Stand on almost any main street in the world and look up. The older buildings have decoration: cornices, carved lintels, columns that hold nothing up. The newer ones are smooth boxes of glass, concrete or metal, with windows in long strips and a flat roof. Somebody decided, about a hundred years ago, that this was how buildings should look. They did not decide it out of taste alone. A new way of building made it possible, and an argument about what buildings are for made it seem right.
Now look at a few more. A school in a Burkina Faso village, made of clay bricks and with a roof that shades the wall. A housing estate in India with courtyards instead of corridors. A titanium museum that seems to ripple. All are called contemporary, and they disagree with one another and with the glass box. A graduate of this course walks past a building and can say roughly when it was made, what it was arguing against, and what technology let it exist.
Through the nineteenth century a building's face was mostly decoration borrowed from earlier ages: Greek temples for banks, Gothic spires for universities. Around 1900 a group of architects began to say this was dishonest. In 1896 the American Louis Sullivan wrote that "form ever follows function" in an essay on the tall office building. In 1910 the Viennese architect Adolf Loos gave a lecture called Ornament and Crime (later misdated to 1908), arguing that decoration wastes labour and belongs to an earlier stage of civilisation. (Loos was provocative on purpose, and his argument has plenty of critics.) In 1919 Walter Gropius founded the Bauhaus school in Weimar, Germany, to train designers who would unite art and industry.
Their shared claim: design from what the building does and how it is made, with factory materials, and let the result look like itself. The most famous summary is by Le Corbusier. In 1927 he listed five points of a new architecture, and built them into the Villa Savoye near Paris, finished in 1931. Switch them on below.
A cartoon, not the Villa Savoye. Top: elevation (the view from the street). Bottom: plan (the floor seen from above, with walls as thick lines and columns as dots).
Each point is a freedom that a concrete frame gave. Pilotis are slender columns that lift the house off the ground. A roof garden replaces the pitched roof that sheds rain, which became practical with waterproof membranes. The free plan means the inner walls no longer carry the floors, so they can go anywhere. The ribbon window runs the length of the wall because the wall carries nothing. The free facade hangs the outer skin in front of the columns like a curtain. Notice the pattern: all five follow from one change, moving the load from the walls to the columns. The next section shows why that change mattered so much.
For most of history a tall building was a stack of walls. The wall carried its own weight, the floors and the roof, and the higher you went the thicker the lower wall had to be. This is where the wall's own weight stops you. A brick wall's weight presses on its base in proportion to its height, no matter how thick it is. Add the floors and there is a ceiling, a height where no thickness is enough. The Monadnock Building in Chicago (north half 1889 to 1891) is usually called the tallest load-bearing masonry building ever built, around 16 storeys. Its walls are reported to be about 1.8 m (six feet) thick at the base. Its southern half, finished in 1893, used a steel frame instead.
Simplified and illustrative: a uniform wall (real ones taper), 3.9 m storeys, brick 19 kN/m³, 15 kN per metre of wall per storey (a 6 m floor span carrying about 5 kPa), steel column every 6 m at 165 MPa. Ignores wind, openings, and buckling. Real design codes differ by country.
In this simplified model, 12 storeys need a wall about 0.3 m thick, 16 storeys about 0.76 m, 18 storeys about 1.6 m and 20 storeys nearly 17 m. At 21 the wall's own weight has used up all the strength and no thickness works. The model is optimistic: the real Monadnock's walls were about 1.8 m thick at 16 storeys, because real designers allowed for wind, openings and safety margins. But the shape of the curve is right, and so is the lesson. The steel column that replaces the wall at 16 storeys needs only about 175 cm² of steel in this crude model, a solid square 13 cm on a side; a real one is bigger, but still far slimmer than the wall. That is the real reason the skyscraper appeared in Chicago and New York in the 1880s and 1890s, and why Sullivan could write about the tall office building at all. A frame also lets a floor be mostly window.
A tall building is a vertical beam sticking out of the ground, and the wind is the load. The hard part is not strength but sway: people feel it, windows crack, partitions tear. A cantilever beam's tip moves in proportion to the load and to the fourth power of its height, and in inverse proportion to its stiffness, which is E times I, where I measures how far the material sits from the centre.
That last factor explains the most important tower idea of the postwar era. If you pull the same steel out of a central core and spread it to the outer wall, I grows with the square of how far from the middle you put it. The engineer Fazlur Khan, working at Skidmore, Owings & Merrill in Chicago, turned that into the framed tube (the DeWitt-Chestnut apartments, mid-1960s), the trussed tube (the John Hancock Center, completed about 1969) and the bundled tube (the Sears Tower, 1973, now the Willis Tower).
Illustrative model: 40 m square plan, wind 1.5 kPa on the face, steel E = 200 GPa, thin-walled square section so I = A·s² ÷ 6, with the core 10 m across and the tube 40 m across. Real tubes lose some stiffness to "shear lag", so treat the gain as an upper bound. The H ÷ 500 limit is a common rule of thumb, not a code value.
At 200 m with 0.6 m² of steel, the 10 m core would sway about six metres, which is unthinkable. The same steel in a 40 m tube sways about 0.375 m, inside the rule-of-thumb limit of H/500 = 0.4 m. The ratio is 16 to 1, just the square of the distance ratio, and it holds at every height. You can feel this idea in the Burj Khalifa (2010, 828 m), which uses a buttressed concrete core with three wings instead of a tube, but with the same logic: put the stiffness far from the middle.
The usual story has the movement flowing from Weimar and Paris to the United States, with Mies van der Rohe's Seagram Building (New York, 1958) as the glass box that everyone copied. The fuller story is that other places took the ideas and rewrote them for their own climate and politics.
A small, personal sample, not a canon: choose a region and see who is missing. Dates are for completion or opening unless the card says otherwise.
Look at the pattern in who is left out when you filter. Europe and the Americas are crowded; whole regions have a handful of dots. That is partly history and partly what gets written down. Teachers now try to correct it, and the phrase to look for in a syllabus is "global modernisms".
Le Corbusier's 1920s city plans proposed towers standing in open parkland, separated from the street. After 1945 governments around the world used a version of it for mass housing. The result was often the loss of the street: shops, front doors, and the sight of neighbours. Jane Jacobs attacked it in The Death and Life of Great American Cities (1961), and architects including Robert Venturi turned on the style itself. His 1966 book Complexity and Contradiction in Architecture answered Mies's "less is more" with "less is a bore". The critic Charles Jencks dated the death of modern architecture to the demolition of the Pruitt-Igoe housing blocks in St. Louis, which began in 1972.
That story is simple, and historians now dispute it. Pruitt-Igoe's decline had causes that were not architectural: falling city population, cuts in maintenance money, segregation and poverty. The documentary The Pruitt-Igoe Myth (2011) argues the building was blamed for failures of policy. The honest summary: the design did not help, and it was not the only cause. Where the street went, you can measure.
Illustrative geometry. The enclosure labels are rules of thumb from urban designers, not standards; the examples are round numbers, not measurements of a real street. Winter noon sun is for 51°N (about London).
The height-to-width ratio is the first number an urban designer reads off a section. Around 1:1 a street feels like an outdoor room; below about 1:4 it dissolves. The tower in the park scores 0.5 against the old boulevard's 1.0, with open ground where the street wall used to be, and that is precisely what the tower was built to do. Whether that is a feature or a flaw is the argument of the second half of the twentieth century.
Postmodernism (1970s to 1980s) brought back colour, symbols and quotation. The Centre Pompidou in Paris (1977) showed its pipes on the outside. Then the computer changed what could be built. Frank Gehry's Guggenheim Museum in Bilbao (1997) used aerospace software to describe, cut and fit a skin of titanium curves, and cities began commissioning "icon" museums hoping for the "Bilbao effect". Critics say this made architecture a brand. Defenders say it showed that a good building can restore a city's confidence. Both are partly right.
A second direction pulled the other way. Since the 1990s the climate crisis has made energy and carbon a design problem, and this is the subject of Environmental Systems in Buildings (34/94). A third, represented by Doshi and Kéré, makes the community and the local material the starting point, and measures success by whether the building is used and maintained. The Pritzker jury's choices of Doshi in 2018 and Kéré in 2022 are signals of that shift.
These threads run side by side today: the parametric skin, the net-zero box, the earth-brick school, the careful reuse of an old building. A graduate does not need to pick one. They learn to name what each assumes.
Modern architecture is what happens when the frame carries the load: the wall is free, the plan is free, the tower rises. It spread worldwide and was rewritten wherever it landed. A generation then rejected its cities, though the story of why is more complicated than one demolition. Contemporary work argues in several directions at once.
Try this: pick a building near you and write three things on a card: the year, the structure that lets it stand (wall, frame, shell), and one thing it is arguing against. Then check the date.
Sources. Sullivan, "The Tall Office Building Artistically Considered" (1896); Loos, Ornament and Crime (first given as a lecture in Vienna on 21 January 1910; the 1908 date often printed was added at a 1931 reprint: Christopher Long, ACSA proceedings); Bauhaus founded in Weimar in 1919; Le Corbusier's five points (1927) and Villa Savoye (1931): standard histories (Frampton; Curtis). Monadnock Building: north half 1889 to 1891, south half 1891 to 1893, steel frame in the south half, walls about six feet (1.8 m) thick at the base per Wikipedia, Structurae and the Chicago Landmarks listing; sources differ on height (about 60 to 66 m). Fazlur Khan: tube structures, with DeWitt-Chestnut dated variously 1963 to 1966 by different sources, John Hancock Center completed 1969 or 1970, Sears Tower 1973 (Lehigh University and other biographies). Kéré: Gando school completed 2001, Aga Khan Award 2004, Pritzker 2022, first African laureate (Pritzker Prize announcement). Doshi: Pritzker 2018, first Indian laureate; Sangath studio about 1980; Aranya, Indore, 1989 (architecture press coverage; sources vary on the number of homes). Pruitt-Igoe: demolition began 1972; Jencks dated "the death of modern architecture" to it in 1977; a 2022 Dezeen essay and the film The Pruitt-Igoe Myth (2011) contest the date and the cause. Dates for Barragán (1948; Pritzker 1980), Brasília (1960), Yoyogi (1964), MASP (1968), Sydney Opera House (1973), Pompidou (1977), Bilbao (1997), Burj Khalifa (2010, 828 m), Asmara (UNESCO 2017) and Venturi (1966) are standard history; see the fact-check log. Every number in the interactives was computed in node: at 12 storeys the model wall is 0.30 m, at 16 storeys 0.76 m, at 18 storeys 1.6 m, at 20 storeys 16.7 m, and impossible from 21 (1.5 MPa allowable); the tube versus core sway at 200 m and 0.6 m² is 0.375 m versus 6.0 m (a 16 to 1 ratio); the street ratio for the boulevard is 1.0 and for the tower in the park 0.5 (60 ÷ 120); winter-noon sun altitude at 51°N is 15.56°.