Aesthetic Fitness

Think of the number of practices that pride themselves on being “research-led”? And how much they let us know about it? One could be forgiven for thinking the point of in architecture is to produce theories designed to suit the technologies of the time having the greatest potential for mass rollout at minimum cost – and then leave the application to others. And that the sole driver is to display an awareness of which side whose bread is buttered. This is one way of understanding the continual supply of proposals to further automate the design and construction process. Even if unsuccessful, the notion that this is how things should be is kept in our minds. [c.f. The Autopoiesis of Architecture: Vol.1 Chap. 3.7 – Styles as Research Programmes, The Autopoiesis of Architecture: Vol.1 Chapter 2.4 – Architectural Research, The Massively Big Autopoiesis of Architecture Post, The Autopoiesis of Architecture Vol. I]

Contemporary Techniques in Architecture is part of the AD series put out by Wiley-Academy at the turn of the century and is one of four recently given to me. The other three are New Babylonians, Reflexive Architecture, and Emergence: Morphogenic Design Strategies. Stamped inside the cover of the book is the name Garry Emery Design Pty Ltd. This may be the same Garry Emery regarded as Australia’s top graphic designer, whose achievements have been acknowledged by an honorary doctorate from RMIT University and an honorary Bachelor of Arts from the Sydney Graphics College, who’s an Adjunct Professor at RMIT University and Deakin University and whose contribution to architecture and urbanism has earned him the President’s Award from the Royal Australian Institute of Architects? If it is, I can understand why these books ended up in a secondhand bookstore. I hope they work the same magic for me.

The books were published 2000–2002 and now form a historic record of sorts, of who was thinking what and when. It’s all history now, so it’s time to look at them and ask just how prescient they were. Did the highlighted techniques change the world? From just one glance outside my window, I think not. Twenty years ought to be sufficient to determine if something was the zeit-iest geist, the avant to any garde.

It’s been a hundred years now since craftspersons ceased to have a meaningful role in the production of buildings. They were slow and there weren’t enough of them anyway. In his Making Dystopia, James Stevens Curl notes that by 1935 Modernism was accepted not only in Germany but in Czechoslovakia, The Netherlands, Austria, The Soviet Union, Hungary, Poland, Belgium, Brazil, Scandinavia, France and the United States of America. Contrary to what Curl would have us believe in the early chapters of his book, it’s not as if everyone was swooning over Gropius’ factory-lite aesthetic or Le Corbusier’s FIVE POINTS. People called clients wanted buildings built and saw a quicker and cheaper way of getting the job done. It’s that old Cost-Time-Quality triangular chestnut and, when a way of building offering cost and time advantages comes along, well… something’s going to get compromised. To see the spread of Modernism as some sort of stylistic victory, and history as some flow of styles, is to serially avoid seeing the elephant. It is fact that a sea change in the way of building occurred over about the ten years 1925-1935. [c.f. The Persistence of Beauty]

It may have been 100 years since craftsmanship was shafted, but it’s only been fifty years since how buildings are constructed ceased to be a major determinant of building design. It’s been about twenty since architecture hasn’t even required a built reality and, as these books now prove, we’ve had two decades of contemporary techniques but no comparable sea change to show for it. Why then the hype? What function does it serve? Is it just to tell us that design is on the case? And working to make the world better for us? That’s nice. Unless it’s not.

Contemporary techniques will always excite publishing houses and, regardless of any potential benefit to architecture, construction or humanity, will serve to reinforce the by now well-established division between design and production. None of this media churn works towards perfecting prototypes of actual buildings that might do us for a century or three. We have low expectations anyway, if any, and this leaves us in a continual state of imperfection reinforcing the position of designers in a system where there are no people in the middle. Next on the list out of architecture after craftsmanship and construction come materials and materiality. Consider the following examples – remembering that we haven’t even got past the introduction to the first book yet.

From left to right, they show 1) how modulations in a single surface can produce differences in spatial configuration and internal lighting, 2) and 3) how different porosities can produce “aesthetic effects” of various transparencies and lighting variation in one surface and 4) how differences in the gradient of a component can produce different lighting levels. All four “contemporary” techniques are working towards eliminating the difference between walls and windows, while the first works towards eliminating the difference between walls, floors, ceilings and windows. This is all necessary groundwork if someday we are to accept 3D printed buildings and the new improved grey goo. Research likes Funding.

But before we ponder the societal implications of an architecture that functions as media content reinforcing the architect’s titular role as head of a system of production, we need to come to grips with the language. The following words head the first essay titled Deleuze and the Use of the Genetic Algorithm in Architecture. The namecheck doesn’t make me anticipate a new Formalism. [c.f. A New Formalism]

Said organisational processes seem to want to be understood as mimicking the organisational processes of Nature to create things intensely producible using the organisational processes of Industry. It’s the design vs. production divide reappearing yet again, and trying not to look like the Form vs. Function divide, or the Architecture vs. Building divide or, going back still further, the Art vs. Science divide. [c.f. Divide & Conquer] Extrapolating, the endgame seems to be a naturalistic architecture 100% made by machines.

Genetic algorithms save the designer some brainwork by computing millions of possibilities [but, and who’s to know that’s not what designers intuitively did anyway?] and presenting them for possible selection according to their aesthetic fitness. Despite the evolutionary terminology and analogies, this is not how Nature works. Or rather, Nature only operates according to (visual) aesthetic fitness if it offers some evolutionary advantage. Nature at least admits the existence of other types of fitness. [c.f. The Snail is Not Trying to Look Beautiful, Architectural Myths #8: Learning from Nature]

Genetic algorithms require that the solution space for a required building first be defined in terms of virtual genes for manipulation. This alone is many design decisions in one but “The architect should add points at which spontaneous mutations may occur” is another. Moreover, “unless one brings to a CAD model the intensive elements of structural engineering, a virtual building will not evolve as a building.” The author frets a bit about this, but not for long.

The gist is that the automated process must be designed to produce the right outcomes – i.e. those that delight. This requires setting the parameters for the desired solution space so one doesn’t get a jellyfish when one wants a giraffe, but a bird might be okay even if a bird-fish was what we wanted all along. It all seems like a lowering of expectations for surprise has no meaning if one has pre-set the terms for what counts as it. The bigger question is why should shock or surprise continue to be a determining characteristic of architectural production at all? In this sense, these contemporary techniques don’t really replace any old ones but merely offer a new way of achieving something known and questionable. “All potential configurations” turns out to be a very narrow set.

Bookending the book is an article calling for the “crash testing” of buildings. In the case of vehicles, crash testing can now be simulated with a high degree of precision as long as all known variables are considered. There’s that phrase again – all known variables. The author suggests that this technique can create ‘an instrumental technique for the discipline’ capable of producing highly performative effects that can validate the effectiveness of their generative techniques.

So at the front of the book we have architects wanting to be more like God, setting up the conditions for populations to run wild, and at the other end we have architects wanting to be more like Industry, fine-tuning the manufacture of the product and the reactions to the product but without ever questioning the need for the product. Since structural and envelope performance can already be simulated to a high level of precision, the beneficial effects of crash testing must be ones users don’t know about yet, assuming of course that users haven’t already gone the way of craftspersons, construction and materiality.

Between the beginning and the end is Emergent Structural Morphology that develops theoretical frameworks and computational environments to relate computational thinking to the design process. Again the same stance. The more the design process can be framed in terms of a set of parameters that an anticipated production process is comfortable with then, I’ll wager, the more efficient/profitable it will be. Again, the user and any possible benefit they might receive is not mentioned. I’m not sure what the illustration heading the article illustrates.

This next description reminds me of a mid-eighties machine translation one struggles to make sense of despite all the nouns being present and connected by a familiar grammar.

In MoSS (Morphogenic Surface Structures) the designer specifies the grammar and guides surface growth by defining shaping forces and boundary conditions. Again this seems like what designers did anyway. The MoSS programme did produce a free-form honeycomb truss in which vertices of generated surfaces are joined to form an adaptive three-dimensional cellular structure. The structures may be produced with sheet materials or as a matrix to be filled with structural foam. That was 2002 or prior. I couldn’t find any recent examples of applications of free-form honeycomb truss but the field of inquiry still exists at various universities.

GENR8 is a user friendly interface for evolutionary algorithms. Users can interrupt, intervene and resume the business of form breeding but grammatical evolution automates the [now grunt, rote, distasteful, unsatisfying, uncreative] work of constructing grammars. GENR8 has one mapping process that maps a genome to a grammar, and another – a phenotype – that interprets the grammar and constructs a surface. Again, one could argue that this is what designers did anyway. Again, there is that same concern for the continuing role of the designer.

Twenty years on, Carbon Tower Prototype maintains a strong internet presence due to a virtual demonstration prototype of a 40-60 storey building. Compressive loads are carried by an array of vertical columns and cores and the structural skeleton is a composite mesh formed of continuous pultruded sections is hung from it and (along with Kevlar cables) supports the floor slabs.

Here’s what it looks like.

Here’s the only part I have a problem with.

The physical and technical demands placed on building materials are nothing compared with those faced by materials in the aerospace and defence industries, two industries that happen to be the world’s most evolved and budgeted – just saying. [Research likes Funding.] I’m all for using materials so their inherent characteristics are utilised to the fullest but I fail to see the sense in contriving a building to make that specific point, unless it’s to show the military-industrial complex that one is open for business, a bit like the Dymaxion Houses designed to keep the share price of aluminium up and the aircraft manufacturing industry ticking over.

By 1949 the world was not housed in elegant boxes of travertine and onyx but in simple shells constructed roughly and quickly from sand, lime and aggregate stuccoed over. We’re fast approaching The Two-Decade Test for Carbon Tower Prototype. I’ll wager that in 2022 we won’t be living or even working in buildings constructed anything like it. One of those inherent design requirements was probably cost. Using aerospace materials will only be worthwhile if those materials have no properties that aren’t being fully utilised. Aerospace materials will always be lightweight or, if they aren’t, will justify their weight through thinness or availability. Building materials simply aren’t subject to the same constraints. To be fair, aircraft aren’t made of concrete and tanks aren’t made of GRP. I’m all for performativity in buildings but we need to judge by appropriate criteria. Drawing our architectural design aspirations from the natural world and drawing our architectural production aspirations from the world of industry aren’t helping. What peacocks and architecture have in common is an ornamental aesthetics that merely signifies fitness for reproduction for its own sake and no greater purpose. We overestimate evolution.

• • • 

The plan had been to have one post per book but I see it’s going to take more time. I’ve only mentioned three of the fourteen articles in the first book and have yet to find out what Interactive Opportunities are or were, what a Vigorous Environment is or was, what Potential Performative Effects could possibly mean or have meant, what Lumping is or was, and what Cecil Balmond had to say twenty years ago about The Digital and the Material although, given that as the title, I expect it will be on-theme.