The forgotten formula for lovable cities

Everyone loves the streets of old cities like Verona, Bruges, and Prague. The reason might seem obvious: they are lined by beautiful historic buildings. But people also love the streets of Osaka, Taipei, and Hong Kong. While the buildings of the latter cities are mostly modern and nondescript, their streets are still full of character and enchantment. 

The Austrian architect Camillo Sitte (pronounced ‘zitter’) sought to explain what makes some public spaces more successful than others in a lively, opinionated masterpiece called The Art of Building Cities, published in 1889. Sitte argued that the best public spaces are physically enclosed and fairly small, creating a sense of shelter and intimacy, defining picturesque views and perspectives, and making public buildings appear larger and more dignified by comparison. He denied the importance of symmetry and claimed, counterintuitively, that monuments should be pushed to the edge of squares, and that great public buildings should generally be entangled in a mass of everyday housing. 

Sitte’s book was explosively successful, seizing the imagination of late nineteenth-century Europe. For a few decades, Sitte was a decisive figure in international urban planning, and planners across the world strove to implement his ideas. But after modernism came to dominate Western planning in the mid-twentieth century, Sitte was mostly forgotten.

Sitte’s attacks were directed against the repetitive grids, sweeping boulevards, and impersonal regularity of most nineteenth-century urban planning. To many people today, however, nineteenth-century urbanism is relatively appealing; certainly, it is a lot more attractive than the motorways, slab blocks, and concrete plazas that came after. This can make Sitte strange to read, but it also makes him even more important now than he was in his own time.

Anyone who wants to build beautiful cities today needs to understand what it is that makes the most beloved cities of the past so successful. Focusing only on the appearance of buildings can only get you so far. As Sitte showed, some of the world’s most impressive buildings are badly let down by their surroundings, and form part of inhospitable and uninviting public spaces. Equally, neighborhoods consistent with Sittean principles can be wonderfully charming even if their constituent buildings are nothing special. Understanding these principles may be the key to building new neighbourhoods and cities that people truly love being in.

Create enclosed spaces

The first principle of The Art of Building Cities is the importance of creating physically enclosed public spaces. People feel calmer and more protected in such spaces, which they experience as ‘outdoor rooms’. An enclosed outdoor area can also develop its own character, just as kitchens, bedrooms, and bathrooms have their respective characters. Sitte would have seen a city without enclosed public spaces as analogous to an extremist open-plan house, where the kitchen, bedroom and bathroom are open onto each other and bleed constantly into one another’s character.

While the rooms of a house are created with partition walls, the enclosed spaces of a city are created with buildings. A conundrum for urban designers is that the building wall of a square generally has to be repeatedly broken to provide access. A square is thus like a room with five or ten large doors, permanently left wide open, undermining its sense of enclosure. Sitte was keenly interested in how to handle the entrances to squares so as to mitigate this.

The picture above shows the Grand-Place of Brussels. Most of the entrances are at the corners, and at each corner, only one street enters. These streets continue the plane of the square rather than entering at a diagonal. This means that, from almost any point in the square, the two building walls appear to overlap each other, effectively concealing the entrance to the street and preserving the sense of enclosure. The resulting ‘turbine square’ is so common in old towns that, as Sitte said, ‘it can be considered as one of the conscious or subconscious principles of ancient city building’.

Even better than a turbine square, in Sitte’s view, was one whose entrances had been bridged over, creating a fully continuous building line around the square. This is particularly common in Spain. The Plaça Reial in Barcelona has seven entrances, but six have been bridged over, so that the square feels like a courtyard or cloister. Colonnades, which Sitte thought to be a greatly underrated device in urban design, provide a simple way to achieve this.

London’s Trafalgar Square is an example of the opposite, a ‘square’ that is only weakly enclosed. Wherever one stands, one can generally see two or more of the wide and busy roads that join it. Like many nineteenth-century squares, Trafalgar Square was never intended to be an enclosed space, but rather was designed to improve the flow of carriages between the Strand and the West End.

An even more extreme example of an unenclosed ‘square’ is the Place de la Concorde in Paris, which has no fewer than nine roads entering and exiting it, and buildings on only one side. The space is bewildering and demoralizing for pedestrians, who must work their way across successive roads without buildings to shape their route and shelter them on their way.

Sitte applied this logic to streets as well as squares, arguing that curves and T-junctions should be used to visually terminate streets and give them a sense of enclosure. The picture above shows Broad Street, one of America’s rare curving streets. The combination of the curve of the street and partial T-junction at the end means that Broad Street is visually closed by the Federal Hall National Memorial, creating a famously beautiful and charismatic space. The second photograph shows a more typical Manhattan street, experienced as an endless canyon separating continents of buildings, what Sitte called ‘parallels vanishing at infinity’.

Squares should not be too large

If a square is too large, it loses its sense of enclosure, and its buildings look strangely diminished by their surroundings. Sitte provided a rule of thumb in sizing squares: the dimensions of a square should not be greater than twice the height of the principal building in it. The pictures above show two examples of over-large squares: the Plaza de la Armería in Madrid and Heroes’ Square in Budapest. Both feel demoralizingly vast, with the monumental buildings they are meant to showcase (Madrid Cathedral, and Budapest’s Hall of Art) dwindling to visual insignificance on the horizon, so that they resemble doll’s houses. In both squares, the buildings are so miniaturized that they no longer seem like enclosing walls, leaving the visitor exposed in a huge open space. This problem could be solved by making the buildings much taller, but although the square would then feel enclosed, the buildings would have become terrifyingly huge, as if meant for giants rather than human beings (the courtyard of Les Espaces d’Abraxas near Paris is an example of this pitfall in the wild).

Some of the world’s most famous squares are, in fact, far too large. St Peter’s Square in the Vatican, seen above, was designed by the famous artist Bernini, and it does indeed look impressive in drone footage. For human visitors, however, it is an exhausting place, overwhelming without being impressive. The Basilica looks small and stubby, exactly the opposite of the desired effect. On a summer’s day, the vast, unshaded space becomes relentlessly inhospitable, and visitors huddle under the colonnades for shelter.

By contrast, the small square in front of the Pantheon in Rome exemplifies the features Sitte prized. It is impossible to fully capture its feeling of enclosure in a two-dimensional photograph, but some sense of the intimacy of the space is apparent in the pictures above, which show how this magnifies the Pantheon. The Pantheon is a smaller building than Madrid Cathedral or Saint Peter’s Basilica, but it is much bigger relative to the square. This allows it to retain its dignity, despite the fact that the level of the street has risen by 1.3 meters since Antiquity, swallowing the steps that originally led up to the entrance.

Symmetry is overrated

At the time Sitte was writing, many planners believed that harmonious spaces should have regular, symmetrical plans, an idea which remains influential today. In English we even use the confusing word ‘square’ for any enclosed public spaces, whatever their shape.

Sitte rejected this, believing that symmetrical spaces do not necessarily feel harmonious at ground level, and that asymmetry is not generally disharmonious at all. The pictures above show the perfectly symmetrical Place Vendôme in Paris. Seen from a satellite, it is visually striking and certainly neater than an irregular space. Its chamfered plan supposedly inspired the design of Chanel’s No 5 perfume bottle. Pedestrians, however, generally do not stand at the exact point at which the objective balance of the buildings is reflected in their field of vision: the various points and angles from which they do experience it may even make it feel quite awkward.

Above is the Piazza delle Erbe, a market square in Verona. This ‘square’ has an odd shape, an elongated and slightly wonky pentagon. Nobody is likely to model perfume bottles on it. But at human level, this does not matter as pedestrians can never take in all five sides at once, and so its irregular form is barely even visible. For the most part, the irregularities are ‘seen only on paper’, and to the extent they are visible, they are not experienced as imperfections. Sitte argued that they in fact serve to subtly arouse visual interest, and also create happy recesses that can host resting places, statues, or other public art.

Move monuments to the edges

If you were putting an ornament in your living room, you would be very unlikely to stand it in the middle of the carpet. It would get in the way, break up the space, and might well look lost and isolated. Nor would you place it in front of one of the room’s main features, like the fireplace. It would block the view of the fireplace, and the fireplace would be a disagreeably busy backdrop. Almost certainly, you would put it against the wall, preferably a relatively neutral wall to create a calm background.

Sitte thought the same considerations applied to the placing of sculptures, fountains, and other monuments in outdoor spaces. Statues should not be placed in the middle of squares, where they would block natural routes, break up the space, and look isolated and diminished. Nor should they be placed on axes with key buildings, where they block the view of the architecture. Instead they should line the sides, where they can stand out against relatively neutral backgrounds.

This was implicitly understood in Antiquity, the Middle Ages, and the Renaissance, and Sitte illustrates his book with examples from all three periods. Above, for example, is the Piazza della Signoria in Florence. The numerous statues line the sides of the square, leaving the center free for people to pass through or mill about. This is advantageous for the statues too: Michelangelo’s David (today a replica, since the original was moved indoors) stands out splendidly against the plain masonry walls of the Palazzo Vecchio.

Later planners became fascinated by the apparent neatness of centering statues. Above is the Victoria Memorial in London. The Memorial stands on the main axis of the Mall, blocking the view of Buckingham Palace behind it. This is not advantageous for the Memorial either, whose outline clashes confusedly with the complex detailing of the building. If you search through pictures of the Memorial, you will notice that most photographers picture it from an awkward oblique angle, the least bad way to capture both the monument and the central pavilion of the palace.

Isolating statues against the sky is not much better. A bronze cast of David was later placed on the Piazzale Michelangelo, at the center of a wide square with stunning views over Florence. The statue is enormous, over five meters tall, but it looks strangely miniaturized in its location. Sitte remarked that contemporary tourists often came away with the drastically false impression that it was life-sized. We judge size partly by comparison, and when an object is isolated from comparators, its size becomes harder to assess. This is why the backdrop of a wall can make sculpture look bigger and grander by providing a constant visual key.

In Sitte’s view, the plain edge of a small square was not only an acceptable site for a statue, but the best one. Sitte disliked museums, which he called ‘art prisons’, and thought that only people with a ‘special mental preparation’ could enjoy art under their ‘morbid influences’. The Sittesque city would be one that teemed with art, but woven into the urban fabric as an adornment of ordinary civic life. 

Buildings need context

Sitte’s view on buildings was similar. Nineteenth-century architects tended to place major public buildings in the middle of squares, judging that this gave them greater dignity. Sitte thought this ‘pedantic, futile modern practice’ destroyed the square and squandered the opportunity to frame the public building with vernacular architecture. He pointed out that many historic cities were in fact designed along the lines he proposed. Of the 255 churches of Rome, 41 adjoined other buildings on one side, 96 on two, 110 on three, and two on all four. Only six were freestanding, two of them modern churches of expatriate Protestant communities.

The most striking examples of this arose when nineteenth-century governments actually demolished the vernacular buildings that surrounded historic monuments. This practice, often known by its French name dégagement, was surprisingly common: nineteenth-century planners thought that it displayed the monuments better. It provides a useful experiment on the effect of freestanding versus adjoined monumental architecture.

For example, Notre Dame Cathedral was formerly approached through a narrow street called the Rue Neuve Notre Dame. The street was demolished by the government planner Haussmann in the nineteenth century to create the Parvis Notre-Dame, a large square that the cathedral still faces today. Haussmann intended to improve the view of the cathedral, but the effect is the opposite. The vernacular fabric formed a visual frame, defining and concentrating views onto the facade. It provided scale, making the cathedral look larger by contrast to the ordinary buildings around it. It also prevented visitors from seeing the whole facade until they were very close, close enough that it towered up terrifically before them. Exposed and isolated, Notre Dame looks smaller and lonelier.

Sitte thought that cities should move in the opposite direction. The Votive Church in Vienna was built in the 1870s on what Sitte described as a ‘desolate monster plaza’ that was ‘intolerably exposed to high wind, inclement weather, heat and dust, as well as to the din of the streets, and the incessant clanging of trams.’ The church does indeed have an awkward and lonely feeling. Sitte thought that the architecture of the building was actually very good, and that the defects in its effect were entirely the fault of its location. He proposed building up regular urban fabric on most of the ‘monster plaza’, defining three small squares with well-framed views onto the church. Sitte’s motives were aesthetic, but we should also note the density and spatial efficiency of his scheme.

Another way to understand this contrast is to look at whole cities planned on Sittesque and anti-Sittesque principles. The pictures below show Padua and Washington DC. It is as though a figure-ground switch has taken place: in Padua, the buildings are molded around spaces; in Washington, space is molded around buildings. Monumental buildings in Washington sit alone on their city blocks, untouched by vernacular fabric. Monumental buildings in Padua constantly link up with vernacular fabric to shape small, sheltered spaces from which they can be properly enjoyed.

The rise, fall, and rise of Camillo Sitte

The Art of Building Cities was an immediate success, landing, in the words of Sitte’s English translator Charles T Stewart, ‘like a demolition bomb on the city planning practices of Europe’. Within a year of publication it had prompted the redrawing of urban expansion plans in Germany and Austria, and Sitte was being approached by town officials seeking his advice on their own plans. By Sitte’s death in 1903, advocates of his ideas were to be found in prestigious positions in cities including Munich, Berlin, and Vienna, either teaching or as official city architects. 

Two Helsinki neighborhoods demonstrate the effect of Sittesque ideas on urban planning. In the nineteenth century, Helsinki followed a neoclassical grid plan, with straight streets meeting at right angles. In the early twentieth century, it switched dramatically to a Sittesque approach, introducing curving streets, turbine squares, T-junctions, and asymmetrical spaces. The resulting neighborhood is full of pleasing, characterful spaces and unexpected ‘old town’ effects, even though it is made up of the same tall apartment buildings and wide streets as its predecessor.

But Sitte’s influence was curiously short-lived. In the decades following Sitte’s death in 1903, European urban design came to be dominated by the modernist movement. Everything that Sitte hated about nineteenth-century planning was vastly amplified in modernism. With honorable exception, the modernists treated public space as merely whatever was left over after the buildings had been placed, and had no sympathy with Sitte’s approach of treating spaces as primary and shaping the buildings around them. Instead of lining the edge of a city block with buildings (a ‘perimeter block’) they preferred to orient all blocks in the same direction (‘parallel blocks’), often with a seemingly arbitrary angle to the street. Sitte had decried the gridiron blocks of the nineteenth century for the weakly defined spaces they created. But by twentieth-century standards, the straight ‘corridor streets’ of the nineteenth century look comparatively Sittesque.

Modernists also wanted to reduce the built footprint, leaving more space for greenery, parking, and roads. Sitte’s preference for small squares and enclosed spaces was in tension with this. Many modernists saw him as an impractical romantic. The critic Sigfried Giedion described Sitte as ‘a kind of troubadour, ineffectually pitting his medieval songs against the din of modern industry’.

Ironically, modernism proved to be Sitte’s great vindication, as modernist planning has been repudiated almost worldwide. Architects of all stylistic persuasions have once again become interested in streets and squares, in creating public spaces with intimacy and character. Designers associated with the ‘New Urbanist’ movement are often openly followers of Sitte, and famous New Urbanist towns like Seaside in Florida, I’On in South Carolina, and Poundbury in Dorset were explicitly designed under Sitte’s influence. But the Sittesque revival goes far beyond this. Prominent contemporary practices like Ash Sakula and Peter Barber produce highly Sittesque urban designs, creating intimate infill schemes that recall the old towns and bazaar districts of the Middle Ages.

Mainstream development is moving in the same direction, but slowly and often incompetently. Many recent attempts to make traditional streets and squares look awkward or stilted. This should not surprise us: building cities is an art perfected over hundreds of years. City-builders today are lucky that Sitte distilled and recorded its key principles, just before the art itself was lost. Studying Sitte can help us to relearn it.