5Introduction

Towards a ProgrammingCulture in the Design Arts

Pratt Institute School of Architecture,Carbon-fibre chandelier studio project,autumn 2006In this studio led by Professor Mike Silver,students bent for scripting was applied in aproject that developed new software tocoordinate the movement of a CNCmachines rotating bed with the controlledtrajectory of its servo-controlled hotwire.Here, foam shapes cut on a CNC hotwirefoam-cutter were used as moulds for hand-laid carbon-fibre panels.

In his introduction to thisissue of 4, guest-editorMike Silver celebratesthe flexible language ofcommands and logicalprocedures of computerswhose creative potentialhas until now beenundervalued inarchitecture. He explainshow the happy accidentof late 1990s blobarchitecture is now givingway to a focus onprogramming andcomposing new code,which promises togenerate new andunprecedented modes ofexpression.

6Chandelier fabrication process using epoxy resin,carbon-fibre and Nomex drapes over a CNC foam-cut mandrel.

7

8Proprietary script for generating ruled surfaces.

Ruled Surface

Foam Block

Ruled surface.

You write a few lines of code and suddenly life is better for ahundred million people.

Charles Simonyi, inventor of Microsoft Word1

The first architects to use computers were interestedprimarily in maximising the efficiency of conventionalmodes of production. Designers working on the Worlds Fairin 1964 used a primitive calculating machine to build theUnisphere, and around the same time Eero Saarinenengineered the complex reinforced-concrete shells of his TWAterminal using early structural-analysis software. During the1980s many architectural practices employed the firstautomated systems for drafting construction documents, ascomputers became cheaper and more readily available. In the1990s the buzz surrounding blob architecture embodieddesigns obsession with digital media at the brink of a newmillennium. This trend has now given way to the more urgentproblem of process and code. Many of the contributors to thisissue have expanded on a programming paradigm originallyexplored at MIT by the artist and graphic designer JohnMaeda, whose protg, CEB Reas is featured in this issue.

Not surprisingly, architecture has lagged far behind thesepioneers with creativity and experimentation taking a backseat to the priorities of disciplinary continuity, history andfunction. At MIT, architectures foray into programming wasmostly restricted to the computer replication of known formsand building types exemplified in the work of William Mitchelland Richard Freeman, and the shape grammars of George Stiny.As Greg Lynn has pointed out, these systems were merely anextension of a previously delineated and closed set of potentialRuled surface generated from proprietary script by Japhy Bartlett.

9forms whose characteristics can be stated in advance by anideal mathematics.2 Today, programming in architecture hasbecome a much more open process, one that is inspired by thecapacity to generate new and unprecedented modes ofexpression (see Malcolm McCulloughs essay on scriptedspace). For many architects coding has become the formal andoperative focus of building itself.

Universality = DifferenceIn the 1990s, the association between computers andarchitecture was marked by the proliferation of amorphous,curvilinear forms. But this link is far from necessary orinevitable. It is, in fact, a happy accident precipitated by pre-computational experiments based on folding and thesubsequent appropriation of operationally specific modellingsoftware. How, then, do we determine the relationshipbetween computing and architecture? What forms, practicesand techniques will seem most relevant if the internalstructure of software and the theories that drive itsdevelopment are changed at the same time? Can we evendefine computational architecture? The answer to this lastquestion is that we cannot. As Ingeborg M Rockers essay onpage 16 shows, computers are universal machines. Unlikeclassical machines (clocks, steam engines and tin-openers)they can perform a wide variety of tasks without significantchanges to their physical design. Through the writing of newprograms, very different operations can be executed on asingle device.

Where the information theorist defines universalitythrough the production of difference, for architects it holds

exactly the opposite meaning. The stripped-down andmonolithic style of international Modernism was thought tobe universal for all places and times. Miesian universal spacewas constrained by a homogeneous and infinitely extendedgrid, very much unlike the complex, convoluted matrix ofinterconnected microcircuitry that constitutes todays digitalnetworks. The universality of a programmable computer istherefore measured by its degrees of freedom. What makes itunique, as a device, is the flexible language of commands andlogical procedures that can instantly transform it from onefunction to another. With existing software, designers areoften forced to conduct experimental work using fixedprotocols originally developed to solve the visualisationproblems faced by aircraft designers and Hollywood film-makers. Even in these disciplines, staff programmers are busycreating new codes for challenges that are beyond the scopeof a particular product. Not surprisingly, this has also becometrue for the early advocates of software appropriation inarchitecture. Faced with increasingly complex commissionsand a growing practice, many designers have turned to the in-house creation of proprietary code. Here, one immediatelythinks of Dennis Sheldens redevelopment of CATIA at GehrySystems, and Greg Lynns collaboration with Microstation.

Both examples suggest the possibility of freeing computer-aided design (CAD) from any one particular concept of formrooted in any one particular software. If indeed codes can bechanged to fit specific needs or developed from scratch, thenthere can be no such thing as a computational architecture;only possible architectures actualised by new programs.These programs can be both simple, generative codes that

Finished panels for the Carbon-fibre chandelier project.

Carbon-fibre chandelier project showing the movable light-reflecting ceiling shroud and retractable downlights.

10

produce complexity (see, for example, the articles by StephenWolfram and Karl Chu),3 or complex programs that helpmake difficult tasks easy. Programs have become pervasivemainly because their instructions can be applied to so manydiverse problems (for example, one piece of software in thestealth bomber actually counteracts the planes pooraerodynamics and prevents it falling from the sky even whenthe pilot deliberately initiates a stall). New systems offeedback and control, database-management protocols,genetic design algorithms and a whole universe of scriptingsystems now drive everything from a simple electrictoothbrush to the Internet.

Building Programs and Programs for BuildingThe ability to craft tools that address both the practicalchallenges of building design and the human capacity toimagine new forms is a fairly recent development. As specificprogramming languages become less mysterious and easier tomaster, home-made software will most likely become afamiliar part of design culture. This move by the designcommunity to take an active role in the production of codetranscends the limitations of prefabricated software tools.

To a great extent this shift is becoming a necessity. Large,unanticipated gaps in the computer-aided design andconstruction process can only be bridged by the creation ofspecial software. Many of these tools cannot be anticipated inadvance by outside developers since the most imaginativeprojects begin with ideas that exceed ordinary expectations. Inorder to connect different production processes, materials andfabrication devices, new tools are required. In fact, asarchitect Haresh Lalvani has demonstrated in hiscollaboration with Milgo/Bufkin (see his interview with JohnLobell), unexpected functions for standard digital-fabricationequipment can be created simply by writing code.

In academia, students and teachers have been increasinglydrawn to the possibilities of proprietary softwaredevelopment. For many architects programming has becomethe new drawing. In the carbon-fibre chandelier projectdesigned for the lobby of Pratt Institutes new School ofArchitecture,4 complex EPS moulds for hand-laid compositeswere produced using a large-scale computer-numericallycontrolled (CNC) foam-cutter. In order to precisely constructthe necessary 3-D shapes, new software was developed tocoordinate the movement of the machines rotating bed withthe controlled trajectory of its servo-controlled hotwire. Theability to link disparate material practices (foam-cutting andadvanced composite construction) would not have beenpossible without the new scripts.

With these examples in mind, Programming Culturesexplores the power of code by encouraging artists andarchitects to become more involved in the creation of home-made, task-specific tools. Through individual labour or bycollaborating with skilled developers, designers can harness

the power of universality. In this way universality supportsdifference in opposition to the sterile homogeneity commonlyassociated with rigid protocols and fixed procedures. Whatcomputation must now serve (as the work of Evan Douglisaptly demonstrates) are the founding concepts, intuitions anddesires that can only emerge from a varied and creativepractice. Rather than working through algorithms imposed onthe architect by an external agent or product philosophy,intuition and desire motivate artistic production whileencouraging the spread of individual creativity.

As the projects in this book suggest, it is only a question oftime before software development becomes an integral part ofthe building design process. Certainly, this will be anunprecedented moment in the history of architecture. 4

Mike Silver

Thank you to Evan Douglis and Dean Thomas Hanrahan at the Pratt InstituteSchool of Architecture for their support of the Universal Machines symposiumheld during the spring of 2005, which provided the original impetus for thisissue. Malcolm McCullough, John Lobell and Yee Peng Chia were especiallyinstrumental in provoking questions and inspiring discussion. ProgrammingCultures: Architecture, Art and Science in the Age of Software Developmentwould not have been possible if it were not for the support provided byHelen Castle, the staff of John Wiley & Sons and a generous grant from theGraham Foundation.

Notes1 Steve Lohr, Go To, Perseus Books (New York), 2001, p 2.2 Greg Lynn, Variations on the Rowe Complex, from Folds, Bodies and Blobs:Collected Essays, Books-by Architects Series, Bibliotheque de Belgique, 1998,p 212.3 Where Wolfram defines code as a generative abstraction that emulatespattern formation at a multitude of spatial dimensions within the universe(analytic) for Chu codes are used for the construction of possible worlds(synthetic).4 Carbon-fibre chandelier. Critic: Professor Mike Silver. Advisers: MarkParsons, Evan Douglis, Rob Langoni. Programming: Japhy Bartlett. Design andConstruction: Patrick Weise, Hanna Meeran, Felice Basti, Jakie Nguyen, MiltonHernandez, Genoveva Alvarez, Jonathan Lee, Julie Camfrancq, Michael Chase,Jean Keesler, Eddy Adu, Jong Kim, Tomoko Miyazaki. Research: Mel Sakor,Arta Yazdansets, Hasti Valipour, Kenneth Chong, Caitlin Duffy.

The Universal Machines Symposium This issue of 4 developed out of the Universal Machines symposiumheld at the School of Architecture at the Pratt Institute in February 2005.The symposium continued a long tradition at Pratt of combining interestsin theory and practice, thereby allowing students, teachers and practitionersto address emerging technological and social issues in architecture.

It is therefore satisfying to see that while several of the articles hereaddress highly theoretical issues on computation and architecture, justas many bring that interest into new forms of material practice.

The faculty and programmes at Pratt are diverse and address a widerange of interests, all of which combine to create Pratts unique strength.With the Universal Machines symposium and this issue of 4 we arecontinuing a long-standing commitment to design excellence.

Thomas Hanrahan, Dean, School of Architecture, Pratt Institute

11