Architects don't build

Peter DowntonStudies in Design Research: Ten Epistemological Pavilions Melbourne: RMIT University Press, 2004

Architects make representations. That is the interesting thing about looking at this exhibition through a craft lens. Craftspeople make buildings. Architects make models-and often not very well. Homo Fabe r is an appropriate title for this exhibition, because it doesn't mean man-made, but man-the-toolmaker, a reference to the intelligence of homo sapiens. These models are tools: for designing buildings and for testing thoughts and ideas. And they were never intended to be viewed by a larger audience than the designers themselves.

When working towards a design, architects don't want their sketches or models to be tight, finished pieces full of gloss and shine-they need them to be rough, ugly, inept-for them to function as 'workings-out', a step in solving a problem, nothing precious that might stop the experimental urge. As Michael Ostwald says in the catalogue, 'Architectural models must remain expendable, ignoble objects, for them to be useful. This does not devalue or limit them; on the contrary it is their natural state and their particular strength.' So what is to be learnt about craft from this exhibition, which stated in its brief to the twenty-two invited architecture practices that they were interested in 'dog models' ( design models, as opposed to presentation models) only?

What do we stand to lose?

Part of the motivation for holding this exhibition and symposium was to find out what we stand to lose by not modelling any more, when 3d printing technology becomes cheap enough to replace it, just as computers have replaced drafting boards.

What do we lose when we give up card modelling? Slow, reflective, work, thinking through making. As Mark Burry asks, where is the contemplative moment in a file-to-factory process?

Even looking at an exhibition of models is slow. The well-attended opening was an opportunity to socialise but didn't encourage concentration on the displays. And how many time-poor architects went back? John Fowles, writing in 1969, 'The supposed great misery of our century is the lack of time; our sense of that, not a disinterested love of science, and certainly not wisdom, is why we devote such a huge proportion of the ingenuity and income of our societies to finding faster ways of doing things-as if the final aim of mankind was to grow closer not to a perfect humanity, but to a perfect lightning flash.'

Models for , or models of ?

The distinction between models 'for' and models 'of' was raised in the concluding paper of the symposium by Ranulph Glanville. Design as a verb, rather than a noun. Prepositions, not propositions. Is the model a proposition, or a question? Every drawing an architect makes is a question, argues Glanville. Which section through the building do you draw? Answer: the one that poses the most questions (and then attempts to answer them). Research in universities has generally been centred on scientific knowledge, rather than a broader academic knowledge, but science is knowledge 'of', and designers mainly need knowledge 'for', knowledge for action. As architects, according to Glanville, 'our job is to make the world different: we aren't so much interested in what the world is as in what it could become .' Physical models may illustrate a formal composition in an abstract way-models of . But for architects these mode ls generally come after the design, which is carried out through models that tend to be much less physically representational-models for .

The distinction between models of and models for is confused in computer modelling, according to Glanville. Is the model a proposition, or a question? Selenitsch argues that the computer model is 'sometimes considered to be 1:1, but none of its manifestations are at this scale. It would be more accurate to think of the digital model as scaleless.'

We stand to gain a lot from computer modelling. Environmental modelling-wind, sun, shade-can only be done in the computer. In this era of oil crises and global warming, traditional building techniques are coming to the fore. Research into ancient building arts, practised prior to the era of cheap energy, is read by today's architects. To be sustainable, we need to build smart, design slow, and make simple.

This exhibition came out of a realisation that design models had been under-studied in the academic sphere. Sketching had been studied, books written, exhibitions held, presentation models had been exhibited and published, but not design models. The Homo Faber study gained funding from the ARC (Australian Research Council), and this exhibition and symposium are intended to be the first of three.

The gallery is organised with the four exhibiting academics in the centre, and exhibiting architects surrounding them on all sides. Submissions-words and images-by the architects (of which only a curated, small portion are on display) form part of the research by the four academics, and the catalogue to the exhibition, issued as a CD, contains initial responses to a survey of modelling in architectural practice.

I am an architect. I work for one of the twenty-two companies exhibiting alongside the four academics central to this exhibition-John Wardle Architects-and the biggest revelation for me was Gaudi.

Gaudi's Lesson

Mark Burry, an academic at RMIT's Spatial Information Architecture Laboratory (SIAL), exhibited 'models' (1:10 scale prototypes) of elements of the Sagrada Familia, along with a video of their making that showed the plaster prototypes being constructed, demonstrating the hand craft-laboriously slow and personal-that produced complex geometric decoration. It is an illustration of wise, pragmatic use of complex geometry-all the shapes are composed of swept straight lines facilitating clear communication between designer and maker. Gaudi's lesson: establish a technique, a buildable technique, and apply it in as many complex ways as you like. Gaudi was not an innovator from a technological point of view. He was interested in using known techniques, intelligently-innovation in the design, not the fabrication.

This is a reversal of 'Oh, look at that complex and interesting thing-how can we get it built?' Designing or finding a geometry, whether digital or physical, and attempting to translate it into bricks and mortar, turns it into 'This is how they will build it-I wonder how complex and interesting we can design it?'

Examples in the exhibition of the former method are easy to find. Take Peter Elliot's Ballarat glass-house project. Elliot exhibited a design model, a folded origami-like structure, that bears a striking resemblance to the built project. A photograph of the model shows it being held in someone's hands like a little Christmas decoration. The shot seems to say 'Voila! Here it is -as simple as a piece of folded paper!' But why this literal translation from fold to frame-and-infill, paper to steel-and-glass, when the two have nothing structurally in common? Unlike Burry/Gaudi's prototypes, Elliot's model is part of a story about the building, a motif rather than a display of structural logic. As much as it looks like a test of a folding structure, it is actually a test of a form's aesthetic potential. The exhibition indicates that very few architectural models are made as prototypes.

One of Michael Ostwald's four essays in the catalogue quotes Frank Gehry: 'The ripped edge [of a torn piece of paper] can be beautiful. But you cannot make architecture do that. [If the model is made of] crinkled-up paper. Now, [the question becomes] how do you get that into final form.' (Tonnes of contorted steel is the answer.) The practice representative at the Symposium, Kirsten Thompson presented a single project that demonstrated the five types of models she outlined (one of which was the prototype). The model exhibited at the Museum was of this project, and it had pasta columns. Spaghetti was the only thing she could think of to model the particular radius of material she was designing with, and when it was built she thought 'what beautiful translucency-how can we achieve that?'

Michael Ostwald, another of the four academics, from Newcastle University, exhibited results of his experiments with rapid prototyping, 'dances with the machine'. Ostwald set out to test the relationship between technology and the architect. He took a model of Le Corbusier's Villa Savoye that came with the CAD software he was using, and tried to get a model built, in America, via a website, using a digital technique called stereolithography, which directly translates the complex 3d computer model into a three-dimensional physical object. The machine process involves lasers solidifying liquid plastic. This sophisticated representation is at the cutting edge of architectural practice, and is predicted by many to become commonplace in the future. Ostwald ran into difficulties: the model had to be readjusted first to have a minimum thickness of wall, then translated into a different file type (which lost a lot of detail), and after many more troubles and human assistance (theoretically it could all have been done automatically via the website) the experiment was abandoned.

Denton Corker Marshall exhibited photographs of a model made using stereolithography. Whilst producing a beautiful sculptural form, there is still a large gap between the pragmatics of construction at full scale and computer-generated forms, so this method gets you no closer to a building. Instead of waiting for the day a building can be 'printed out' at 1:1 scale (and there is at least one company in the United States trying to achieve this with poured concrete using a giant computer-controlled gantry), the intelligent thing is to work with the grain. Michael Ostwald uses the term 'CAD carvers' as opposed to modellers-I think this is a very useful expression. To model with the right parameters that ensure smooth processes all the way to the construction site.

The only built example in the exhibition of this approach was by Dale Jones Evans who used CNC technology to make a door-skipping the physical model, as a door is almost model size. Evan's door looks like frozen mercury, hard matter made liquid. Its smooth, fluid geometry was built like a contour model out of laser cut mdf, which was then glued together and finished by hand. When painted in a high gloss silver, the result was miraculous. For joinery, and other small bespoke items, this technology holds great promise. Evans' door relies on a grain-less material, mdf, to manifest the complex computer geometries, but you can't construct a whole building this way.

Art

The models in this exhibition that least represent future buildings were by RMIT academics, Peter Downton and Andrea Mina.

Made of timber and metal, Downton's models are explorative and experimental. They look more polished than the conceptual studies that they are, and can take up to 14 months or more to make. They are not iterative, but cumulative, each one unique. There are two series of pieces and each piece has a title. The first series had the titles given to the pieces after they were built ('An architect's dreamed shelter' for example, or 'An ideal of home') and the second series were generated subsequent to the titles ('Song Station', 'Music Jetty'). They are dioramas, each one on a timber base with its own little scaled-down world including shading structures, furniture, window frames, some with suggestions of hydraulics and operability. Scale is suggested by doors and windows, or a little seat under a protective canopy. You get the feeling of the scale and simultaneously the emptiness, the absence of people. There is something vaguely futurist about these machines for mysterious purposes, brimming with internal logic but opaque to the viewer. Surrealist and dreamlike, these dioramas have extraordinary detail, demonstrating the passion of their maker. Suggesting movement, but frozen, they are empty, figureless spaces. Like scenes from a recently deserted alien city.

Downton is making inquiries into the sources of forms. Not design sources like platonic solids or classical precedents, but the real sources of form-life, and the diverse experience of it. These are archetypal, emotional, memory machines. His research into the processes of design has centred on observing his own making of these models, often designed intuitively without the aid of drawings. He has written two whole books on them, Studies in Design Research and Design Research .

These works are not about representation but ends in themselves. Not designed to exemplify anything or provoke architects to 'build this way!,' but to affect the viewer's life, to enter our dreams and delight us with their odd familiarity. In short, they are artworks.

Andrea Mina, Associate Professor in Interior Architecture at RMIT, enjoys the hunched posture people take to look at his miniatures. He hopes they are transported to another plane of reality, if only for a moment. Mina says his objects are at 1:1 scale and not models at all. Miraculously tiny, he uses a jeweller's eyepiece to glue cat and dog hair into miniature architectural lattices that hold together fragments of marine-life shell and other found organic objects. They encourage close inspection. Even more alien than Downton's empty dioramas, these could be crushed underfoot, destroying a universe in a moment. Grand, mysterious monoliths at a small scale.

There is something about being in the presence of sculpture that a 2d representation just can't match. It has presence. Kerstin Thompson said of modelling 'as a miniature world it is profoundly attractive and fascinating'. Participating firms who chose not to exhibit a physical model were less interesting.

Are Downton and Mina doing craft, not architecture, with these highly finished pieces? That depends on the purpose of the modelling, says Glanville. He argues Downton and Mina are doing in order to learn something about architectural propositions. Using the model as an investigative tool. Asking questions, rather than telling people things. In one of his essays in the catalogue, Downton describes models as machines (models for ). Downton's are machines for speculating about the world. They are modelling ideas about architecture, perception, emotion... Mina's are machines for learning about space, investigating thresholds between inside and outside, tools to leverage understanding.

Where does architecture reside?

Burry and Ostwald are not designing but looking at technique. They have taken a given design-for Ostwald it was Le Corbusier's Villa Savoye and for Burry it was Gaudi's Sagrada Familia-and investigated techniques for designing. Downton and Mina are absorbed in their own architecture, and analysing it. They are designing and making architecture.

But surely, you say, Downton and Mina are not making architecture because they are not working towards buildings? So where does architecture reside? In the catalogue, Ostwald discusses Peter Eisenman-'For Eisenman in the 1990s, the architectural drawing is the 'matter of architecture' itself and it contains more than a model'.

It seems there is strong agreement that architecture does not reside solely in buildings. This exhibition may even be more a display of architecture than a tour of built works! It is even suggested in one of Downton's catalogue essays that buildings may represent the model at a larger scale, rather than the reverse. Several papers and presentations touched on the idea that the model may be the purest expression of the architecture. Alex Selenitsch in the catalogue: 'the end [model] often shows the architecture with more clarity and power than the final intended building which can easily be dissolved through the design process by the demands of commerce and weather.' Greg Burgess loves his models 'like children'. John Wardle likes to tell his clients that the presentation model of their project is on loan to them until the building is finished, when it is to be handed back in return for the full-scale version. According to Glanville, models are smaller and cleaner than buildings. 'We make them so we don't have to build the thing.' Architects use models, because the building has to be built by agents. The very idea of the professional designer only exists with this remoteness-craft work has no such distance. Where does architecture reside? In the words, dreams, ambitions of architects and clients. In what we choose to see in the bricks and mortar, how we analyse and interpret and understand it.

Conclusion

The exhibition consists of four different takes on modelling, by four academics. And a fifth take by 22 architecture firms.

Mark Burry's non-models are the key-they are conceptual models, models for moving into the future. The secret to building the computer's geometry is to model only what can be built. The failure represented by Michael Ostwald's experience of rapid prototyping is a failure of strategy-a failure to know the maker and her tools, whether she is a 'homo sapien' or a machine. To model 'with the grain'. Even CAD software has a grain-an invisible grid that likes orientation in xyz. Balsa has a grain. Card has its own characteristics-but not the same as a building material.

The ARC grant was to research computer models and physical models, and the academics represent two of each field-Andrea Mina and Peter Downton are interested in physical models, and Mark Burry and Michael Ostwald in digital modelling. They set about determining a baseline for how architects use models, although the grant did not give money for exhibitions (and this one was done on a shoe-string).

Would architects be brave enough to exhibit dog models? Yes, and it's wonderful to see.