PREVI: The Metabolist utopia

Forty years later, an experimental district offers us an unparalleled insight into the slow metabolism of the infinite city.


Architecture / Justin McGuirk

PREVI—Proyecto Experimental de Vivienda—an experimental district collectively designed by a generation of radical avant-garde architects who converged on Lima (Peru) in the late 1960s, was a pioneering attempt to reconcile the conflicting forces of informal growth and top-down planning.

In 1965, Peru's architect-president Fernando Belaúnde Terry began consultations for a social-housing programme that would regulate the unstoppable flow of people seeking an urban destination (in the mid-1960s, the informal city, with its barriadas, seemed to have overwhelmed the "urbanised" part of Lima). Under the leadership of Peter Land, and with the support of the United Nations, the project for an experimental neighbourhood was born; it involved the best radical avantgarde international architects chosen from among those who had a solid reputation for social housing. These architects were invited to participate in the creation of this dense urban collage. All of them had been involved in the most interesting experiments in social housing in the early '60s, for example, Runcorn New Town Housing by James Stirling, the house-capsule for Nippon Prefabrication Co. by the Metabolist group, Tube Housing by Charles Correa, which became a model for the Indian region of Gujarat, or Cité Horizontale in Casablanca by Georges Candilis, who had collaborated with Le Corbusier in Marseille.

In PREVI, 13 internationally renowned architects, along with as many Peruvian architects, were commissioned to develop a model neighbourhood of 1,500 dwellings to develop prototypes of urban housing that would internalise programmes for any future transformation. Thus each unit contained the terms of its own growth. This was perhaps the first act that recognised the value of the dynamics of growth adopted in the informal slums. In contrast to a growth model based on large, out-of-scale gestures—from megastructures to gigantic superblocks—the PREVI experiment fielded new dynamics based on a model of low-rise, high-density housing. When, in October 1968, a military coup led to the overthrow of the president-architect who had promoted the previ project, the involvement of the United Nations prevented the project's cancellation. The jury met in 1969 and, while having chosen the winning projects (the international groups selected were Kikutake-Kurokawa-Maki, Atelier 5 and Herbert Ohl), resolved to initiate the construction of all but two of the proposals. In 1974, the first phase of 500 units was finally built and left to its fate of growth and progressive oblivion.

In the north of Lima is a housing estate that could have changed the face of cities in the developing world. Its residents go about their lives feeling lucky that they live where they do, but oblivious to the fact that they occupy the last great experiment in social housing. If you drove past it today, you might not even notice it. And yet the Proyecto Experimental de Vivienda—or previ for short—has a radical pedigree. Some of the best architects of the day slaved over it. Now it is largely forgotten.

PREVI was the product of exceptional conditions. In the 1960s, the population of Lima was growing so fast that government housing schemes were proving woefully inadequate. Instead, people were building their own homes in informal barriadas, which today account for more than half of the city. In 1966, President Fernando Belaúnde, who was also an architect, held an international competition in conjunction with the UN to devise a solution to the city's housing problem. The list of participants reads like a roll call of the 1960s' avant-garde: James Stirling, Aldo van Eyck, the Metabolists, Charles Correa, Christopher Alexander and Candilis, Josic and Woods. These are only the most famous. There were 13 international teams and 13 Peruvian—it was a housing Olympiad of sorts. Never again did so many prominent architects weigh in on the issue of social housing. The profession disengaged, eventually to discover the museum as the pinnacle of its ambitions.

Without malleability you cannot have cultural expression—all you can get is a top-down notion of how people should live

The first day we spend at PREVI the sky is hazy, as it is almost every day in Lima. The light is as diffuse as in a museum, casting no shadows. It is forensic weather. These days previ is a gated community, but in the end everything gets a gate in South America. To our right we pass a football pitch and a playground. On our left is a little square with a tree and some houses with Po-Mo porthole windows. These are James Stirling's. You can still recognise his single-storey design under the two- and three-storey houses that sit here now. The entire day consists of such moments of recognition. "These are Van Eycks." "These must be Alexander's." It's not always easy to tell, after three decades' worth of expansion. The original houses are encrusted with geological layers: extra floors, pitched roofs, balconies, external staircases, faux-marble facades, terracotta roof tiles and bright coats of paint. It's like a form of archaeology, mentally scraping away these accretions. That was the genius of PREVI: it was designed as a platform for change. The houses were not the end but the beginning, a framework for expansion. It was revolutionary. Of course there was a tradition of the working class modifying their modernist offerings, as Le Corbusier discovered to his chagrin in Pessac, but it was never intended. The prevailing solutions to mass housing, the tower and megablock, were singularly inflexible, and they simply couldn't be built fast enough or cheap enough. The barriadas, however, were starting to impress architects with their home-grown ingenuity, and PREVI was conceived as a formal version. The competition was geared towards low-rise, high-density concepts, and the architects were flown in to study the tvvv. Come 1970, after choosing three nominal winners—the Metabolists, Atelier 5 and Herbert Ohl—the judges decided to build all of the entries (except, ironically, Ohl's, which turned out to be too difficult to realise). The first pilot scheme would consist of 500 houses, so that the designs could be fully tested on the ground, and in the second phase the best would be rolled out by the thousand. Except the second phase never happened.

For that reason, many considered PREVI a failure. Imagine investing in 24 different designs and construction methods—from brick to prefabricated concrete panels—in the hopes that the economies of scale would make up for it when the scheme was standardised. In the end, previ became an anomaly: a housing laboratory containing so many design ideas, that was so diverse and adaptable that it can probably never be repeated. Turning down an alley, we find ourselves in a little square overlooked by half a dozen houses. From the seams on the prefabricated concrete walls, they are easily recognisable as the Danish entry, designed by Knud Svenssons. The owner of the white house with the blue trim, Juana Mazoni, opens the gate and leads us across a patio into the house. The living room is tennis-ball green, and it's an Aladdin's cave of plump sofas, lace doilies and gilded mirrors. Above this heavily nested space, and in striking contrast to it, is a modernist waffle ceiling—hung with a chandelier (see page 24). It was when she first saw that ceiling back in the mid-1970s that Mrs Mazoni realised she was moving into a special house. Somewhere she even has a faded photo of her and her husband with Svenssons. He promised to come back and help them build another room above the living room (to replicate their beloved waffle ceiling) and they're still waiting.

Every family in PREVI has a story about their house. They may not know the architect's name, but they know which international team designed it. It's common to refer to the different typologies by nation: so and so lives in one of the Dutch houses, or in a French house. It's a form of orientation. But more than that, the families within each typological quadrant have a sense of solidarity. During the football World Cup (in which Peru rarely features) they will support the country whose house they live in. Sometimes they even organise their own tournaments, on the pitch by the gate.

Who knows if such competition prevailed among the architects themselves, back in 1969, but they certainly took markedly different approaches to the problem of dense, low-rise housing. Christopher Alexander, who went on to make his name with the Pattern Language books, spent two weeks living in a barriada before designing his house. It's not easy to tell that from looking at them, but inside there were very few walled-off rooms because he had apparently observed that Peruvians socialise in alcoves. Instead, he expected smaller spaces to be curtained off as needed. Aldo van Eyck, meanwhile, noticed that women were the heart of the home, and placed the kitchen in the centre of the floor plan. He also took a more proscriptive approach to how the owners should expand, creating diagonally walled courtyards to discourage people building on top of them. He failed of course. Outdoor space is not sacred to a family of eight with another generation on the way.

It's almost possible to read the success of each typology based on how extensively they've been transformed. The Metabolist houses—designed by the trio of Kisho Kurokawa, Fumihiko Maki and Kiyonori Kikutake—appear not to have been that successful. Though long, they are uncomfortably narrow, and their ground floors are generally rented out as shops and restaurants rather than homes. Made of cheap breeze blocks, it's hard to imagine that they could have had the same budget as James Stirling's generous courtyard houses, with their prefabricated concrete walls. One of the latter has been transformed into a four-storey school. Elsewhere, the original houses are like kernels inside popcorn. Some of these homes are extraordinary works of transformation whose occasionally surreal suburban grandeur belies their setting. Tinted windows and hacienda styling may not meet with architects' approval but they speak volumes about the owners' pride and aspirations. Therein lies one of previ's great successes. People didn't move out as their financial situations improved. Residents stayed, and turned a housing estate into what feels like a middle-class community.

PREVI may have been largely forgotten but the lessons have not been lost. It just took 30 years to learn them. Today, there is a new orthodoxy emerging, at least in Latin America, that says housing should be built with a view to expansion and adaptation. Chilean practice Elemental's "half a house" model, deployed at the Quinta Monroy housing in Iquique, is the descendant of previ. Although in many ways it is not fair to compare previ with the much humbler Quinta Monroy, the comparison is rather eloquent on the situation in which architects engaged in social housing now operate. While building someone "half a house" is an ingenious solution to an extreme problem, it also reveals the lack of wholehearted government support that architects have today compared to that idealistic moment in Belaúnde's Peru. The pragmatic compromises made by dedicated architects such as Elemental reflect a weakening of ideology and the failure of the state. And as governments have neglected their social responsibilities, architects have courted private patronage. The closest we have come in recent times to a generation-defining housing competition is Ordos 100, a vanity project by a Chinese billionaire in an uninhabited patch of Mongolian desert.

There are signs, however, that the tide may be turning. If PREVI marked the shift from a dogmatic modernist approach to housing the poor to one that capitalises on the evolutionary, organic nature of informal settlements, that ethos has now been mainlined by a new generation of socially motivated architects such as Urban Think Tank, Jorge Mario Jáuregui and Elemental. And in them, Van Eyck and Candilis, Josic and Woods et al finally have their successors.