“For the purposes of this short discussion, let us use the term ‘technologies of mediation’ in place of ‘representation.’ The problem with the latter term is that the re- of representation always lowers the discussion to a secondary status subservient to the content conveyed. The drawings, models, notations, texts, and calculations that the architect produces are not secondary to the building, but constitute the primary acts of architectural design as mediations translating between conceptual ideas, interpretive parties, and aesthetic articulation. . . .”
*image courtesy the author.
For the purposes of this short discussion, let us use the term “technologies of mediation” in place of “representation.” The problem with the latter term is that the re- of representation always lowers the discussion to a secondary status subservient to the content conveyed. The drawings, models, notations, texts, and calculations that the architect produces are not secondary to the building, but constitute the primary acts of architectural design as mediations translating between conceptual ideas, interpretive parties, and aesthetic articulation. Furthermore, technology is always at play in any system of representation, even if it is hidden in the disciplinary naturalization of technique. By shifting the terms toward “mediation,” many technological aspects rise to the surface.
A key thinker for this discussion is the German media theorist Friedrich Kittler. In much of his work, Kittler lays out a counterhistory of art and technology by paying close attention to when and how a technology of mediation goes through a transformation. One of his key points is that all mediation involves modes of storage, manipulation, and transmission.1 These distinct modes often require different mediums, media, and technologies. For instance, painting may be a technology that stores optical information by drying colored liquid pigment into a woven canvas matrix, but in order to transmit the optical information stored in a painting across physical distance, a reproducible print copy is necessary, which requires a different technology of mediation such as engraving. The relations between these two different technologies, that which stores and that which transmits, forms a mediated exchange, so as one technology changes over time, there are repercussions for the other. Kittler points out that in order to understand the entire system of mediation, one has to look beyond simply the art form that appears to hold the content and has to consider the relations between storage, manipulation, and transmission. Painting was transformed by the advent of photography, as it was previously transformed by earlier technologies of transmission such as etching and engraving. What photography replaced was not painting, but the underdrawing in painting—that is, the drawn composition that sought to fix spatial relations through a bounding contour line. This organizational drawing, preceding and under the pigment painting, was what engraving focused its mediation around. Light, color, and textures were much more difficult to store and transmit through engraving. (This is not to say that etching and engraving did not attempt to capture the light and texture of painting, and through these struggles generate a whole world of effects, but these unexpected excesses are for another discussion.) Photography in a way released painting to focus on the aesthetics of optical perception through patches of light and color without a reliance on underdrawing, thus opening painting toward techniques of manipulation that initially foster Impressionism and Postimpressionism.2
Let us now consider a few differences in the technologies of mediation as the discipline of architecture incorporates parametrically regulated digital models. I would like to suggest that the important change is not the one to digital computation. To work digitally is to describe continuity through discrete numeric calculations, something that architectural representation has been involved with to a degree ever since Renaissance architects shifted their technology of mediation from graphic to numeric computation.3 If being digital is not the key difference, neither is it the use of a model, as architects have used models to simulate design at a reduced scale for centuries. The third term, parametric, might be the locus of difference, but an entity defined parametrically is likewise not a novel condition. Parametric representation is fundamental to the calculus and differential geometry of the 19th century. If we want to step closer to the present day, we have the car manufacturing of 1960s France that, between Pierre Bézier and Paul de Casteljau, set down much of what was to become the contemporary computational definition of NURBS (Non-Uniform Rational B-Splines) curvature through parametric representation.4
One crucial point Kittler makes in a collection of lectures, published in book form as Optical Media, is that the computer is the first technology of mediation that automatically combines storage, manipulation, and transmission into a single system.5 This automated combination is a good place to begin to unpack the difference. The contemporary digital model is a hybrid mongrel of multiple histories of mediation wrapped into a single technology. Designers now have a model that can be manipulated through the geometric transformations of drawing—this is new. This model is automatically measured and stored through vectors normal and tangent to a surface, thus no longer coincident with the plane of projection—likewise new. Furthermore, these vectors are optically transmitted as a full-scale surface simulation parametrically regulated in real time, leaving no visual residue of the computational process—again, new.
To further clarify these technologies of mediation, it is not even “the computer” that we should be discussing, but instead the conglomeration of hardware components, software environments, interface dialogues, graphics and codes, display monitors, printers, routers, mills, and cutters. These all combine to form the technological system of mediation for contemporary architectural design, and it is crucial that the entire assembly be understood as part of a single technology of mediation.
An extension of this discussion is that, in theory, anything modeled in a digital space can be materially output through a 3-D printer or CNC mill. The physical material output might be a mess, or might fall apart because thresholds and tolerances have been crossed or a nondimensional surface self-intersects, thus producing significant problems in the physical realm, but all of this must be kept to the side. The change in our technology of mediation here is that the designer does not need to mathematically process a representation to usher a design toward construction. This translation is now handled automatically by the computer. This translation was a key component for a technology like descriptive geometry: “Get it flat so that it can be measured.” In a digital model, the plane of measurement is no longer the plane of projection. A design can be looked at from all sides, at constantly changing variable distances, with no single view privileged by the requirement that it be flat and measured on that side. The previous technology of mediation, descriptive geometry, required the designer to focus the majority of his or her attention on a flat, projected visualization that would also serve as the primary representation for measurement. For architecture these are horizontal plan cuts, vertical elevations parallel to the plane of the facade, and vertical sections perpendicular to the floor and wall surfaces. These drawing conventions have a discursive history in addition to their relations regarding the physical realities of gravity, but as orthographic projections they were also key drawings for the fact that they could be measured, translating visual information into mathematical information through geometric similarity. To repeat an earlier statement, the digital model no longer requires the plane of projection to be coincident with the plane of measurement. This is a significant conceptual and aesthetic change for architectural design. It does not negate the disciplinary discourse and history of drawing conventions, nor does it negate gravity and the organization of horizontal movement. But it does shift the pragmatic necessity that previously required the drawn view to also be a measurable projection. And in this pragmatic shift, there is an opening of alternate conceptual and aesthetic potentials for architecture through the digital model.
The orthographic projections and sections of traditional architectural representation will continue to communicate the information that they store; they will likewise provide a crucial link between changing historical developments of architectural thought. But they are no longer enough. They do not capture the qualitative differences in a design developed as a digital model. The traditional modes of mediation must be supplemented and extended by experiments in mediation that seek to transmit the conceptual and aesthetic difference of a computational model. These experiments can begin by exploring what is implied through the fusion of storage, manipulation, and transmission operating in a digital model. Architects now have a design mediation that is: one, spatial-vector based not planar-position based; two, fixed in scale and variable in resolution; and three, simulated as rendered real-time updates for visual evaluation.
A designer working through a digital model is manipulating spatial-vectorial information and judging these manipulations as real-time simulations at a specific resolution. The resolution is variable, both globally and locally. The updated simulations may be run forward or backward, but always remove the visual index of stages between transformations unless instructed to do otherwise. A single projection is incidental; a cut is momentary. The plan and section offer partial analytical representations. But the changing character of these surfaces resists a single moment of analysis. What mediation will be able to communicate the character and qualities of the aesthetics operative in these variations? Is architecture now free of the underdrawing in a manner similar to painting? And if this is so, what strange hybrid beast is released by computational mediation?
The computer is not just another tool. This oft-invoked analogy assumes that the tool has no effect on the conceptual or aesthetic agenda of the user. The conservative attack against the computer in architecture would thus seem to have some justification, as there is indeed a difference and something is lost. While this may be true, the question is not to hold onto what is lost, but to begin to take advantage of what is gained and to attempt to understand where threads of continuity with the past can be found. We would be wise to take a moment to consider how previous representational traditions interwove themselves into new developments of representational technology, and to consider the ways in which we may develop continuity with disciplinary traditions. Simultaneously, I would like to suggest that there are potential developments for future architectures that we will only be able to engage when we no longer require computational representations to mimic those produced by previous technologies.
1. Friedrich Kittler, Optical Media (Cambridge, UK: Polity Press, 2010), 26. ↩
2. Jeremy Gilbert-Rolfe, Beauty and the Contemporary Sublime (New York: Allworth Press, 1999), 22–23. ↩
3. Mario Carpo, “Drawing with Numbers: Geometry and Numeracy in Early Modern Architectural Design,” Journal of the Society of Architectural Historians 62, no. 4 (December 2003): 448–69. ↩
4. Gerald Farin, Curves and Surfaces for CAGD: A Practical Guide, 5th ed. (San Francisco: Morgan Kaufmann, 2002), 1–8. ↩
5. Kittler, Optical Media, 26. ↩