openwork examples at the choir. The perforated treatment shows that a flyer could be visually articulated while remaining part of the support system. It also undermines the idea that ever-lighter appearance necessarily means a simple march toward structural perfection. ^[3]

The building changed after distress. A partial vault collapse near 1500 led to reinforcement by an iron chain, and lower members were added to parts of the flying-buttress system. These measures belong to the cathedral’s structural history, not an embarrassment to be edited out. ^[3]

For readers, Amiens supplies a practical test: compare adjacent campaigns. Differences in arch depth, openwork, meeting height and masonry bonding may identify changed design or repair. The most ornate flyer is not automatically the newest, strongest or most original.

Openwork flying buttresses at Amiens Cathedral’s chevet, photographed from the aisle roof in 1980. Acroterion, Amiens Cathedral chevet buttresses 1980; CC BY-SA 4.0. JPEG prepared; no crop or retouch.

Beauvais and the limits of height

Beauvais Cathedral pushed its choir vault to nearly forty-seven metres. In 1284, flying buttresses twisted and broke as part of a collapse of the choir vault. The cathedral’s own history describes wind as a likely trigger, while the apse and timber roof behaved differently from the failed bays. ^[11]

The episode should not be reduced to “bad buttresses.” Exceptional height, slender proportions, wind exposure, construction sequence and the interaction of many members shaped the event. Reconstruction altered the system, and the planned nave was never completed.

Stone flyers and horizontal metal bracing at Beauvais Cathedral, photographed in 2010. Txllxt TxllxT, Flying buttresses and metal bracing at Beauvais Cathedral; CC BY-SA 4.0. JPEG prepared; no crop or retouch.

Beauvais is valuable precisely because it resists technological triumphalism. Gothic masonry could be ambitious, empirical and adaptable, but surviving height does not prove that every original decision was adequate. Failure and reinforcement are evidence about how the building worked.

England and phased buttressing

Flying buttresses are often taught through French cathedrals, yet their use varied across regions. English buildings might conceal flyers beneath aisle roofs, combine them with heavy walls, or add them in later campaigns. A Gothic elevation without conspicuous external arches may still possess effective lateral support.

Research at Salisbury identifies five different flying-buttress systems constructed or added between the main thirteenth-century campaign and the fifteenth century. ^[8] Historic England’s study of the tower and spire connects documentary purchases of lime for new flyers with later buttress construction and changes near the crossing. ^[12]

Flying buttresses above an aisle roof at Salisbury Cathedral, photographed in 2010. Charlesdrakew, Salisbury Cathedral flying buttresses 2; Public domain. JPEG prepared; no crop or retouch.

Salisbury therefore supplies a counterexample to any claim that one cathedral has one buttressing design. Structural need can change when a tower rises, a vault is rebuilt or a roof is altered. Regional appearance also reflects weathering, local stone, aisle height and architectural preference.

Construction and centring

A flyer was built as an arch. Temporary timber centring supported voussoirs until the masonry course closed and compression could develop. Builders then protected the upper surface with sloping coping, dressed joints and integrated the arch with its bearing blocks.

Sequence mattered. The outer pier needed enough weight and stability before the flyer could push against it. The upper wall, adjacent bays and vault construction had to reach compatible stages. Removing centring too early or loading one side without its counteracting system could redirect force through unfinished work.

Evidence survives unevenly. Putlog holes indicate scaffold levels; changes in stone or mortar can identify campaigns; joints reveal whether a flyer is bonded into or merely abuts another mass. Later repointing may conceal these clues, making careful survey more valuable than stylistic guessing.

What structural studies can and cannot prove

Limit analysis tests whether a plausible compression line can fit within a masonry form under stated loads. The Mallorca research defines the flyer’s task as taking central-vault thrust over the aisle toward external buttressing, then examines possible mechanisms rather than assuming effectiveness from shape. ^[9]

At Tournai, engineers considered flying buttresses together with outer buttresses and unusually slender choir piers because defects had concerned observers since early in the building’s life. ^[10] The coupled analysis is the important lesson: strengthening one member without understanding the rest may simply move a problem.

Models depend on assumptions about geometry, loads, cracks and support movement. They do not recover a medieval designer’s intention automatically. Survey, archaeology, documentary history and monitoring must constrain the model; the model can then test explanations that visual inspection alone cannot settle.

Reading the force-path diagram

D23 begins at the high vault springing. Downward load continues into the interior pier, while a lateral component reaches the clerestory wall. The flyer receives that component in compression and carries it above the aisle roof to the outer culée and pier. The foundation spreads the combined force into the ground.

Solid arrows mark the principal explanatory path, while dashed arrows identify variable lateral actions. The pinnacle is labelled as counterweight, not as an independent cure. A water line along the coping shows that drainage can share the assembly without defining its structural purpose.

The comparison panel shows an attached buttress thickening the wall continuously. Use three checks in a real building: locate where support meets the upper wall, identify the gap it crosses, and trace the receiving mass to ground. If any part is hidden by roofs or chapels, treat the conclusion as provisional.

A flyer carries lateral force above the aisle to an outer pier. The real system also includes the vault, high wall, roof, pinnacle and foundations.