Investigation
Informed Space
Previous Chapter | Next Chapter
A Culture of Virtual SpaceIn the 1980s, a new place surfaced, almost literary, with the introduction of digital technologies. It was a place constructed by space altering technologies, like the fax-machine, the modem, the mobile phone, the satellite transmission and the personal computer. Technologies that were at the forefront of a digital revolution that hit the consumer markets as they anxiously awaited the arrival of George Orwell’s 1984. It was the creation of a new place that was neither entirely inside the private sphere of the family, nor entirely outside in the public. It was a place that was process driven, collaborative, synthetic and parametric – a place with just as many ways in as ways out. You could unplug from your digital workspace cubicle, drive 25 minutes to your tract home, and then plug-in to that same workspace at home. It was a place where geography was history, where distance disappeared into the milliseconds needed for a file transfer or a faint echo on the telephone from the satellite hookup. It was a place from where space apparently had disappeared – but only apparently. Space was still there, more present and to more use than ever before.
For the architects, it was also much more than the introduction of new hardware in their arsenal. For them, it was a movement into a post-industrial era,(1) accompanied by a digital twin, who was soon to untie politics, ideologies and aesthetics from its physical materiality. It was a move that ‘shifted’ the attention away from what for long had been the physical domain of the architects, toward a new virtual materiality of ‘information space’ that was dynamic, active, and which had greatly accelerated the ‘symbols in space’ that Robert Venturi found in Las Vegas.
Computing SpaceIt is noticeable, how the first computers in the service of architecture were focused on data management rather than on design manipulation. They were more the machines of spatial symbols than of spatial forms.
As early as 1977, the professor of design science John S. Gero quoted two compendia with 452 and 925 computer programs in environmental design.(2) Besides the more general office management functions, they were either for analytical work in the early part of the design process – siting studies, user surveys and area programming – or at the very end of the design process – providing the architect with two- and three-dimensional renderings. Out of the 1377 applications surveyed, about one tenth could be used for strictly two-dimensional site planning, while none of the applications could be used for the manipulation of three-dimensional objects. This should, of course, come as no surprise since the survey was made in 1977, but it is nevertheless worth noting that the introduction of computers in architecture was diagrammatic and programmatic, and not for formal aesthetic expression.
One use of the newly introduced computer resources was for the method of space planning. In the same 1977 book as Gero’s survey of computer applications in architecture, a young William J. Mitchell, later to be professor of architecture and media arts and science at MIT, introduced procedures of space planning that could distribute work areas in large office buildings with regard to needs, inventory, distance, key functions and so forth. The use of the computer application was exclusively analytical and diagrammatic – it had no formal potential besides creating a patchwork of distributed office space. Mitchell and software developer Jeffrey M. Hamer, from a private office in Los Angeles, broke down the process of space planning in to five steps: 1) information gathering and database creation, 2) data analysis, 3) design development, 4) design documentation and 5) post-occupancy activities.(3) These five steps were again either located in the very beginning of the design process, or at the very end – either gathering the operands and constructing the design space, or presenting the design product in perfect line perspectives. Mitchell and Hamer mentions several systems: DOMINO, which dated back to the early 1970s, had a constructive approach and could distribute space according to area, shape and adjacency to other areas. Another system, GSP (General Space Planning), was able to handle ‘objects’ with different shapes and relations to other ‘objects’, and did this through operations, like scanning for possible object locations, boundary comparison between objects and perimeter adjacency to another object.(4) What these early applications had in common was that they processed architectural data by specific rules and presented a result, which would form the basis for a later phase in the design process. They were machines of actualization – they were fed with data, which they handled by rules that often were too complicated for the architect or engineer to handle by himself. Therefore, if these systems were not for interactive formal design, they were for immediate design programming. They were systems to handle all the operands in a design space and were in their essence spatial.
We could argue that these first systems were not developed to accommodate the immediate sculptural design process, but rather to construct a design space, which could contain all the different operands of a given project, be it issues of economy, organization, zoning laws, area disposition, light distribution, height limitations or structural engineering. All this may seem trivial some 25 years later, but it serves to show that the first use of computer application in architecture was diagrammatic, and that these applications were based on an understanding of a parametric design space.
The Case of AutoCADJohn Walker, one of the founders of AutoDesk, proposed in September 1983, in an internal paper entitled “Low Rent 3D”, to advance the spatial capabilities of the application AutoCAD.(5) He wrote:
Many of our competitors, have or will be introducing 3D packages around the time of COMDEX. If we do not have a credible response to queries about 3D, we may be in trouble selling our package. While all drafting is 2D, and almost all users will spend all their time with AutoCAD working in 2D mode, 3D is important more from a marketing perception standpoint than a technical one.(6)
It is clear that the architectural design process was not the primary aim for the development of AutoCAD’s spatial capacities. Rather, spatiality and 3D was a ‘sales gimmick’ that was needed to better push their 2D software packages, as Walker made clear: “The impact of rotating an object in 3 (D) space at COMDEX is many times that of zooming in on a flat drawing … Just think of the difference in our COMDEX literature of ‘Now with 3D’ versus ‘Now with crosshatching and dashed lines’.”(7) The development of AutoCAD progressed as follows:
Release Version name Release date Spatialization
1 Version 1.0 December 1982 2,5 Dimensional
2 Version 1.2 April 1983
3 Version 1.3 August 1983
4 Version 1.4 October 1983
5 Version 2.0 October 1984
6 Version 2.1 May 1985 3D Level 1
7 Version 2.5 June 1986
8 Version 2.6 April 1987 3D Level 2, 3D Line/Face
9 Release 9 September 1987 Spline Curves
10 Release 10 October 1988 3,0 Dimensional, UCS, B-Splines
11 Release 11 October 1990
12 Release 12 June 1992
The effect of John Walkers entrepreneurial adventures had, needless to say, a profound influence on the way architects design today. What began as a nice sales feature, advanced as a 2D application through different levels of spatiality, to a full 3D spatiality with space curves, wire frames and solid Boolean modeling. Between Walker’s proposal for 3D Level 1 (2,5 D) and the full 3D capacities of Release 10 in October 1988, was a period of 5 years in which, the virtual space of the computer developed from being a projection that substituted the architect’s eye, to being a fully spatial modeling environment. It was not only for the final photorealistic renderings, but also for creating a place to structure the entire design process.
This development came to define Computer Aided Design (CAD), which was present in Walker’s statement from May 1986:
Computer aided design is the modeling of physical systems on computers, allowing both interactive and automatic analysis of design variants, and the expression of designs in a form suitable for manufacturing … As computers become more powerful, and tools evolve that allow us to build larger and more complex software systems on them, the artificial barriers that keep us from modeling the real world will fall.(8)
The focus of AutoCAD had changed. From being a clever drafting system with the ability to display rotating objects at COMDEX, it had transformed into a modeling system – with the potential to construct design spaces with a large degree of simulation or prediction. In the words of Walker: “computer graphics has nothing at all to do with CAD, except as the servant of design, simulation, or presentation.”(9) So, forget the nice graphics, the easy perspectives and the photorealistic renderings for a moment, the conceptual change to the architectural design process was greater – by far.(10)
John Walker writes:
Now we’re at the threshold of the next revolution in user-computer interaction: a technology which will take the user through the screen into the world inside the computer – a world in which the user can interact with three-dimensional objects whose fidelity will grow as computing power increases and display technology progresses. The world inside the computer can be whatever the designer makes it; entirely new experiences and modes of interaction can be explored and as designers and users explore this strange new world, they will be jointly defining the next generation of user interaction with computers.(11)
We could argue that the ‘drafting system’ was intended to ease the manual process for depicting the architectural idea in the mind of the architect, while the ‘modeling system’ relocated that idea to the design space of a spatial application. In so doing, it offered the architect a design space ‘outside’ his own mind. In the first spatiality, the design space was still inside the architects mind and projected onto the world. In the second spatiality, the design space in the mind of the architect was overlapped by the idea of a model. In the third spatiality, the design space took the form of an environment that the architect is immerged into as the creation of a virtual reality. In other words, the progression of this media has moved between three places. From being inside the architects mind as an representation of his idea, to being outside the architects mind in the model that can be shared among others, and finally creating a synthetic place for the architect to be within his very own idea as a design environment, where everything is collected and visualized.
Programmatic and Formal LimitsAs the new virtual place has become real, and we have the proper applications to construct that reality, what do we then do with it?
One position could be the architect Michael Speaks’ critique of the ‘formal limits of the American Avant-Garde’, and what he calls a ‘just there’ modernism. Speaks argues that the American avant-garde cannot escape the constraints of form, why it lacks the connection to what is ‘out there’. This makes the produced architecture a formal exercise that is only confronted by the real, when it is finally build.(12) Opposite to the American avant-garde, Speaks positions a ‘Dutch’ programmatic and pragmatic approach to design. This approach confronts the real from the very beginning – often by letting reality define the very frame in which the design process takes place. Speaks critique could characterize two different way to construct and use a design space, even though it is a simplified and rather categorical view. Architects like Peter Eisenman and Greg Lynn could be said to manipulate their design spaces in a formal way, through applications that support a gamut of geometric operations. Eisenman has in the back of his book Diagram Diaries a list of 24 formal operations by which, he designed his projects, while Lynn has demonstrated how to use the character animation software MAYA for the creation of ‘animated form’. Opposite Eisenman and Lynn, we could position the Dutch design group MVRDV, which never or at least very rarely demonstrates formal manipulations as a source for architectural design, but instead use the construction of datascapes. MVRDV collects data from what is ‘just there’ and manipulates that data to construct a spatial argument that eventually turns into architecture. In their book FARMAX from 1998, the group even shows an example of a custom application Castle Maker from 1995, which optimizes the constraints of light and program, and subsequently suggests different solutions to problems regarding building volumes. This is a design method that has a striking resemblance to the diagrammatic applications of the early 1970s, which calculated shadows from building volumes, thus signaling a kind of diagrammatic nostalgia that we could compare to the mechanical nostalgia, which I described earlier. In this way, MVRDV has constructed design spaces with operands like noise distribution on highways, line of sight in Amsterdam, floor area ration, population density, traffic patterns and fire escape regulations.
Another position from where to observe the use of an ‘informed’ design space, could be architect and critic Sanford Kwinter’s suggestion to ‘leap in the void’. He writes:
… Let research follow the real, let it be encumbered by no moral and aesthetic preconceptions, and design will follow as an integrated process ... The late 20th century may one day be known as the dawn of the algorithm. If so, we wish to be the first to embrace the new rationality that sees space and matter as indistinguishable, as active mediums shaped by both embedded and remote events and the patterns they form.(13)
Kwinter’s critique is a suggestion to leap into a design space embedded with operands of the real, but also to do so without preconceptions: without the formal, programmatic or dogmatic constraints of architecture, neither to the initial method, nor to the final result.
Based on Kwinter’s strong suggestion we could ask where to find that informal component that Kwinter suggests. We could look at two projects that never got further than the constructed design spaces: Greg Lynn’s Port Authority Gateway from 1995, where he made extensive use of the application MAYA from Alias Corporation(14) and MVRDV’s Noise Scape from 1997, where they used a traffic engineering software from the Dutch firm Dgmr called SRM2.(15) Both Lynn and MVRDV found their informal components in the real siting of their project – Lynn in the movement of people at the location, and MVRDV in the noise generated by cars on the highway. However, there is a crucial difference hereafter. If we ignore for a moment that they reached different formal expressions at a later stage in the design process, we could look at the way their methods were structured relative to a design space. Lynn embedded his design space with data from the real, which stayed animated or alive throughout the process, until he stopped the process, and his animation became ‘freeze frame architecture’ – an architectural manifestation of an instance in space. MVRDV on the other hand embedded data from the real, just as Lynn, but embedded them into a design space with a much more rigid structure, defined by precise rules and calculations. They only accept the informe to a certain point in the process, after which they turn off the data stream to do their calculations. While Lynn keeps room for the informe in his methodological structure, which may easily accept inaccuracy and interruptions, without distorting the entire argument, the case is different with MVRDV. They depend on the outmost accuracy in the way their application handles their datascapes: if the data or the computation is false, they have an invalid design product. In other words, MVRDV’s method is just as limited as Eisenman’s and Lynn’s, just at another place in the design space. A place that we, with reference to Speaks, could call ‘the in-formal limits of the Dutch avant-garde’.
When Greg Lynn presented another of his design spaces at the AnyWise conference in Seoul, South Korea in 1996, it sparked a discussion, which pinpointed the critique that Speaks raised above. After Lynn’s introduction to his project House Prototype, he had a discussion with the architects and theorists Ignasi de Solà-Morales and Jeffrey Kipnis:
//Lynn: In the simulation that ran on the analysis of the site, I put in over 100.000 particles, every one of which has more than a dozen parameters that can be set to define how they will interact with each other. Each particle has parameters that can be attached to any of the five fields. Each of the five fields has over 1.000 parameters for generating forces on the site ... The internal complexity of the system and the fact that there is no one-to-one quantitative alignment with any aspect of the project make it a very creative process ... We evaluate the process continuously. This is just a design method. I would never show this as a way of validating the house because I am involved in a completely unscientific enterprise.
Ignasi de Solà-Morales: I am not convinced that it should be a design method. First of all, you didn’t show a project, a final form, and second, you have only shown that it is some kind of trial, not something that ends with a decisive project. Your own input is part of the decision making in this evaluation. Everything is involved in strategies that are not objective but that you evaluate …
Jeffrey Kipnis: Let me hold you accountable to the question, Greg. Because you stay at the level of dynamic animation, we could be fascinated by what we see, but because you do not resolve it as a fixed static object with materials, structure, and construction, at which point we see its real consequences, we’re left fetishzing the video rather than really understanding its design consequences. Is this true or not? …
Lynn: I want to resist answering that question. In other situations in which I have shown material like this, the response has been ‘well, are you saying architecture has to move in order for this to be an interesting design approach?’ I would say no …
Kipnis: You say no, but you do not show us what happens when you take the motion away.(16)//
What provoked the panel was that Lynn’s project first seemed to replace the architect with an automatic process for generating form. However, the architect designed this process, so Lynn operated from another place and in another space than traditional architects: He accepted the limits of his procedures, which were ‘completely unscientific’, as they were nothing but formal unfoldings. Therefore, we could argue that Lynn designed the space in which, the process took place rather than the traditional objects of architecture. Secondly, both Solà-Morales’ and Kipnis’ remarks show that they presume that the space of the design process equals the space of the real building: if the design space is animated, then the final building must be moving. This is of course not the case. Just as Schindler’s unit system only is visible through its influence on the building, and not left after the carpenter has removed all the nails and strings from the construction site. Therefore, if the space of the design process on the one hand, has a value that goes further than ‘just being a design method’, and on the other hand, not necessarily has very much to do with the final real building.
I would argue that we are dealing with a design space. A design space that creates an informed place for integrating how it was done with what was done.
Previous Chapter | Next Chapter
(1) The leader of MediaLab at MIT, Nicholas Negroponte, gives a broad historical account of this ‘phenomenon’ in Being Digital from 1995.
(2) Gero, J.S., ed. (1977), Computer Applications in Architecture, London: Applied Science Publishers, p. 5-7. The categories in the study of 452 applications were: feasibility study, architectural programming, relational planning, site planning, two-dimensional graphics, three-dimensional graphics, cost control, environmental control, circulation analysis, text manipulation, project control, office management and, finally, evaluation. In the study of 925 applications the categories were: management and project control, accounting and cost control, information handling, quantities and schedules, site and land use studies, plan layout and analysis, structural engineering – frames, structural engineering – constructional elements, services engineering – pipe work, ducts and transport, services engineering – heating and cooling, services engineering – lighting, sound and electricity, graphics and, finally, integrated systems.
(3) Mitchell, William J. & Jeffrey M. Hamer (), “Space Planning,” in Gero, ed. (1977), ibid., p. 29-30.
(4) Ibid., p. 54-5.
(5) Walker, John, ed. (1994), “The AutoDesk File: Bits of History, Words of Experience,” from http://www.fourmilab.com/autofile/. AutoDesk is the company that manufactures the applications AutoCAD and AutoDesk VIZ (3D Studio VIZ), which have become standards in spatial modeling.
(6) Ibid., p. 207. COMDEX is a yearly five-day technology convention in Las Vegas, which is labeled as ‘the global technology marketplace’.
(7) Ibid., p. 207 & 209.
(8) Walker, John (1986), “CAD: the Heart of Computer Science” (May), in Walker, ed. (1994), ibid., p. 308-9.
(9) Ibid., p. 308.
(10) Another development that we should not forget is the development from ‘projected spatiality’, via ‘static spatiality’ (AutoCAD release 10), to ‘dynamic spatiality’ (MAYA).
(11) Walker, John (1988), “Through the Looking Glass” (September), in Walker, ed. (1994), op. cit., p. 446.
(12) See Speaks, Michael (1998), “It’s Out There … The Formal Limits of the American Avant-Garde,” In HDA Dokumente zur Architektur 10 (theme: Other Space) (1998). I do not entirely agree with this point.
(13) Kwinter, Sanford (1998), “Leap in the Void: A New Organon?,” in AnyHow, ed. Cynthia C. Davidson (1998), Cambridge, MA: The MIT Press, p. 27.
(14) See http://www.alias.com.
(15) See http://www.dgmr.nl. An application called SRM1 is available.
(16) Davidson, Cynthia C., ed. (1996), AnyWise, Cambridge, MA: The MIT Press, p. 107-112.
|
|
