“Let us go back to the past; it will be a step forward” (Verdi)

Eduardo Souto de Moura 

An interview with Eduardo Souto de Moura
by Pedro Borges de Araújo, Eurico Almeida and Carlos Machado e Moura

The launch and great success of minimalist frames are inseparable from your work, so much so that initially the myth that you were the designer of this product spread across Portugal. What led you to start using minimal frames?

When frameless glass doors began to appear, I realised that glass is a structural element. Throughout the history of windows and openings, frames have been load-bearing and the voids filled in with glass. But at some point in time the contrary started to happen: structural glass did away with the frame, which from then on only served as a sealing device. It became a finish for the glass itself, a corner protector. I’d been asking myself why frame brands would not take advantage of this possibility until one day when a client showed me the Vitrocsa doors and asked me about the feasibility of this product.
I replied: “That’s exactly what I’m looking for. And as I’m designing your house, I’ll use them there. This way, you can test them and find out if you’re interested in representing them for Iberia.” And that was that.

Do you agree with the idea that the technical and formal purity of this type of frame comes as an almost natural continuity of your architecture?

Obviously, it was in line with the architecture I had been practising. Before living in my home, which I designed, I had this idea – which I still entertain, although more moderately – that architecture is made by walls with holes in them. Holes which, when extended down to the floor become doors and, when not, are windows. So, walls, doors and windows make up the trilogy all architecture consists of historically. Until the Modern Movement and Neoplasticism emerged, to which I adhered early in life because I’d always grappled with the scale of buildings. Whenever I tried to design a window or a door I couldn’t find the right proportions because these were always in reference to the old architecture. And I would always lose out. Because the old architecture has a ceiling height of three or four metres and the walls have a thickness of half a metre or a metre. Now we have spaces with a height of two forty and thirty centimetre walls. Everything is lyophilized. So the grammar of this type of language no longer consists of walls with holes but rather walls and non-walls. And a non-wall would require the appropriate expression to strengthen this positive-negative dichotomous image. I was striving to give it a minimal presence so it could be as transparent as possible and the frame could lose its prominence. It was a happy marriage.

This direct relationship between the architectural language and the form of the frames is something that clearly emerges from your work…

Well, I owe that to Siza. Siza’s work has a much more classical language which is completely different from mine, even though he himself was a Neoplastic at an early stage. Just look at the Leça Swimming Pool and the houses where the wooden frames never appear in the middle of the walls, they are always against the sides and the wood folds into the corners. Precisely so that the windows fold into the corners, so that they are a negative of the wall, a non-wall, not quite a hole.

It was when I worked with Siza that I learned to abide by the principle of the “overall work”, i.e. to understand that the work is not made up of separate layers but is rather a whole. You don’t build the foundations first, then the pillars, then the walls, then clad them and finally fit the frames into the holes. With Siza we’d integrate all the 1:20 scale engineering and architectural designs: the plaster, the waterproofing, the door and window frames and even the electrical wiring. All this was done on a model on which Siza would mark where he wanted each component with a biro. He called it the “overall work”, a method in which the chances of error are less because you’re able to control all the elements gathered before you.

However, I recall that Rossi, when I was a student of his, criticised this kind of architecture precisely for this approach. He thought that detail was unnecessary, that each specific development should be entrusted to a specialist. But I was accustomed to this way, to the overall work process, a continuum that is really Loos’ principle.

So, I’m still doing it today, I try to bring together all the information on a project. This leads to the frames having their proper place vis-a-vis the double wall, the blind, etc. All this is dealt with both on the plan and in the section, and the frame is part of a system. I’ve even developed an aluminium kit for Desenho Ibérico that never went ahead, which included a frame in which the sill and the blind were already designed.

So, if we define architecture as walls with holes, then the holes are part of the architecture. It’s something incorporated into it, we won’t be buying a tie without bearing in mind the shirt, the jacket and the shoes. At least, that’s how it was until now.

Is it the reason why it’s so difficult to design windows?

The great difficulty in designing windows lies in the fact that the window is part of a grammar and an organisation of elements. Nowadays the elements have changed so much that we can’t apply the same rules with different materials. Since the wall has stopped having its own thickness, we might be driven to use more and more glass. A window placed above, that doesn’t extend to the floor, can actually be seen as something romantic, like in the film “O Pátio das Cantigas”(1). So, the loss of the wall thickness is the key issue for me. It’s not just the length and the height that determine the proportion of a building, but also its depth. And, with this trilogy, my one by one metre windows work well in the Monastery of Bernardas because the walls are one and a half metres thick. But the same windows wouldn’t work in the Patio Houses because they miss one of the proportions. They don’t have the thickness, it’s like a large model. And in this case I give up, I’d rather do a non-wall.

What’s the main factor in designing a window? The cost, the ease and convenience of use, the human scale, or is it something else?

None of those, you can’t choose one. It’s a number of countless factors that vary according to circumstances: cost, convenience, scale, context.

When I design a window, first of all I consider what will look good. Not just to contemplate aesthetically, but in terms of logic and comfort. For example, I don’t like to sit with my back to a window, it fills me with a weird, insecure sensation. In fact, in a café people tend to sit in the corners and sides, and the middle remains empty. And in a lounge, you have the position of the sofas and the TV. Besides, the size of a window also depends on how much we want to see outside. I try to create a framing, like a film maker, seeking to not make it overly large, or the interior will vanish and the exterior will be always present, or too small, because then the outside becomes something remote.

And what’s interesting in the evolution of architecture is that we have the possibility of not making choices but rather varying the interior environment. Today I’d rather have more of this, less of that, more or less sun… There are mechanisms that provide us this control, for our comfort. That’s exactly the word: Távora used to say that “architecture is where we feel good”. But at the same time, when we design the elevations, we check whether a given glass panel is absurd or over the top… So it’s these interior and exterior factors that set a window’s boundaries in terms of height and width. All this is done by sensitivity. It’s an equation that needs to be calibrated so that all variables fall naturally into place, so the window is not too heavy, too expensive, too opaque when it should be transparent. And frankly today, I don’t know why, I tend to close off more and more. Because I like it…

Frames have known great development, trying to keep a minimal expression but covering increasingly large glazed surfaces. When exactly did transparency become an architectural need and began to change how windows are designed?

There are two major ambitions and transparency is one of them. One is to counter gravity and everyone has attempted it by making their works lighter. Távora, in his famous class on gravity, “along the thread of the pyramids”, which started with the Egyptian pyramids and ended with Niemeyer’s inverted pyramid, explained that the most stable form corresponds to the pyramid, which changed over time, until Niemeyer makes the inverted pyramid and Pei the glass pyramid. (2) Therefore, there’s this obsession, this vertigo of lightness on which Italo Calvino wrote a very interesting essay. (3) The other ambition is the issue of light. Technical constraints prevented openings from being larger for a long time. Windows used to be tiny, the Ancients needed gigantic stones to create a one-metre opening until evolving techniques made it possible to use different materials: stone and wood, arches, concrete and iron, maybe carbon in the future. All this development came about so that people would have more light.

It is no coincidence that the Modern Movement originated in the Nordic countries, which have stronger links to Calvinism, and where night falls earlier and they need more light. So transparency comes about on the strength of two assumptions: on the one hand, the need for light to work and live; on the other, a religious, moralistic issue of not having anything to hide. The adoption of a correct stance, since there’s no absolution, no escape. As for Southern countries, with their Catholic culture, we find three window obscuring layers: firstly shutters, then thinner curtains, and then drapes. While some need to behave because neighbours are looking in, others have confession to save themselves.

And with the Dom-Ino structure, which is set back, the façade is independent and self-supporting, and can be entirely in glass. This is particularly useful for the real estate sector since office buildings have problems at the rear, from lack of light in the areas furthest away from windows. That’s why extending the glass until the ceiling allows for deeper illumination. All these issues are instrumental. I’m convinced that the History of Architecture has variants but it always starts from a need. Perhaps it’s a pragmatic illusion of mine, but I believe that everything is born out of a primary function, which then gives way to digressions or semantics that bestow a given meaning on a use.

All this takes on a different symbolism when companies do not see a functional need in glass but a question of status because glass is expensive and modern. That’s why the Seagram was entrusted to Mies, the architect who most used glass, just as bronze was chosen as a cladding, an imposition made by the client. All this has progressively had different expressions and statuses.

In turn, the German Pavilion at the 1935 World Fair in Brussels, entirely glazed with the Swastika on the front, was a way of publicising the Nazi regime as modern, transparent and progressive. In their letters to Goebbels and Rosenberg calling for the Bauhaus not to be closed down, Gropius and Mies used precisely the justification that the superior capacity for production of the Bauhaus over other centres was a way of demonstrating the superiority of the Arian race. In short, transparency becomes a mark of power.

Do you face technological challenges today, development avenues to explore, besides those semantic dimensions?

There are developments but they’re much slower or not applied at all. Whenever I board a plane and see the stewardess close the door, making it slide with just two fingers, I wonder why houses don’t have this technology. It doesn’t exist.

Countless solutions used in aeronautics in WWII contributed to the technical innovations of post-war Modern Architecture, as indeed the book by Jean-Louis Cohen shows (4). Neufert himself was an engineer in the German Army during this war, and applied a rational systematisation to German industrial and military architecture.

So I’d say there are development vectors, and one is the economy. A one and a half metre thick wall in the Middle Ages is reduced to forty centimetres in the 19th century, to a 10 centimetre panel in the Case Study Houses, and today to a 2 centimetre one in glass. And from a physical point of view, it may even perform better.

But on a psychological level, this is not the case. The evolution was not so quick because the house is a very conservative institution. Many of the typological elements of the Greek or Roman house subsist after thousands of years – the day-time and night-time areas, the public and private areas, the courtyards, etc. And a Sumerian house from 5,000 BC is quite similar to an Alentejo house. The door, the roof, the area for the animals, the sleeping quarters, etc. are the same. I say this with our Hellenic and Roman culture in mind. Of course, there are other cultures in which the relationship with the space is completely different.

However, as Soutinho said in a class: “Astronauts love to return home”; despite the contribution of all the technical innovations to a person’s well-being, they don’t actually create well-being.

Jofebar has sought to reconcile industrialisation with a certain level of craftsmanship, making prototypes and developing project-specific solutions. Does this approach make sense under market conditions?

There’s a very interesting topic to do with the fact that the Modern Movement borrowed the basis of its design from industry, even if it hasn’t used the industrial logic subsequently. I’ve never understood why; it’s a mystery to me. For example, in the metal structure of Mies’ Barcelona chairs, the two welds for the X take two days to complete. This is everything but an industrial choice in a chain system. And the metal tubes in Le Corbusier’s grand confort sofas would be flattened if they were bent. They would require very large radiuses, like the Cassina LC4, which is the only chair with a correct curvature radius. So, they’re expensive, not only because of their design status, but also because their manufacturing process is irrational from the industrial point of view at their origin. That’s why they are not simply a consumer item but have become a work of art. This rationality aspect is the starting point but never the end. Maybe that’s why these objects are not comfortable. And, quite interestingly, that principle also applies to architecture. We could consider why the houses in the History of Architecture have seldom been lived in…

How can we reconcile today the solutions designed specifically for each project with the operational needs for certification, safety, performance and economy that are imposed, e.g. by the European Community?

That’s a really interesting question because I think that virtually the entire economy works for profit. From Russia to China through to America, there’s a capitalist system of capital gain generation which drives the economy. No-one’s really very interested in focusing on people’s happiness. If they were, waste would be used to put an end to hunger and poverty. So, in my view, this whole issue of safety, performance, economy, etc. is mainly a way to increase profit. And even ecology is a form of marketing that works because it’s fashionable. Does it make any sense that in Zurich thermal insulation for walls is thirty centimetres? If it were for the cold, you’d wear a jumper, it’s that simple.

Or the irrational measures currently imposed in France, making every house be prepared for people in wheelchairs. Since every room needs to accommodate a turning radius of one and a half metres, but the living area has not increased, hallways, kitchens and bathrooms have to take space from the lounge which is reduced to a table and four chairs. Conviviality is gone. This accessibility principle applies to all apartments, even to three-storey buildings without a lift, which cannot be accessed by a wheelchair. It’s a contrivance! And what about the limitations on glazed surfaces? I designed a standard residential building in Bordeaux, facing a lake, and wanted a wide oblong window to provide a view over that beautiful scenery. But I couldn’t exceed one square metre, which is the limit imposed by law on residential windows, established to restrict energy consumption. To have a wider surface, I’d have to install photovoltaic solar panels to make up for the additional use of energy. Other options, such as restricting maximum heating power, are not acceptable. There is a clear option for increasing the sales of panels, and the subsidy granted to their installation is withdrawn if these are not visible from the street, as was the case with a design by a friend of mine.

I’m not saying that ecology and safety are just a business, but in construction they amount to economic gain: more concrete, more iron, more insulation thickness, more double and triple glazing, more investment money are always needed.

So, you’re not happy with the state of the art as far as frames are concerned?

Not at all! And to start with, as for the term “state of the art” I have my doubts that architecture is an art. It must, first and foremost, meet a function, whereas art doesn’t have to meet anything, it stands on its own. I’m not denying it might be art, but I reject the idea that there’s an underlying artistic principle in architecture. I’m called upon to solve a problem, I don’t go about making houses to express a feeling… it’s only afterwards that the project evolves.

But no, I’m not happy at all because there are millions of problems to solve, starting with the materials themselves. For example, now there’s a focus on electric cars, but lithium batteries are made from cobalt from the Congo mines. The ore is being depleted and its extraction raises a series of human problems. (5) One day soon they will bar batteries and electric cars will be no more. And the situation with aluminium is similar. As far as I know, the consumption of energy required to get to the purification state of pure aluminium, with all the treatments and anodizations, is extremely high. We have to focus on simpler solutions. There are no stone frames, they have to be in wood. There’s nothing more environmentally-friendly than wood, because it is planted, cut and planted again, it’s a good insulator and its repair is quite artisanal. At a time when so much attention is given to the classification of materials, as was the case with the Clermont-Ferrand Theatre, it’s the right way to go. And glued laminated timber is so resistant that, in a breakage situation, it doesn’t break at the glued sites. This development suggests the possibility of mixed situations with metal and wood. Reinforced concrete is nothing but reinforced stone. And I’ve designed tables in reinforced wood which you look at and see three centimetres because the inside is completely in iron. These tables have never moved except at the mitred joints. So I’m quite sure we’ll be making frames in reinforced timber. I also think there are no good or bad materials.

In the 1980s the Iran-Iraq war aroused a fear there might be a shortage of petrol. The automobile industry responded with tiny, very light, low-consumption vehicles, such as the Citroen AX which had wonderful publicity on the Great Wall of China. (6) Now it would appear that petrol won’t come to an end, that it can self-regenerate and be produced from schist. As such, petrol was never spoken of again, as there’s an abundance of it and its price has decreased, changing the paradigm all over again.

There’s new materials to be discovered for sure, extremely hard plastics, many products to succeed wood, iron and even aluminium. I remember when Siza and I were collaborating in Berlin, he proposed to use wooden frames and was told it was fine because they were cheap. And Siza asked: “That’s cheap?” They replied: “We were afraid you would want aluminium.” Because aluminium was expensive and a job well done is really expensive. We’ve come a long way but there’s still a lot to be done.

What pointers or ways can you suggest for future developments?

Stravinsky quoted Verdi saying: “Let us go back to the past, it will be a step forward.” (7) He was referring to music but I believe it also applies to architecture and construction. For instance, when I’m due to design a house and I’m at a loss, I open the Inquérito (8) and I draw from the photos. I don’t copy the photo but as I draw I interpret. Unconsciously, of course, the copying mustn’t be a conscious act, otherwise it would be ridiculous.

So, I believe we could recapture an old practice in teaching architecture, which is to redesign buildings. There was the “statue drawing” which you learned to do by reproducing Greek statues, as I did. And in Fine Arts Schools, all modern architects used to draw Greek temples to learn how to draw and study the proportions. Even though some think this is a dangerous practice, the architecture they were doing had nothing to do with that exercise. Távora has never designed a Greek temple, or Viana de Lima for that matter! When I worked with Siza and he inquired about some proportion – for instance for a balustrade for a church in São Miguel de Ceide – he’d ask Távora who, in turn, would draw Palladio’s balustrade by hand, whose proportions he had studied. Even Le Corbusier would draw his villas and then apply Florentine Renaissance lines to them.

When it comes to frames, I think it’s important to cultivate their history and the practice of their design to understand what may be inferred from them for the future. The same goes for cars. For example, there’s a book of all the drawings of the Volkswagen Beetle and the Porsche 911 on tracing paper and, when you superimpose the sheets, you see that the differences between the various models are minimal: one, two, three centimetres. They’re constantly redrawn but they’re always modern cars and they’re always the 911 model.

The same happens with frames. I remember the window I designed in brass for the Pousada de Santa Maria do Bouro which Távora subsequently used when he dealt with the extension to the Parliament Building, also metal on metal with no rubber weatherstripping. I used it again in another project but applying rubber. Now Francisco Vieira de Campos, Cristina Guedes and João Mendes Ribeiro have used it in the Arquipélago – the Contemporary Arts Centre in Ribeira Grande (Azores), always changing and improving on the model. This detail is the result of transposing to brass a wooden frame that existed in Bouro with a very interesting system. It has a reversed tooth, inwards, which lets water in, running off along some grooves, to be collected in a half-round and then come out through a hole. It’s the British theory, which Jorge Gigante advocated saying: “In construction, water must get in unhindered so we know where it is, and then we take it out”. As far as I’m concerned, it bothers me to put the trench coat on under the jumper, but sooner or later the water gets in and we need to take it out. It’s smart reasoning. In the same way, the sash windows of Braga which also tilt, are a very intelligent system. I believe it could be a source of inspiration, a starting point for a new solution.

One needs to place some trust in tradition. In the same way that animals evolved over millions of years, in construction there are also features of vernacular architecture that have been improved over time. And this evolution has made them more natural.


1. “O Pátio das Cantigas” is a Portuguese film from 1942, directed by Francisco Ribeiro, “Ribeirinho”, that takes place in a typical Lisbon neighbourhood.

2. This refers to the unbuilt project by Oscar Niemeyer for the Museum of Modern Art of Caracas (Venezuela) in 1954, an inverted pyramid in reinforced concrete, and to the Pyramid for the Louvre Museum in Paris, designed by Ieoh Ming Pei in 1984-89. Together with The Great Pyramid of Giza, they were the subject of Eduardo Souto de Moura’s “A Triangular History”, executed by Jofebar for the Lisbon Architecture Triennale 2016.

3.  This refers to “Lightness”, the first chapter of Italo Calvino’s: Six Proposals for the Next Millennium [original: Lezione americane: sei proposte per il prossimo millennio, Milano: Mondadori, 1988]

4. Cohen, Jean-Louis: Architecture en Uniforme: Projeter et construire pour la Seconde Guerre mondiale, Montreal: CCA, 2011

5. http://www.washingtonpost.com/graphics/business/batteries/congo-cobalt-mining-for-lithium-ion-battery/

6.  https://www.youtube.com/watch?v=fOn4digOIvM

7. Torniamo all’antico e sarà un progresso. Written by Giuseppe Verdi in a letter to Francesco Florimo, 5th January 1871.

8. “Arquitetura Popular em Portugal”, published in 1961