“architects do not make buildings, they make drawings for buildings.”     –Robin Evans¹

This essay is an attempt to discuss the role of drawing in architecture. By taking a look at how drawing and architecture have been closely connected, from the Renaissance to the Russian avant-garde and into the last three decades, some of the underlying forces and processes could come into focus. It is not only an attempt to discuss, but also to try and understand the relationship between drawing and architecture and what that means for present and future architecture. 


Already Alberti writes in his De re aedificatoria about the important role of lineaments, linear constructions projected “in the mind” in preference to the physical material reality of buildings. But it is a bit blunt to simply oppose those projected constructions to actual buildings. Robin Evans writes in The Projective Cast: “What connects thinking to imagination, imagination to drawing, drawing to building, and buildings to our eyes is projection in one guise or another, or processes that we model on projection”²

Projections are the way how the architect bridges the gap between the idea and the material. So let’s take a closer look at some of the projections.

The invention of perspective enabled painters and architects to recreate their surroundings in a natural but scientific way by geometrical means, finding a mathematical logic in everything. The Golden Section and its Divine Proportions illustrate this by putting Man, and thereby the creator of man, God, at the center of all creations. The fact that man is forced in an uncomfortable and unnatural pose to prove this mathematical logic was being ignored.(fig. 1) 


Now it became possible to represent space on a two-dimensional surface in a natural way. By the introduction of depth and thereby space in the drawing, also time was introduced in the drawing. The viewer could “travel” through the represented perspectival space which creates certain sequences and a possibility of a narrative. Architects could add a new tool to their palette. Stan Allen writes about this “In architecture, the smooth space of mathematical reason allowed the architect to reverse perspective’s temporal vector and project precisely imagined constructions into the future.”³  But there was unease about how perspective could be used in the realization of actual spaces. The use of perspective relies on the vanishing point to complete the composition. But when the viewer is not at the predetermined viewpoint, the composition could fall apart. That is why the perspectival projection was used more conceptually in Renaissance architectural theories of perception and proportion. “[…] design is not visualization (empirically “testing” successive versions) but rather the manipulation of a series of highly abstract devices–primarily the orthographic projections of plan and section–that serve to describe and construct the space.”⁴  These projections, orthogonal or perspectival, are no autonomous objects; they represent relationships between aspects of the drawing. These relationships can be interpreted and reconfigured in different ways and representations while “reading” the drawing. In analogy to the reading of drawings; can the production of drawings be called “writing”?

El Lissitzky explained “Perspective limits space; it has made it finite, closed.” Axonometry “has extended the apex of the finite visual cone of perspective into infinity […]”⁵ The perspectival projection on a plane represents a three-dimensional space, whereas the axonometric projection is “the ultimate illusion of irrational space” that makes it possible to navigate through and around the subject, “constructing new worlds”.

Although perspective is a mathematical invention and has scientific roots, it has mostly been used in a pictorial way, imagining the known world according to worldly principles. When the revolutionary avant-garde of the early twentieth-century wanted to construct new worlds they chose to use axonometric projection, it communicates abstract information, is measurable and precise. Axonometric projection has its origin in the military, was taught at engineering schools and became very important during the industrialization period. If perspective is related to the École des Beaux-arts than axonometry is related to the École Polytechnique. Axonometry is not trying to map vision; it deals with construction and measurability. These very aspects attracted the avant-garde.

Perspectival drawing eventually ends in the vanishing point, closing it up in space and time. Axonometric drawing has its vanishing point at infinity, which “suggested a continuous space in which elements are in constant motion.”
Attempts to realize axonometric works in a physical built form have had similar results as the anamorphous projections on church domes; the impact and energy of the drawings was lost. When visiting El Lissitzky’s Proun Space in the Van Abbemuseum in Eindhoven (fig. 2&3), it can feel quite disappointed. When experiencing the space physically, the potential and dynamic that it could generate from looking at the drawings wasn’t there. Or when looking at the model of Van Doesburg’s Café Aubette it also looked much more static and flat than the drawings suggest. So there is a difficulty in translating these avant-garde axonometric visions to a true physical experience. This difficulty resulted in a role in the margins of architecture for axonometric experimentation for approximately fifty years; until the so-called de-constructivist experiments of for example Daniel Libeskind (on paper) or Peter Eisenman (also in models, fig. 4) among others.

What was the backdrop of the rise of a new avant-garde who became the “starchitects” of today?

In the early 60’s innovations in arts and film inspired for “architectural speculations on paper”. The ’68 strikes fueled a “distrust of […] institutions and social conventions” and architecture came to distrust conventional type and program; and the conventional “collaborating architecture–would soon, and easily, be replaced by a more radical, democratic, free, uninhibited world.”⁶  The perspectival drawing method became synonymous for the collaborating corporate architecture and had to be opposed. The oil crisis of the 70’s slowed down the building production and talented promising architects took on teaching positions, “where the graphic experimentation […] condensed into a primary mode of research.”⁷ 


Architecture was preparing itself to take the lead in radical social and political reinvention. This condensation led to the objectification of the drawing as a means on its own, as a strategy to launch the architectural debate. The drawings were able to reveal a reality that goes beyond any built reality. These multiple interpretations and readings of the drawings show a richness that is outside the scope of the physicality of the building. 


Daniel Libeskind gives a very clear description of the state of architectural drawings at the time of his 1979 project Micromegas: The Architecture of Endspace:

“Architectural drawings have in modern times assumed the identity of signs; they have become the fixed and silent accomplices in the overwhelming endeavour of building and construction. In this way, their own open and unknowable horizon has been reduced to a level which proclaims the a priori coherence of technique. In considering them as mere technical adjuncts, collaborating in the execution of a series made up of self-evident steps, they have appeared as either self-effacing materials or as pure formulations cut off from every external reference.”⁸

The drawing becomes, once again, more than depicting the various views of the building, they construct a narrative and show not only the appearance of the project but also its meaning in a broader sense. A dialogue is opened by this narrative that questions the formal and spatial dimensions, subjective content and interpretations are taking into questioning the supposed objective realistic perspectival view.

Spatial and physical continuity, best represented in the perspectival view are undermined by cinematic and choreographic techniques like in Bernard Tschumi’s Manhattan Transcripts (fig. 5). Which is a mix of an anti-hero involved in the Russian avant-garde meets Georges Bataille and Jacques Derrida, has the subversiveness of Archigram and is on a Situationist dérive through Manhattan. All represented in photographs, conventional architectural drawings and motion diagrams. Or Rem Koolhaas’ AA thesis project Exodus (fig. 6), “A work set in three equal registers: a scattershot set of emaciated, anemic parodies of traditional architectural drawing; a storyboard of colorful collages; and a text/script.”⁹  Or like Peter Eisenman’s House I–XI series ( fig. 7), who used the analogy of language with its fixed set of words, to liberate design from the burden of functional and contextual requirements by proposing the transformation of a fixed set of elements like walls, columns and stairs into a “process-driven design”. So by explaining parts of the projects/process in the drawing instead of “depicting the resultant form” the “evolution of the concept” becomes the subject of the project. Patterns emerge that are as much part of the project as the final form.

 In these drawings all preconceived conventions about architectural drawings are being destroyed, and all that is left is a state of architectural anarchy. The drawings are almost autonomous works of art. The Micromegas works can be seen as the final step in axonometric drawing and fit perfectly in El Lissitzky’s description as “the ultimate illusion of irrational space”. It is as though a multitude of Prouns have collided that are in a constant flux. Or consider Libeskind’s Chamber Works drawings of which John Hejduk writes:
“This phenomenological work of Libeskind has turned the body inside out, but with a difference.
Before there was the possibility of transcendence, of the release of a soul that could be free in the unknown heavens. Now, the new drawings produce the antithesis. […] But here we also have the soul being discarded, carried by a landscape of inexplicable meaning. […] 
Libeskind silences all; we see the very soul.”¹⁰

So, did Libeskind’s drawings announce the end of/the final stage of the evolution of the architectural drawing? Is it like Malevich and his White on White (fig. 10) or Black Square paintings? Libeskind himself writes about his Chamber Works drawings:

“[…] This work in search of Architecture has discovered no permanent structure, no constant form and no universal type. I have realized that the result of this journey in search of the ‘essentials’ undermines in the end the very promise of their existence. […]”¹¹

Libeskind certainly maneuvered himself in a very difficult position. He not only silenced all, he almost silenced himself.  In Pandora and the modern scale model machine Albert C. Smith writes that by engaging in the construction of three scale model machines Libeskind wants to re-enter the realm of architecture. In his Three Lessons in Architecture (fig. 11,12&13) he builds a Reading machine to teach (himself?) the process of building. He then builds a Memory machine, which “consists of what can still be remembered in architecture.” And finally he builds the Writing machine that “teaches the artless and scienceless making of architecture”. These machines can be seen as an attempt to provide Libeskind a new framework and a set of reference standards to test his concepts.
It certainly proved impossible to compete with these drawings on a similar platform. That’s when the attention shifted to the diagram during the nineties. Which allowed for continuous, folded surfaces and smooth transitions, as opposed to conflict and instability. 
What is the position of architectural drawing today? The rise of the computer in the design studio and the development of Building Information Modeling (BIM) could change the role of the architectural drawing once again. In the communication between architect, client and contractor, 3D computer models are becoming more and more important to express the ideas of the architect. Not only in photo-realistic representations but the complete building can exist in virtual reality. The architectural drawing could be less and less designed to express the idea of the architect, but a result or revelation of complex processes of which the architect is the director. Does that lead to good architecture? Sometimes a bad script results in a good movie, sometimes a good script results in a bad movie and sometimes a good script results in a great movie. It all depends on the director.

1 Allen, Stan Practice: architecture, technique and representation p.1
2 Ibid p.183 (notes)
3 Ibid p.7
4 Ibid p.7
5 Ibid p.17
6 Kipnis, Jeffrey  Perfect Acts of Architecture p.10
7 Ibid p.11
8 Libeskind, Daniel Countersign p.14
9 Kipnis, Jeffrey  Perfect Acts of Architecture p.14
10 Libeskind, Daniel Countersign p.122
11 Ibid p.111


Figure 1. De architectura libri deci, source Smith, Albert C. Architectural Model As Machine
Figure 2. El Lissitzky Proun Space drawing, 1923, source: http://www.usc.edu/schools/annenberg/asc/projects/comm544/library/images/305.html
Figure 3. El Lissitzky Proun Space, Berlin 1928, source: http://www.tate.org.uk/research/tateresearch/tatepapers/07autumn/berndes.htm
Figure 4. Peter Eisenman, axonometric model House X, 1975-78, source http://blog.miragestudio7.com200708peter-eisenman
Figure 5. Bernard Tschumi, Manhattan Transcripts 1976-81, source Kipnis, Jeffrey Perfect Acts of Architecture   
Figure 6. Rem Koolhaas a.o., Exodus 1972, source Kipnis, Jeffrey Perfect Acts of Architecture 
Figure 7. Peter Eisenman, House VI 1976 source: http://www.kjabaird.blogspot.com/
Figure 8. Daniel Libeskind, Micromegas, Little Universe 1979 source Libeskind, DanielCountersign
Figure 9. Daniel Libeskind, Chamber Works 1983 source Kipnis, Jeffrey Perfect Acts of Architecture
Figure 10. Kazimir Malevich, Suprematist Composition White on White 1917 source: http://en.wikipedia.org/wiki/Image:Marevich,_Suprematist_Composition-_White_on_White_1917.jpg
Figure 11. Daniel Libeskind, Reading Machine 1985 source Smith, Albert C. Architectural Model As Machine
Figure 12. Daniel Libeskind, Memory Machine 1985 source Smith, Albert C. Architectural Model As Machine
Figure 13. Daniel Libeskind, Writing Machine 1985 source Smith, Albert C. Architectural Model As Machine


Allen, Stan (2000) Practice: architecture, technique and representation. Routledge
Forty, Adrian (2000) Words and Buildings. A Vocabulary of Modern Architecture Thames & Hudson
Kipnis, Jeffrey (2001) Perfect Acts of Architecture Thames & Hudson
Libeskind, Daniel (1991) Countersign Academy Editions
Smith, Albert C. (2004) Architectural Model As Machine Architectural Press