Béton Brut

Brutalism has been hated for two reasons that are actually different reasons.

The buildings are falling apart because France and Britain couldn’t pour concrete. The buildings are ugly because the 20C couldn’t draw. Those are different failures with different villains, and conflating them lets every one of the villains off the hook. Ando and Zumthor and Kahn solved both problems with the same material and the same aesthetic. The theory was always available. The culture lost its nerve.

The standard account goes like this. Postwar architects, drunk on socialist optimism and Corbusian theory, inflicted cold gray monoliths on unwilling populations. The buildings turned out to be unlivable and ugly. The public revolted. The style died. Good riddance.

Prince Charles got a career out of saying this in a plummy voice. Every British broadsheet recycled it through the 80s and 90s. By the demolition of Robin Hood Gardens in 2017 the narrative was so settled that “brutalist” had become a pejorative for any large concrete building regardless of when or why it was built.

It’s wrong. Or more precisely, it’s compressing two different failures into one and then misdiagnosing both.

Failure one: the buildings are falling apart. The concrete is spalling, the rebar is exploding, the facades are streaked with rust. This is a trades and maintenance failure. It has nothing to do with the aesthetic.

Failure two: a lot of the buildings are, in fact, ugly. Not because concrete is ugly, but because 20th century architectural culture produced an enormous volume of genuinely bad buildings across every material and every style. This is an aesthetic failure. It has nothing to do with the concrete.

The two get conflated because the spalled rust-streaked facade and the badly proportioned building are often the same building, and the public (reasonably) reads the combined impression as “brutalism looks like this.” The fixes are completely different, the responsible parties are completely different, and pretending the two failures are the same failure lets everyone involved off the hook.

Start with the concrete. Come back to the aesthetic.

Corb had no choice, and then everyone copied the accident

Start with Corb, because everyone else was copying him. The Unité d’Habitation in Marseille (1952) is the origin document. Le Corbusier specified raw concrete (béton brut, literally “raw concrete”) as the exterior finish and wrote extensively about honesty of material, expression of structure, the beauty of the trace of the formwork. He meant it. He also had no choice. Post-war French construction was operating on bad aggregate, inconsistent cement supply, crews who’d spent the previous decade either at war or building fortifications, and formwork cobbled from whatever lumber hadn’t been burned for heat. The Unité’s walls are pitted, voided, streaked, and visibly irregular because the pour was uncontrollable. Corb’s move (and this is a genuine theoretical move, not an accident) was to reframe the failure as a virtue. Board-marked texture. Honest expression. The hand of the worker.

Then everyone copied the *look* without understanding what they were looking at.

British brutalism inherited both Corb’s aesthetic and Corb’s concrete problems. The Barbican. Trellick Tower. Robin Hood Gardens. Park Hill in Sheffield. The Hayward Gallery. Most of the LCC estate work. Same bad mix designs. Same inexperienced crews. Same theological commitment to leaving the material “honest,” which in practice meant unsealed. Add a maritime climate with freeze-thaw cycles eight months of the year and you have written the specification for spalling.

What unsealed reinforced concrete actually does

Water enters the surface through capillary action and microcracks. Every concrete pour has microcracks. They’re inherent to the curing process. The water reaches the reinforcing steel. Steel in contact with water and oxygen rusts. Rust occupies roughly six to seven times the volume of the steel it came from. That expansion cracks the concrete from the inside out. The cracks expose more rebar. The process accelerates.

Meanwhile atmospheric carbon reacts with calcium hydroxide in the cement paste to form calcium carbonate, a process called carbonation, which drops the pH of the concrete from around 13 to around 9. At pH 13 the steel is chemically passivated and doesn’t rust even in the presence of water. At pH 9 the passivation layer is gone. Carbonation proceeds at roughly 1mm per year in typical urban air. Rebar cover in a lot of postwar British concrete was 20 to 25mm when it was supposed to be 40 to 50mm, because the crews couldn’t be bothered to place the spacers correctly.

Do the math.

The buildings were structurally compromised within thirty years of their pours. The public saw rust streaks, spalled facades, chunks of concrete on the pavement, and correctly concluded that something was wrong. They were told by the press and by a generation of post-modernist architects that what was wrong was the style. What was actually wrong was that nobody had sealed the concrete, nobody had placed the rebar correctly, and the mix design hadn’t accounted for the climate.

The Japanese got it right

Tadao Ando is working with the same material Corb used. Cement, aggregate, water, rebar. The aesthetic goals are nearly identical: raw concrete as finish, expression of the formwork, no cladding, no paint, no ornamentation. The outcomes are not identical at all. The Church of the Light (1989), the Row House in Sumiyoshi (1976), the Naoshima projects. All of them still look approximately like they did when the forms came off. Thirty, forty, nearly fifty years in.

The difference is not talent. The difference is that Japanese concrete trades treat the pour as a precision operation. Specific aggregate sourced from specific quarries. Water-cement ratio measured, not estimated. Slump tested at the truck. Formwork built from sealed plywood with taped joints so the paste doesn’t bleed through and leave streaks. Vibration on a schedule, not when the foreman remembers. Pour rate controlled so you don’t get cold joints between lifts. A penetrating silane or siloxane sealer applied within weeks of form stripping, reapplied every seven to ten years for the life of the building.

Ando’s concrete is sealed. It has always been sealed. The sealing is invisible (silanes and siloxanes penetrate the surface and bond chemically, they don’t form a film) and it’s what allows the “raw” aesthetic to actually survive as raw rather than becoming a ruin.

American outcomes split on exactly this axis. Paul Rudolph’s Yale Art&Architecture Building (1963) is corduroy-textured concrete, hammered after curing to expose the aggregate, never systematically sealed, and has been a preservation headache since the Carter administration. Louis Kahn’s Salk Institute (1965), poured with teak formwork on a cliff above the Pacific, is maintained by a foundation that actually budgets for resealing, and looks spectacular sixty years in. Same decade. Same country. Same material. Different maintenance cultures. Different outcomes.

Peter Zumthor’s Therme Vals (1996), in the Swiss Alps, is the European counter-example that proves the rule. Swiss trades. Properly specified. Sealed. Still perfect.

The fix has been available the whole time

Penetrating silane sealers entered the commercial market in the 1960s. They cost between $2 and $4 per square foot applied in current dollars. They’re invisible. They last seven to ten years between reapplications. They prevent something like 90% of the water ingress that drives spalling. Every state DOT in North America uses them on bridge decks, because bridge engineers have done the rehabilitation math and concluded that sealing is cheaper by an order of magnitude than letting the deck go and patching it.

British architectural practice of the brutalist era treated sealing as a betrayal. The material was supposed to weather honestly. Sealing it was cheating, was dishonest, was bourgeois, was a refusal of the theoretical commitment. This is a theological argument, not an engineering one. It has the structure of a religious objection to blood transfusion: the principle is clear, the consequences are predictable, and the adherents would rather suffer the consequences than violate the principle.

You can see exactly the same thinking in current exposed-mechanical discourse. “Authentic” slides into “unmaintained” and then into “failing on a schedule.” The aesthetic commitment becomes a prohibition on the maintenance the aesthetic requires to survive.

The trowel-and-vibrate problem is the other half of the failure, and it’s also been solved. Self-consolidating concrete (SCC) has been commercially available since the 1990s. SCC flows into formwork under its own weight with minimal vibration, which means you can get consistent architectural finishes without depending on a vibration crew that may or may not show up sober. European precast yards adopted SCC readily because precast is a factory operation and the economics are obvious. European cast-in-place trades, particularly in the UK and France, have dragged their feet for decades. Partly because the unions are structured around skilled manual labor that SCC partially obsoletes. Partly because the culture of those trades treats new materials with suspicion bordering on hostility. You still get cold joints and honeycomb on new civic projects that have absolutely no excuse for it in 2026.

Three piles

The brutalist inventory sorts into three piles.

Pile one. Buildings by architects who worked with competent trades and specified proper maintenance. These look fine. Ando, Kahn, Zumthor, late Rudolph when he had the budget, most of the Japanese inventory, the Salk, the better German work. The aesthetic is vindicated. The theory was correct.

Pile two. Buildings by architects who had the theory but not the trades, and whose clients didn’t budget for sealing. This is most of the UK inventory and a lot of the French. The buildings are failing. They were always going to fail. The failure is a construction and maintenance failure, not an aesthetic failure, but the public reads the stained spalled facade as “brutalism looks like this” rather than “concrete that wasn’t sealed looks like this after forty British winters.”

Pile three. Buildings that were demolished before the failure could be diagnosed properly, which means the demolition itself became the evidence that the style was unworkable. Robin Hood Gardens is the canonical example. Smithson’s design was not the problem. The problem was a housing authority that deferred maintenance for thirty years while telling residents the building was the reason their lives were hard, and a political consensus that found it easier to blow up the evidence than to admit that social housing had been deliberately defunded into dysfunction. The building came down in 2017. A section of it is in the V&A now, as if it were an archaeological artifact rather than something we chose to destroy last Tuesday.

The other thing, which is that a lot of them are actually ugly

Separate question. Keep it separate.

A building can be structurally sound, properly sealed, beautifully detailed, and still ugly. A lot of brutalist buildings are ugly, and a lot of modernist buildings are ugly, and a lot of postmodernist buildings are ugly, because 20th century architectural culture produced an enormous volume of genuinely bad buildings. Something was wrong, and the something was not confined to brutalism, and pretending it was is the move that lets the rest of 20th century architecture off the hook.

My read on what went wrong, in descending order of confidence.

The profession lost proportion. Not as a skill (some people still had it) but as a shared commitment. Classical architecture had roughly two thousand years of accumulated work on what ratios of height to width, window to wall, base to shaft to cap, read as right to the human eye. That body of knowledge was not mysterious. It was taught, it was refined, it was the thing a draftsman absorbed in the first year of training. The modernist break threw it out on the theory that proportion was arbitrary cultural baggage and that new materials demanded new rules. New rules never arrived. What arrived was a generation of buildings where nothing is talking to anything else and the eye has nowhere to rest.

The profession lost ornament and didn’t replace it with anything. Adolf Loos published Ornament and Crime in 1913 and the argument metastasized for the next sixty years into a blanket prohibition on decoration. The stated reason was that ornament was bourgeois, dishonest, a distraction from structure. The actual function of ornament was to give the eye something to do at close range, to provide a gradient of detail between the overall massing and the individual brick, to signal that a human hand had been near the work. Stripping it out didn’t produce honest architecture. It produced blank walls that the eye slides off. The late-century attempt to put ornament back (postmodernism, Graves and Venturi) failed because it was ironic rather than felt, and nobody wants to live inside a joke.

The profession got captured by photography. Buildings started getting designed to look good in a single black-and-white hero shot in Architectural Review, taken at the right angle with the right lens at the right time of day, with no people in it. The experience of the building at street level, in rain, at eight in the morning, when you’re late for something, became a secondary consideration. This is not a brutalism problem specifically. It’s a 20th century problem generally. But brutalism is where it becomes most visible, because the buildings are big and the hero shot and the street-level experience are maximally divergent.

The profession adopted an ideology where comfort was suspect. The Bauhaus inheritance, transmitted through the Ivy League architecture schools after the European émigrés arrived in the 1930s and 40s, treated making a building pleasant to be in as a kind of aesthetic cowardice. The good building was supposed to challenge you, to refuse easy pleasure, to make you think. Applied to a museum or a monument this is a defensible position. Applied to a housing estate or a university library it is a category error, and the category error ran the discipline for fifty years.

The profession served the client it had, and the client it had was the postwar state. Social housing, civic centers, universities, hospitals. The client wanted maximum square footage for minimum cost, produced quickly, on sites that were typically either slum clearance or bomb damage. The architectural response that worked inside those constraints was large, repetitive, concrete, and lightly detailed. The response that didn’t work inside those constraints was anything with hand-crafted materials or load-bearing masonry or ornament or proportional refinement. The buildings look the way they look partly because the brief selected for it. But a better profession would have resisted the brief harder, or found ways to make the cheap building beautiful the way Scandinavian architects often did in the same period with the same budgets.

Put that all together and you get the actual indictment. 20th century architecture, taken as a body of work, is probably the worst built inventory any century has produced since humans figured out how to stack stones. Not because the materials were bad. Because the culture lost its nerve, lost its proportion, lost its ornament, and replaced all three with bad theory.

Brutalism is inside that indictment. Not uniquely. Alongside everything else built between roughly 1925 and 1995. The buildings that survive the critique (Ando, Kahn, Zumthor, Utzon, Scarpa, Aalto, Barragán) survive it by not having abandoned proportion and by not treating pleasure as suspect. They’re a small minority of the built inventory of their century, and they are the reason we know the problem was cultural rather than material.

The trades failure and the aesthetic failure are different failures. The trades failure is fixable with money, specifications, close supervision, and maintenance budgets. The aesthetic failure is fixable only with the slow recovery of a profession that currently does not know what it lost. That recovery is underway in some places, stalled in others, and actively resisted by the architectural academy in most places, which still treats “proportion” and “ornament” as words for the history section rather than the design studio.

Boston City Hall: the failures compound

Boston City Hall is the ur-example of both failure modes at once, plus a third one on top.

Kallmann, McKinnell and Knowles won the competition in 1962 and the building opened in 1968. The design is actually coherent as architecture. The three-part section (heavy civic functions up top, ceremonial floors in the middle expressed as projecting concrete volumes, public functions at the bottom) is a legible translation of Le Corbusier’s La Tourette into a civic program. As a drawing it works. As a model it works. As a building in Boston it does not work, and the reasons are instructive.

The weather problem first. Boston gets roughly 110cm of precipitation a year, distributed across rain, snow, freezing rain, and the specifically destructive category of water that falls as rain, pools on horizontal surfaces, and then freezes overnight. The City Hall design has enormous horizontal concrete surfaces: the projecting floor slabs, the top of the upper volume, the plinths, the plaza edges. Every one of those is a place where water sits, soaks into unsealed concrete, and then expands as ice. The cantilevered elements in particular catch wind-driven rain on their undersides, which is the worst possible exposure because water wicks upward into the slab against gravity and has nowhere to drain. The rust-staining on the undersides of those cantilevers was visible within a decade and has been progressive ever since.

The architects knew Boston’s climate. They designed as if the climate were Marseille’s. This is not a small oversight. Corb at La Tourette was working in the Rhône valley, which is dry, mild, and forgiving. Translating that geometry directly to coastal New England means every horizontal plane becomes a water-collection feature and every cantilever becomes an ice-expansion test. The building is fighting the weather with its own shape, and the shape loses, because physics.

The circulation problem is worse because it’s the more visible failure and the one the public actually experiences. City Hall Plaza is something like 4.5 hectares of brick pavers across an almost featureless slope, designed as a grand civic gathering space on the model of the Piazza del Campo in Siena. Sienese piazzas work because they’re small, enclosed, shaded, and sited at the convergence of medieval street patterns that funnel pedestrians through them as a matter of course. You cannot avoid walking through the Campo if you’re going anywhere in Siena. Boston’s plaza is none of those things. It’s enormous, exposed, unshaded, uphill from the subway, and sited such that pedestrians have every reason to go around it rather than through it. The prevailing winds in winter turn it into a venturi. The brick paving in summer turns it into a radiant heater. There is nowhere to sit that isn’t deliberately austere concrete. The scale is calibrated for rendered figures in an architectural drawing, not for actual human beings who want a coffee and a bench.

Jane Jacobs had published The Death and Life of Great American Cities seven years before City Hall opened. Everything she wrote about eyes on the street, short blocks, mixed use, and the failure of superblock planning was available to the designers. They either didn’t read it or didn’t believe it applied. The plaza was designed as an abstract civic gesture, a place where the citizenry would gather for, presumably, civic purposes, as if civic purposes manifested spontaneously given enough open space. They do not. Public space that isn’t programmed and scaled for actual use becomes hostile space, and hostile space gets abandoned, and abandoned space attracts the uses that thrive on abandonment. Boston spent the next fifty years trying to retrofit humanity into the plaza with farmers markets, ice rinks, and seasonal installations, which is the civic equivalent of putting throw pillows on a concrete bunker.

City Hall fails three times. The concrete is losing to the weather because the geometry was wrong for the climate and the sealing was inadequate for either climate. The plaza is losing to the humans because the circulation was designed for an abstract citizen rather than actual Bostonians in February. And the massing is losing to the eye because the building has no proportional relationship to anything around it and no close-range detail to reward the viewer who gets within ten meters of it. All three failures are blamed on “brutalism.” None of them is actually a property of raw concrete as a material. They’re properties of specific design decisions that could have gone differently, and did go differently in places like Scandinavia where civic buildings of the same era were detailed for the weather they actually faced and scaled for people who actually had to use them.

The cruel part is that the building itself, considered as a diagram, is coherent. Considered as a mass in the sky at a distance, it has presence. It’s the middle distances where it collapses. And middle distances are where buildings actually get experienced.

Why the wrong narrative won

None of this is secret. The concrete literature has been clear on sealing and mix design for fifty years. The preservation community knows exactly what went wrong with the 1960s pours. The Japanese results are sitting right there, available for inspection. Ando is not doing anything Corb couldn’t have specified. The proportional and ornamental failures are equally well-documented, by Scruton and Alexander and a half-dozen others who got ignored because the academy didn’t want to hear it.

The narrative that brutalism failed because it was ugly, full stop, is more useful than the accurate narrative because the accurate narrative names too many villains. The architects failed at proportion and refused ornament and got captured by the hero shot. The builders poured badly and skipped the sealer. The politicians cut maintenance budgets for forty years. The academy taught the students that pleasing the eye was a form of intellectual cowardice. Every one of those parties benefits from the story where the buildings are just ugly and the public just didn’t get it, because “ugly” is a verdict that names no one.

The material is fine. The specific buildings that work, work. Ando, Kahn, Zumthor, the Salk, Therme Vals. Concrete poured properly, sealed on schedule, proportioned by someone who still had the skill, detailed at human range, maintained for the life of the building: it does what Corb said it would do. It expresses its structure, it ages with dignity, it accumulates a patina rather than a pathology.

What doesn’t work is three different failures stacked on top of each other. The trades couldn’t pour. The architects couldn’t proportion. The clients wouldn’t maintain. Fix any one and you get a better building. Fix none and you get the Barbican in February.

Béton brut is fine. The Euro trades just can’t pour concrete. And the 20th century just couldn’t draw.