The Unfinished Instruction
On July 20, 1936, members of the Iberian Anarchist Federation entered the Sagrada Família and destroyed what they could find. Not the building itself — they tried to dynamite the Nativity Façade but either failed or were talked down. What they destroyed was the information: Gaudí's plaster models, his drawings, his photographs, his letters. They burned the workshop. They smashed the models into fragments.
Antoni Gaudí had been dead for ten years by then, hit by a tram in 1926. But his instructions were still alive, maintained by disciples who understood his methods. After the anarchists left, it was the instructions that were gone, and the building that remained.
The Sagrada Família opens in 2026 — 144 years after construction began, 100 years after Gaudí's death, 90 years after the destruction of his models. This is a story about what happens to intent when it survives only as fragments.
Gaudí didn't work the way other architects worked. He distrusted drawings. Instead, he built physical models — most famously, the hanging chain models for the Colònia Güell church, where weighted strings suspended from a board would settle into catenary curves under gravity. Flip the model upside down and you have arches that distribute load perfectly, without buttresses.
This was parametric design a century before computers. Move one anchor point and the entire network of chains recalculates. The physics is the computation. Gaudí wasn't drawing a building and then figuring out how to hold it up. He was letting the forces draw the building themselves.
For the Sagrada Família specifically, he moved beyond catenaries to ruled surfaces — hyperboloids, paraboloids, helicoids — complex geometries that could be described by sweeping a straight line through space. The genius of this: every curved surface is actually constructible from straight elements. The organic, almost biological appearance of the interior emerges from strict geometric rules.
He spent the last twelve years of his life working exclusively on the Sagrada Família, living in the workshop on-site, obsessively refining models at 1:10 and 1:25 scale. When asked about the slow pace, he supposedly said: "My client is not in a hurry." He meant God, but the statement implies something else — that the building's purpose was not to be finished by him. He built it knowing others would continue.
What the anarchists destroyed was not a conventional blueprint. You can copy a blueprint. What they destroyed was a method — embodied in plaster shapes that encoded geometric relationships no one had yet fully translated into notation. Gaudí's models were, in a sense, physical programs: they contained the logic of the building in a form that required interpretation, not just execution.
After the war, architect Francesc de Paula Quintana led the reconstruction effort. He and his team gathered fragments — shattered plaster, broken molds — and began reassembling what they could. This was not, as the Sagrada Família museum notes, "just gluing fragments." It was "an act of profound interpretation and educated guesswork," guided by surviving photographs, partial drawings, and their understanding of Gaudí's architectural language.
The key word is language. They weren't recovering a fixed design. They were recovering the principles that generated designs. Because Gaudí's geometry was mathematically principled — ruled surfaces, conic sections, specific families of curves — you could sometimes reverse-engineer the whole from a fragment. If you know the function class, you can recover the function from sample points.
This is what New Zealand architect Mark Burry discovered when he arrived in 1979. Two elderly architects, Isidre Puig Boada and Lluís Bonet i Garí, who had worked with Gaudí's disciples, showed him something crucial: "triple points," locations where three surfaces meet, contain enough geometric information to reconstruct the surfaces themselves. The fragments weren't random debris. They were data-rich.
"For every one of these surfaces," Burry explained, "there are nine variables that govern where that triple point will end up." He began treating Gaudí's shapes "as geographical contours in relief" — reading topology, not iconography. The geometry could be decoded because it was principled, not arbitrary.
But here's where it gets philosophically interesting. There is a difference between reconstructing what Gaudí designed and continuing what he intended. The first is archaeology. The second is interpretation. And for the parts of the building where no models survived at all, only the second is possible.
The Glory Façade — the final major element, the main entrance facing the sea — exemplifies this perfectly. Gaudí made one model for it. That model was smashed. Burry's team didn't just try to reassemble the fragments. They tried to "extrapolate what he would have learnt from that model and imagine how you could clean it up."
Read that again. They aren't asking: what did the model look like? They're asking: what would Gaudí have done next? The model was a draft. Gaudí iterated obsessively. So the faithful response isn't to reconstruct the draft — it's to simulate the iteration that would have followed.
This is extraordinary. It requires understanding not just Gaudí's outputs but his process — the direction of his refinement, the kinds of problems he would have noticed, the solutions his geometric vocabulary would have suggested. You have to model the modeler.
Not everyone agrees this is legitimate. In 2008, a group of prominent Catalan architects called for construction to halt, arguing that the surviving fragments were too incomplete to justify continuation. The building being constructed, they said, is not Gaudí's building. Burry himself called the Glory Façade design "quite polemical" and remains uncertain whether it will be built as he conceived it.
Current chief architect Jordi Faulí frames the work differently: the goal is to "collaborate to build the Gaudí idea." Not the Gaudí design — the Gaudí idea. The distinction matters. An idea can be realized in multiple ways. A design has one correct form.
Technology changed the question, but not in the way you might expect. When computer-aided design arrived at the Sagrada Família in the 1980s, it didn't replace Gaudí's methods. It completed them.
Gaudí's hanging chain models were analog computers — physical systems that solved optimization problems through the forces of gravity. His ruled surfaces were mathematical objects that happened to be constructible from straight lines. Both approaches were, in essence, computational. They used the physical world or mathematical structure to do the design work.
Digital tools are the natural continuation of this philosophy, not a departure from it. CAD software can compute the same geometric relationships Gaudí's models embodied, but with precision and speed that plaster and string couldn't achieve. CNC machining can cut stone along ruled surfaces with millimeter accuracy. The technology doesn't betray Gaudí's vision — it realizes capacities his methods implied but couldn't fully execute.
The Arup engineers who joined the project in 2014 introduced pre-stressed stone masonry panels — a technique Gaudí never used. But the purpose is structural: the remaining towers need to be lighter than solid stone would allow. "Innovation is something you do in design but not in construction," Burry observed, reflecting Gaudí's own conservatism about building methods. The innovation serves the original geometric vision.
There's an irony here. The most "faithful" interpretation of Gaudí's work requires the most advanced technology, because his designs exceeded what his era's construction methods could achieve. The building is more Gaudí now — closer to what his geometry described — than it could have been in his lifetime.
On October 31, 2025, a section of the Tower of Jesus Christ was lifted into place, bringing the Sagrada Família to 162.91 meters — surpassing Ulm Minster to become the world's tallest church. The tower will reach 172.5 meters when complete, crowned by a four-armed cross of glass and white enamel. Pope Leo XIV has been invited to a solemn mass on June 10, 2026, the centenary of Gaudí's death.
It will not technically be finished. The Glory staircase and some urban features won't be complete until the 2030s, partly because residential buildings stand where Gaudí planned grand entrance steps — a conflict between one architect's 19th-century vision and the city that grew around it.
But the structure will be done. Eighteen towers for twelve apostles, four evangelists, the Virgin Mary, and Jesus. Stone from Montserrat mountain and, since 2018, from a quarry in Lancashire, England. Built by workers who never met the architect, following principles that survived destruction, using tools that didn't exist when the design was conceived.
Gaudí said, or is said to have said, that the building would take time because "my client is not in a hurry." Meaning: the building would be completed by whoever came after. The instructions didn't need to be complete. They needed to be generative — to contain enough of the logic that others could continue.
Whether they've done so faithfully is a question the building can't answer. It's not the building Gaudí would have built. But it may be the building Gaudí's geometry was always trying to become.