Did you notice that the cover of Log 50 is white?

For the past nine issues, since

Log 41, the pink, a screening of the warm red that graced the cover from num-
bers 31 through 40, has been slowly fading. Fading to white and an unknown
future. Or fading to white and a return to the beginning. (Long-time readers
of Log may see the game.)
Editor Why did we do this? Well, first, what color can possibly follow the wel-
Cynthia Davidson coming jolt of warm red? Second, at the time of Log 41, pink was very, very
“fashionable.” And metaphorically, facing an always unknowable future, fad-
ing gave us an out. A possible fading away rather than an abrupt departure.
Managing Editor
So what happened? A lot happened. In the realm of architecture journals,
Patrick Templeton San Rocco announced it would cease publication, PRAXIS published its final
issue, and AA Files went through some temporary contortions. These are
Editorial Interns miniscule things in light of the dramatic shifts in global politics and economic
Harish Krishnamoorthy sanctions (not to mention the COVID pandemic), but they are registers of dis-
Anna Renken course in architecture. So Log kept fading the color of its cover and expanding
its pages. The last three numbers were the largest we’d ever published, until
Protagonists now. Log 50 is 256 pages, and the number of authors in this issue could barely
Thomas Daniell be squeezed onto the contents page.
Throughout history cultures have ascribed importance to particular num-
Todd Gannon
bers – lucky seven (lucky eight in China) or unlucky 13, for example, or the
Catherine Ingraham devil in triple sixes. Fifty is considered a milestone in a marriage, a golden
Sanford Kwinter anniversary, even though 50 is the atomic number of tin. It feels like a mile-
Manuel Orazi stone here, too. This isn’t just Log 50, it’s the marker of the journal’s 17 years
Bryony Roberts in print, nearly a generation of writing that is beginning to form a record of
Julie Rose architectural theory and production in the first two decades of the 21st century.
Sarah Whiting If we weren’t all still working remotely due to COVID, Patrick in
Brooklyn, Anna in St. Paul, Harish in Houston (and Bangalore), and myself
Postcard and Cover Story in Connecticut, we would invite you to a launch party for Log 50 and raise a
A model lifts a Styrofoam log glass to you. The toasts would be endless: to dedicated supporters like the late
Harry Cobb and Charles Jencks; to new contributors like Erin and Ian Besler
to promote the Dow Chemical
(great Log 49 postcard!) and Erin Manning (great Zoom event Erin!); to the
Company, Midland, Michigan, author who wins the prize for greatest number of footnotes, Sylvia Lavin
1949. Courtesy Science History (Pier Vittorio Aureli may be in second place); to those who write 8,000 words
Institute. and to those who write 250; to the eight protagonists who keep us honest (see
names at left); to every intern and editor who has spent time with us, and to
our benefactors and readers.
Architecture is an act of love. A love of life, a love for each other, and a
love of learning. Nothing has made that more clear than the global pandemic
that separates us all, leading journalists to declare the decline of cities. But
architecture is at work. And Log is at work. We will see you again, in a bright
new color, when we put out Log 51. – CD

Log 50 Copyright © 2020 Anyone Corporation. All Rights Reserved.

ISSN: 1547-4690. ISBN: 978-0-9992373-8-0. Printed in USA. Log is published
three times a year by Anyone Corporation, a nonprofit corporation in the State of New
York with editorial and business offices at 41 West 25th Street, 11th floor, New York, NY
10010. Subscription for 3 issues: $45 US; $49 CAN/MEX; $69 International. Distributed
by Idea Books and Motto. Single issues are $18 plus shipping. The opinions expressed
herein are not necessarily those of the protagonists or of the board of the Anyone
www.anycorp.com Corporation. Send inquiries, letters, and submissions to [email protected].
Log 50
Fall 2020 Model Behavior

Stan Allen 16 Thinking in Models

Sean Anderson 227 Compound Tenses
Phil Bernstein 238 Canonical Models of Architecture
Cynthia Davidson 9 Notes on a Concept: Model Behavior
Joe Day 173 One to One
Penelope Dean 249 Business, Actually
Dora Epstein Jones 115 Models In and Out
David Erdman 139 Mottle Behavior
David Eskenazi 179 Tired . . . and Behaving Poorly
Marshall Ford 135 Blue Foam
Forensic Architecture 217 Operative Models
Todd Gannon 125 Mind the Gaps! Toward a Pedagogy of Models
and a Model Pedagogy
Erik Herrmann 39 Role Play
Eric Höweler  49 Verify in Field: Models and Other Useful Fictions
Christian Hubert 151 Model Behavior?
Ferda Kolatan 71 The Chunk Model
Jimenez Lai 131 Ephemeral Models’ Permanent Ghosts
Mark Lee 55 Models and Models of Models
& Sharon Johnston
Alice Loumeau 206 The World Model
John McMorrough 145 Not a Role, Model
Michael Meredith 62 A Conversation about Models
Kiel Moe 157 Architectural Agnotology & Broken World Models
Rizal Muslimin 203 Computational Puppets
Jason Rhys Parry 208 Simulating an Assassination: Four Explosion Models
Shane Reiner-Roth 191 Internet Browser
Jesse Reiser 103 Voodoo
Paulette Singley 75 Dollhouses and Other Bad Objects
Tyler Survant 195 Modeling and Remodeling the Oval Office
Patrick Templeton 185 A Model Painter: Behind the Scenes with Amy Bennett
Neyran Turan 163 Unnatural Models
Tom Wiscombe 91 The Inner Life of Models
& Marrikka Trotter
Andrew Witt 28 Shadowplays: Models, Drawings, Cognitions
Kechao Xiang 256 Log of Curiosity
Observations on Models: Samantha Ding, 190... Log Team, 184... Anna Renken, 150...
Alican Taylan, 172... Davide Tommaso Ferrando, 90...
Cynthia Davidson
Notes on a Concept:
Model Behavior
In the summer of 2019 – what now seems like a prior life-
time – we at Log began to talk about the idea of model behav-
ior. While this concept might first be thought to mean “good
behavior,” for us the idea recognized the rapid changes – even
before COVID – occurring in architecture and its production
of form. Working with Elisa Iturbe on Log 47: Overcoming
Carbon Form, it was obvious that the carbon-based models of
habitation that architecture has essentially produced since the
Industrial Revolution need to be changed because the behav-
iors they sustain are, in the face of climate change, no longer
morally tenable. And working with Bryony Roberts to frame
Log 48: Expanding Modes of Practice, it was clear that a new
An example of a late 19th-century generation of practitioners, one that both looks different and
string model of the type developed thinks differently from a profession historically “modeled” by
by Théodore Olivier, now in the
archives of the mathematics Institut
white males, is already introducing new modes, or models, of
Henri Poincaré, Paris. From Andrew architectural thought and action.
Witt, Formulations: Architecture, At the same time, we were working with Andrew Witt on
Mathematics, Culture, forthcoming.
his forthcoming Writing Architecture series book, Formulations:
Courtesy Andrew Witt.
Architecture, Mathematics, Culture. In chapter two he discusses
the development of physical visualizations of the principles of
ruled surfaces and descriptive geometry in late 19th-century
France. The resulting mathematical forms – string models and
wireframe models – were soon in demand in engineering and
architecture schools, creating what Witt calls a “globalized
model market [that] signaled a newly vital visual culture of
mathematics.” Add to this the creeping technological changes
over the past 30 years that now constitute the wholesale
embrace of computation and digital modeling – the visual but
nonphysical models of Rhino or BIM, for example – and model
behavior, or what models do, what they influence or prove,
raised a question. In architecture, what is model behavior?
Neither the model nor its behavior is particular to archi-
tecture. Economics, climate science, politics, art… the list
goes on and on, so prevalent is the concept and use of models
today: from statistician Nate Silver (now billed as a “forecast-
ing expert”) talking to the insurance industry on “how mod-
els might change the future,” to First Lady Melania Trump’s
“Be Best” program, which, the New York Times reported, is

9
On April 8, 2020, Rachel Maddow, in her
eponymous MSNBC nightly news show,
discussed responses to the day’s mod-
els of COVID infection rates alongside a
graphic that said “MODEL/BEHAVIOR.”
The forward slash both separated and
linked these words, seemingly verify-
ing our interest in questioning model
behavior in architecture.

“intended to encourage children to model good behavior.”

Data-driven models both reflect and predict behaviors,
whether of markets, ecologies, voters, or buyers. And each of
these affects architecture, which produces physical and digital
models to work out spatial concepts and project future visions
of reality that also reflect and predict behaviors.

Models in Question
All around us today, cultural, scientific, and historical models
are being called into question. When climate change mod-
els spawn new forms of science denial, when a US president
edits a NOAA hurricane model with a Sharpie to reflect his
personal forecast, is the kind of imagination and expertise
that underlies architectural models still seen as projecting
believable futures? Are there simply too many models – “So
many damn models telling us so many damn things about our
now certain damnation,” as Green New Deal activist Billy
Fleming writes – or too few convincing models? According to
new media researcher Wendy Hui Kyong Chun, “To combat
global climate change we need to act on the basis of models
[that] cannot be entirely verifiable.” This is the very point
that climate change skeptics use to argue that those models are
wrong. The architect Marco Vanucci, writing at opensystems-
a.com, offers a disciplinary perspective:
Architects mediate the complexity of the world and their ideas
through different instrumental modalities. Whether perspective
drawings, proportional relationships, descriptive geometry, material
prototypes, scaled models, maquettes or three-dimensional models
– models serve the purpose of collecting and indexing information
into measurable and rational systems so that the architectural proj-
ect can be nurtured, refined, measured, and its behaviour simu-
lated, tested and possibly predicted before getting built.
The model works, at the same time, as an instrument of mea-
surement, representation and forecast. It measures the extensive

10 Log 50
 characteristics of matter, it represents the built form via other
media, and it predicts its intensive traits, its behaviour. The model
operates in the Deleuzian domain of the virtual: it is non-actual,
nonetheless it is real.
In other words, the model is not the architecture itself, but its
effects, its behaviors, are real.

Definitions
Merriam-Webster’s Dictionary lists 14 definitions simply for
the noun form of model, the first being “a usually miniature
representation of something,” and the fourteenth, “obsolete: a
set of plans for a building.” Each definition implies an action
or code of conduct – that is, a behavior. For example, Henry
Ford named his first automobile, in 1903, Model A, as in the
second definition, “a type or design of product.” Celebrity
Padma Lakshmi exemplifies definition six, “one who is
employed to display clothes or other merchandise,” and in
Historian Hillary S. Kativa writes in her second career as a host of television’s Top Chef, she may
“It’s a Mad, Mad World: Dow and the be seen as a role model. Model stems from the Latin modulus,
Age of Consumption,” Distillations (sci-
encehistory.org) that this image in Dow
meaning measure or standard. The concept of a standard also
Chemical Company’s postwar adver- relates to conduct, that is, good behavior. Merriam-Webster’s
tising campaigns laid “the groundwork defines behavior as “the way in which someone conducts
for a lifetime of consumption in the
hopes that the child who played with a
oneself or behaves.” The second and third definitions are of
toy [model] house today would buy and greater interest for the concept of model behavior: “anything
furnish a real one tomorrow.” Both that an organism does involving action and response to
miniature and full-scale model homes
stimulation” and “the way in which something functions or
are designed to expand the behavior of
consumption. Image courtesy Science operates.” In the latter sense, the many forms of architec-
History Institute. tural models can be said to display a range of behaviors, from
structural to material, aesthetic, and social. These, in turn,
elicit the “response to stimulation” by those who engage a
given model.

Idea as Model
In 1976 and again in 1980, two exhibitions called “Idea as
Model” were mounted at the Institute for Architecture and
Urban Studies in New York. In the catalogue Idea as Model,
published after the second exhibition (Rizzoli International,
1981), architectural historian Richard Pommer writes, “Why
not an exhibition of drawings instead of models? Drawings
are closer to ideas – sketches to initial conceptions, eleva-
tions to proportional schemata, plans to geometric diagrams,
views to dreams. Models by contrast are usually dumb, for
show and the study of appearances. Perhaps the hope was to
present difficult ideas more effectively to the public, but that
would have been no more than another form of ‘propaganda

11 Log 50
Sou Fujimoto Architects, Architecture is Everywhere (foreground), Chicago Architecture Biennial, 2015. Fujimoto’s “ana-
gogical” models play with found objects, including an ashtray, potato chips, and ping-pong balls. They demonstrate his
belief that architecture is a creative activity and are likely to convince a viewing public that architecture can, at the very
least, be found in everything. In the background is MOS’s House No. 11 (Corridor House), a small house that assembles
the circulation spaces found in luxury housing developments. While the architects claim that details such as mechanical
systems were worked out, making Corridor House a possible dwelling, the 1:1 model in the gallery was more a conceptual
one than a model “house for sale.” Photos (top; bottom middle, right): Tom Harris. Courtesy Chicago Architecture Biennial.
Photo (bottom left) © Diego Hernández.

12 Log 50
for persuading clients.’” Arguably the growth of interna-
tional architecture biennales and architectural curating since
that time has proven Pommer wrong. If the marketplace is a
measure, the general public responds more to architectural
models than to drawings. In fact, not long after the 1976 exhi-
bition, a small market for architectural drawings and mod-
els arose, giving these contemporary objects a new role more
like that of art. Recalling “Idea as Model” in OASE 84: Models
(2011), architectural historian and curator Stefaan Vervoort
refers to the work of artists, including Rita McBride and
Thomas Schütte, rather than architects to demonstrate that
the model “provides a dialogue” between ideation and mate-
rialization, in essence denying architecture such a concep-
tual role. But London-based architect and theorist Teresa
Stoppani, in a chapter titled “Model: from object to process”
in her book Unorthodox Ways to Think the City (Routledge,
2019), calls the architectural model a “transitory object,”
essentially occupying another in-between condition. “The
shift from the in-formational (the potentiality to produce a
form) to the orientational reveals the transitory presence of
the model,” she writes. “The model in-forms the construc-
tion of the building, and once the building is built the model
becomes redundant. But does it? . . . Liberated from the
one-to-one relation with the object that it has enabled or
informed, the model can then reengage in relations that play
on scale and repetition. Still charged with the potentiality of
producing otherness, the model can be never-finished and
remains active also when it appears complete.” This poten-
tial for otherness suggests that the architectural model can
“provide a dialogue” beyond informing the construction of
a building.

Modeling Architectural Concepts

The focus of The Architectural Model, by historian Matthew
Mindrup, is summed up in its subtitle, Histories of the
Miniature and the Prototype, the Exemplar and the Muse (MIT
Press, 2019). In the last chapter, “Modeling Architectural
Concepts,” Mindrup defines the conceptual model as one
that demonstrates “an idea about architecture or a guide for
user’s decisions about the design of an architectural form, its
organization of spaces, or its structure.” But his three types
are more abstract: the allegorical, which weaves figurative
meanings with architecture (for example, Peter Eisenman’s
“cardboard architecture,” Daniel Libeskind’s Memory
Machine, Ben Nicholson’s Appliance House); the analogical,

13 Log 50
 which involves the “transfer of comparable qualities,” say
from everyday objects to a model or building (Le Corbusier’s
Unité d’Habitation, Steven Holl’s Stretto House, Santiago
Calatrava’s Kuwait Pavilion), and the anagogical, which
Mindrup defines as “the process by which architects tempo-
rarily suspend their identification of a found object in order
to interpret it as a model of architecture . . . [or] in philosoph-
ical terms as a phenomenological reduction of bracketing.”
As examples, he offers the models of an unnamed Hermann
Finsterlin project, Peter Zumthor’s Thermal Bath House and
Bruder Klaus Field Chapel, and Kurt Schwitters’s Castle and
Cathedral with Courtyard Well. The behaviors of these concep-
tual model types would seem, from Mindrup’s telling, to be
minimal due to their likely limited audiences.
The idea of interpreting a found object as architecture
(minus the phenomenological bracketing) reminds me of Sou
Fujimoto’s installation of nearly 40 tiny models in the first
Chicago Biennial in 2015. Although placed between two full-
scale models of houses by MOS and Tatiana Bilbao that any-
one would identify as architecture, Fujimoto’s “anagogical”
models were found everyday objects staged to suggest archi-
tectural forms, which in turn suggested that architecture can
be found in everything and influences everything. A careful
arrangement of processed Pringles potato chips suggests a
landscape; an assemblage of ping-pong balls suggests a theater
or an observatory; an ashtray turned upside down becomes
an architectural form simply by adding a few scale figures
around it. Fujimoto showed that an architectural concept can
spark the public imagination.
A floor-to-ceiling display of New York Perhaps, in hindsight, “The Architectural Imagination”
architect Alex Gorlin’s collection of exhibition, which I cocurated with Mónica Ponce de León
architectural miniatures. Writing
about models and miniatures in Gran
in 2016, was actually a moment of reflection on the archi-
Bazaar (October–November 1985), tectural model. The 12 architects we commissioned to design
architect Piera Scuri said, “The fun- speculative projects for four sites in Detroit were required
damental difference . . . between an
to produce a model that would fit on a four-by-eight-foot
architectural model and a miniature
lies in the relationship with reality. table. This resulted in scenarios that ranged from the scale-
While the model, representing reality, less to the siteless, from experimental geometric forms to
brings us closer to it, the miniature
field conditions, from polished presentation models to models
replaces reality with a dimension
cleaned of the frightening features of that exhibited a lifestyle to an interactive model one could
reality.” Is this distancing from reality “enter” by wearing a virtual reality headset. Each model in
why so many architects (and architec- “The Architectural Imagination” not only projected ideas
ture critics) collect miniatures? Photo:
Eric Petschek. Courtesy Alexander
but also initiated discussions, elicited reactions, and, as then
Gorlin Architects. Thanks to Teresa City Planning Director Maurice Cox told us, “changed the
Stoppani for the Scuri reference. conversation in Detroit.”

14 Log 50
T+E+A+M, Detroit Reassembly Plant,
commissioned for “The Architectural
Imagination,” 2016. The architects’
ideas about recycling demolition debris
to produce a new, aggregate building
material were evident in the model.
Their addition of tiny furniture – a
frequent move by many young offices
today – injected the model with known
settlement behaviors.

Model Behavior
When we announced Model Behavior as the theme of Log 50,
to guarantee some measure of response to the subject we
invited some 40 individuals to address the topic, hoping half
of them would accept. Ultimately, we received more than 165
submissions, a record number for an issue of Log and a clear
indication of deep interest in models. But as we read through
them all, it was obvious that how a model might elicit or
project behaviors was not always a primary concern. Rather,
the making of models, whether they have digital or physi-
cal properties, and the unmasking of the largely invisible
transactional models that underpin the systems of architec-
tural education and practice (models that could be said to
standardize behaviors) came to the fore. Model airplane kits
and dollhouses and artists’ models entered the conversation,
perhaps leaning more toward a model’s object quality than its
behaviors (though behaviors do ensue, even if unpredictable
in this context). Even role models surfaced, though not in
architecture (nor in politics). A lot of history washed ashore,
and was sent back out with the tide. What appears here, in
these pages, are thoughtful assessments of the state of the
model in architecture in its many forms and behaviors. But I
am convinced that model behavior, the effects and influences
of models, is an ongoing issue for architecture, especially in
a time that demands real change in how we live together.
While acknowledging the diverse and complex roles of mod-
els today, Log 50 reminds us that we cannot lose the unique
spatiality that architectural models bring to the table. Only by
understanding and acknowledging the behaviors those spatial
qualities project can architecture fully participate in changing
the world. For its participation begins with models.
Cynthia Davidson is a “model citizen”:
she always votes.

15
Angelique Firmalino, rendering from Master of Architecture thesis advised by
Stan Allen, Princeton University School of Architecture, Spring 2020. Drawing
courtesy the author.
Stan Allen
Thinking in Models

We were using a lot the color coding for drawings, line drawings by
hand or models, I remember tons of models.
– José Oubrerie, recalling working methods in
Le Corbusier’s atelier, 1957–65

Models run the world

– Promotional copy for Modelshop, a business
intelligence software platform, 2020

For a brief period in the early 1990s, before the paperless stu-
dios at Columbia’s GSAPP, and before the protocols of digital
design work had been fully established, I owned a copy of a
software program called ModelShop. I had little desire to learn
AutoCAD, which was already widely used for drawing pro-
duction in professional offices. Programs such as ModelShop
and Form-Z, which both appeared in the late 1980s, prom-
1.  ModelShop was released in 1989. It was ised something else: they offered new design tools.1 Mac-based
positioned as a “concept modeler” but
ultimately lost market share due to its
(and therefore aimed at designers rather than professional
inability to integrate 2D CAD input. See offices or engineers), ModelShop was intuitive and easy to use.
“ModelShop,” Macintosh Garden, https://
macintoshgarden.org/apps/modelshop. As the name suggests, it transposed, quite literally, the opera-
Thanks to Greg Lynn, who tracked down tions of physical model making in chipboard or basswood to a
this material in response to an e-mail
query. The exhibitions and catalogues of computational space. The user started with a palette of sheets
the “Archaeology of the Digital” series or square members that could be cut to size, extruded, and
Lynn curated in collaboration with the
Canadian Centre for Architecture remain assembled. The desktop icon was a diminutive X-Acto knife
the best sources for information on the blade. Google “modelshop” today and you will land on the
early use of the computer in architecture.
2. “The architectural model is perhaps one site of a “No-code AI Modeling Platform” for business intel-
of the oldest tools for architectural design, ligence software.
but it is also one of the least preserved
and studied.” Massimo Scolari, “The Idea I thought about this – both the restructuring of design
of Model,” in Oblique Drawing: A History
of Anti-Perspective, trans. Jenny Condie
protocols brought about by computer modeling and the
Palandri (Cambridge: MIT Press, 2012), shadow of big data – when the studios and fabrication facili-
137–84. This remains the case. Forty years
on, the Institute for Architecture and Urban
ties at Princeton’s School of Architecture were abruptly shut
Studies’ 1981 publication Idea as Model is still down by COVID-19 last spring. I was advising a thesis stu-
a primary source, and Models, a 2011 issue of
OASE, is the best contemporary collection. dent, Angelique Firmalino, who had proposed working exclu-
See Idea as Model, ed. Kenneth Frampton sively through physical models as an open-ended experiment
and Silvia Kolbowski (New York: Rizzoli
International, 1981); and OASE 84: Models to highlight the role of medium and means of representation
(2011). Thanks to Angelique Firmalino for in design work. Drawing, she noted, has been extensively
this bibliographic research.
theorized, whereas the use of models is discussed infrequently
and often in disconnected and inconsistent terms.2 She set

17
 out to test a hypothesis: that working exclusively through
physical models might lead to an architectural proposition
more immediate and tangible than could be achieved through
drawing alone. After a promising start, and in particular an
emerging interest in the low relief – a design syntax derived
from the flatness of photographic representation and realized
through operations of 3D milling – students were shut out
of studios and shops, confined at home, and obliged to work
Screenshot of ModelShop, Macintosh- exclusively in the virtual realm of the computer. The disci-
based 3D modeling software released pline was thrust into a massive real-time experiment on the
by Paracomp in 1989. Image courtesy
Macintosh Garden.
effects of working in virtual design space.
For Firmalino this should have been a disaster; in fact, it
was the opposite. The forced reboot opened up a rich territory
for design experiment and produced a new visual convention
derived from the space of computer modeling and photog-
raphy. It called attention to the fact that, even before the
lockdown, there was a complex interplay between computer
modeling and physical modeling. Fabricated models are built
in the computer before they are output as tangible objects.
This begins to suggest to me that the differences between vir-
tual models and physical models may not be as consequential
as often thought. As her advisor, it provoked me to reconsider
the implications of a comprehensive shift from drawing to
computer modeling in design work today.
I tend to be suspicious of the language of the “paradigm
shift.” New tools and new techniques do not always replace
older ways of working and thinking but are instead layered
onto the existing, creating hybrid working procedures.
When new strategies become available, the toolbox expands
rather than contracts. And to speak about a paradigm shift
is to assume a level of consensus that simply does not exist
across the diverse spectrum of architects working today.
Nevertheless, among a younger generation of architects who
entered the field in the 2000s, it is increasingly clear that there
is a distinctive way of working and thinking that depends in
large part on computer modeling as its primary design proce-
dure. For these architects, there is no separate category called
digital design. All design today takes place in digital space,
including many practices that reject the precepts once associ-
ated with computation: curvilinearity, formal complexity,
the nonstandard, mass customization, etc. As digital design
operations have become entirely absorbed into the everyday
workflow of all architects, a realignment has taken place that
is conceptual as much as it is technical. And for a number of
younger architects, these new design procedures have both

18 Log 50
 enabled and reinforced a preferred aesthetics of immediacy,
abstraction, and telegraphic legibility: an emergent sensibility
3.  My notion of a shift toward a simplified, more than a paradigm shift.3
telegraphic reading of the building’s
profile is similar to but somewhat distinct
The deep historical and conceptual affinity between
from R.E. Somol’s concept of shape. I am drawing and design is well-documented. It is registered in
sympathetic to his arguments, but would
say that, while he is more concerned with Alberti’s theory of lineaments and elaborated through a long
affect and graphics, I am more interested history of descriptive geometry and projection. It is implicit
in design techniques used to realize those
affects; moreover, the available tools and in the abstract geometric scaffolds of Rudolf Wittkower
the culture of modeling have advanced and Colin Rowe and their elaboration by Peter Eisenman. It
considerably since 2004, when his piece
was published. See R.E. Somol, “12 Reasons undergirds John Hejduk’s translation of painterly space into
to Get Back into Shape,” in Content, ed. Rem architecture, Daniel Libeskind’s preoccupation with the trace,
Koolhaas and the Office for Metropolitan
Architecture (Cologne: Taschen, 2004). as well as theories of the index that persisted through the
4.  See, among many others, Hal Foster, 1990s and into the first generation of digital design work. And
ed., Vision and Visuality (Seattle: Bay Press,
1988). for my generation – close readers of Robin Evans – it was
5.  Note, for example, the exhibition of almost impossible to imagine ambitious design work without
Beaux-Arts drawings at the Museum of
Modern Art, “The Architecture of the École a close attention to drawing and projection.
des Beaux-Arts,” on view from October
1975 to January 1976, or Architectural Design
In the ’80s and ’90s, discussions of representation also
47: America Now: Drawing Towards a provided an important point of contact between architecture
More Modern Architecture (1977), guest
edited by Robert A.M. Stern. and advanced critical theory in art history and the visual arts.4
6.  An architecture curator I contacted “Representation” is at once a category of technical operations
estimated that less than five percent of
archival holdings are models. More often
specific to architecture as a discipline (questions of drawing,
photographs of models are collected. projection, and descriptive geometry as described by Evans
See Véronique Patteeuw, “Miniature
Temptations: A Conversation with CCA and others) and, at the same time, a conceptual template to
Curator Howard Schubert on Collecting understand the production of meaning in architecture. If
and Exhibiting Architectural Models,”
OASE 84: Models (2011): 123–27. architecture is understood primarily as a system of represen-
tation – a medium of communication among other media – it
follows that it is best unpacked using the tool kit of semiotics
and theories of language. The long shadow of the semiotic in
the architectural theory of the late 20th century was bound up
with drawing and theories of representation; it is no accident
that postmodernism in architecture coincided with a renewed
interest in drawing.5
If drawing acquired legitimacy, in part, through the
reflected prestige of visual arts theories, models suffered not
only from a lack of discursive support but also from their
suspect affiliation with less-than-serious pursuits like toys,
model trains and airplanes, or dollhouses. Add to that their
fragility and tendency to disintegrate over time. Architects’
drawings are archived, conserved, and treated as objects of
historical study. Models are collected and archived too, but
in far fewer numbers.6 Without easy access to study models,
historians typically trace the evolution of a project primarily
through sketches and design drawings. José Oubrerie, for
example, has noted that models were regularly used in Le
Corbusier’s studio but rarely saved. SANAA, a contemporary

19 Log 50
 office known for its extensive use of models, purges the accu-
mulated study models in its shop several times a year. As a
7. An exception to this is OMA, an office result, the model has been written out of history.7
where models are fundamental to the design
process. See Albena Yaneva, Made by the Office
Despite claims to the contrary, for those who respect the
for Metropolitan Architecture: An Ethnography history of the discipline, drawing and the conceptual habits
of Design (Rotterdam: 010, 2009).
8.  Robin Evans, “Translations from
associated with drawing persist when designing on the com-
Drawing to Building,” in Translations puter. Drawing’s conventional orthographic logic is embedded
from Drawing to Building and Other Essays
(London: Architectural Association, 1997), in contemporary drafting and modeling software: CAD pro-
153–93. See also Evans, The Projective grams simulate and automate the categories and operations of
Cast: Architecture and Its Three Geometries
(Cambridge: MIT Press, 1995). analog drafting. But behind all of these new drawings is a 3D
9.  Thomas Demand, “A Conversation with model, and it is in the realm of the model that design unfolds
Anne Holtrop and Bas Princen,” OASE 84:
Models (2011): 55–56. today. In a drawing culture, an imagined form is described
10.  “In a more general sense, a model is a first in two dimensions. Each drawing is partial; as the object
structure, a pattern, along the line of which
something is shaped. . . . This is all the becomes more complex, more drawings are required. The
more so when the complexity of something architectural imagination is located in this space of translation
increases or the scientific sphere becomes
so minute that any kind of observation between the measurable two-dimensional space of the page (or
would fail. . . . Generally, a model is a the screen) and three-dimensional form, built or projected. As
theoretical complexity in itself which
either brings a visual form or a conceptual Evans has described, in both historical and operative detail, it
order into the components of complex
situations.” “Models,” in Oswald Mathias
is projection that enables this translation, a process that archi-
Ungers, Morphologie: City Metaphors, 2nd ed. tects have internalized over time as a conceptual template.8
(Cologne: Walther König, 1982; 2011), 11.
11.  Discussion of the differing instrumen-
By contrast, design takes place today primarily through
talities of drawing and models also tends the slow buildup of an abstract three-dimensional construct
to get confused by a lack of clarity around
what a drawing is, as well as a lack of
situated in a boundless yet measurable space: a computer
attention to the different instrumentalities model that is constantly rotated and viewed from different
of drawing types. Drawing practice was
systematized very early on. The Vitruvian angles, but also in more or less proximate viewpoints – now
categories of ichnographia, orthographia, and close in, now far away, at times from inside. The process of
scenographia were revived in Renaissance
architectural treatises, and, although computer modeling includes moments of orthographic sta-
the terms are rarely used, the categories bilization that refer to plans or elevations, but in a computer
persist. See Werner Oechslin, “Geometry
and the Line: The Vitruvian ‘Science’ of model, these are simply pauses in a more extended, fluid pro-
Architectural Drawing,” Daidalos 1 (1981): cess. Even as it smooths and internalizes the workflow, the
21–34. Moreover, not all two-dimensional
graphic instruments are drawings; digital computer model collapses the space of projection, becoming
renderings, for example, are images of
computer models filtered through rendering
more direct and immediate. Operations of transcoding and
and image software. See John May, transposition replace translation, with its inevitable linguis-
“Everything is Already an Image,” Log 40
(Spring/Summer 2017): 9–26. tic overtones. Scale becomes a property of output; all design
is now 1:1. With analog drafting, the number of drawings is
always finite and each one incomplete, a partial approxima-
tion of a future building or object. A single digital model, by
contrast, can generate an infinite series of two-dimensional
drawings. Moreover, the model is indifferent to its mode of
representation. It can be viewed and output in perspective or
parallel projection, on paper or on-screen, as a still drawing, a
time-based video, or a 3D print. The capacity to generate mul-
tiple copies of itself is another defining property of the model
understood as a tangible diagram.

20 Log 50
 Tangible Diagrams: Model as Object
“Model” is a really multilayered term: I feel most comfortable
using it in its most abstract sense: we are thinking in models. . . .
Scientific models are needed to eliminate transient and unintended
side effects in order to see the objective more clearly, and more
importantly, to be able to repeat the experiment that proved the
theory. Models accompany us from our earliest experience in this
world: toys are mostly abstract versions of things around us.
– Thomas Demand9

Models are tangible objects, real and immediate, with specific

material properties, and at the same time they stand in for
something else – they are miniature ciphers of buildings or
ideas for buildings. Like buildings, they are concrete, physical
things; like drawings or diagrams, they are abstract devices
that designate possible future realities. It is precisely this
in-between status, this insistent toggling between real and
abstract, that makes the model such a powerful design tool.
Thomas Demand, Farm 58, 2015. Although no longer exclusively made by hand, models still
Framed pigment print. © Thomas retain the qualities of tactility and presence associated with
Demand / Artists’ Rights Society
(ARS), New York / VG Bild-Kunst,
making. For this reason, models are often associated with an
Bonn, 2020. Photo courtesy Matthew improvisational design method that foregrounds materiality
Marks Gallery. and bricolage. But the materiality of the model is not the
materiality of the building. Buildings are complex material
assemblages, whereas models can be (and often are) made
from one material. Scalar reduction erases detail and renders
the model intrinsically abstract and diagrammatic. Its com-
pactness is the compactness of an idea or phenomenon made
immediately visible in a simplified form. This is the instru-
mentality of the model in scientific culture, where complex
systems are “modeled,” typically as data-driven computer
simulations.10 The model is, in this sense, related to the dia-
gram but distinct. A scientific model is a dynamic construct.
Architectural models are insistently object-like, but they
retain the conceptual power of the scientific model, that is to
say, the ability to figure and fix future forms and patterns.
Drawings and models signify differently. Drawings are
embedded in specific disciplinary histories and are interpreted
and read according to known conventions.11 Drawings imply
projection and translation, while models are immediate and
present – they resist interpretation. Buildings and models
are both things in the world that communicate without the
intermediary of a codified system of representation. There
is no translation at work here: neither spatial (as described
by Evans) nor linguistic. The process of signification is

21 Log 50
 telegraphic, like a diagram. The model’s closest kin among
drawing types is the axonometric, which also describes a
12. This point is underscored by the fact measurable three-dimensional object floating in space.12
that Scolari devotes an entire chapter of his
history of oblique projection to the model;
Today, the category “model” needs to encompass GIFs,
for Scolari, the objectivity, scalar reduction, 3D computer models, and complex data sets. To rethink the
and abstraction of the model align it with
the antiperspectival operations he describes: model in this way – as a tangible diagram – is to see physical
“At the beginning of the eighteenth century, models, 3D computer models, and scientific models not as
the model was considered the equivalent of
parallel projection drawing.” Scolari, 158. separate categories but as points on a spectrum. There are
13.  Demand, 55–56. consequential differences, of course, and I do not want to
14.  Jesús Vassallo, “Digital Collage and Dirty
Realism in Architecture,” in Seamless: Digital minimize the importance of the model’s physical presence.
Collage and Dirty Realism in Contemporary The immediacy of contact, the fluidity of perception that
Architecture (Zurich: Park Books, 2016), 117.
Also in Log 39 (Winter 2017): 48. comes with bodily interaction, and the tactile awareness that
15.  John Hejduk, Mask of Medusa, ed. Kim comes of making something by hand remain fundamental to
Shkapich (New York: Rizzoli, 1989), 69.
16.  “A model embodies ideas quite literally. the agency of the model. But I also want to suggest that this
It forces you to resolve things in three same immediacy, which is such a vital property of a physical
dimensions to some degree. But it is
simultaneously still ‘open’: you can see model, is reflected back onto a virtual model. In the imagina-
other things, discover other possibilities tive space of the designer, the distance between physical and
in it.” Adam Caruso, “On Models and
Images,” interview by Job Floris and Hans virtual is narrowed, if not bridged, through the use of mod-
Teerds, OASE 84: Models (2011): 129. els. Physical and virtual models alike are “abstract versions of
17. A generation younger than the protago-
nists of the Düsseldorf school (Andreas things around us,” and they reinforce the idea that creative
Gursky, Thomas Ruff, Candida Höfer, work involves “thinking in models.”13
Thomas Struth, etc.), these photographers
are equally indebted to the work of Bernd
and Hilla Becher, as well as to the American
photographers associated with the New
Photography and Graphics: Model as Image
Topographics, in particular Lewis Baltz. I The introduction of the digital blurs the division between observa-
owe many of the points made here to ongoing
discussions with Jesús Vassallo, who has
tion and action, between representing the world and proposing new
written insightfully on contemporary archi- worlds. Once we assimilate the ideathat a photograph can be manipu-
tecture’s dialogue with photographic practice.
See his Seamless and Epics in the Everyday: lated or constructed, it no longer matters whether it actually has been
Photography, Architecture, and the Problem of or to what degree. When photography’s illusion of neutrality dissi-
Realism (Zurich: Park Books, 2019).
pates, the act of making a picture becomes more than ever a charged
statement about how the world could or should be.
– Jesús Vassallo14

Models have a specific relationship to the body. Their object-

like character means that they can be picked up, held, and
manipulated. We are, as Hejduk observed, much larger than
the model: “We hover about it and walk around it, we hold it
in our hands, etc.”15 This is the power of the model as a tool in
the design studio; its compactness induces clarity, while min-
iaturization triggers a process of free association.16 The mod-
el’s lack of detail leaves space for imagination. At the same
time, the concreteness of the model endows the imagined
object with a sense of reality: the way, for example, that light
strikes a surface or the parallax effects of a moving viewpoint
can be simulated and apprehended faster and more effectively
with a physical model. Models are quick and effective design

22 Log 50
OFFICE Kersten Geers David Van ciphers. But it is also true that, in large part, models circulate
Severen, Centres for Traditional in the world beyond the studio through the filter of photogra-
Music, Riffa, Muharraq, Bahrain, 2018.
Drawing courtesy the architects.
phy. And today, a photograph of a physical model, the graphic
output of a computer model, and a photograph of a building
are all processed through the same graphic software.
Photographs, like models, traffic in the tension between
the real and the abstract. A photograph has a documentary
value as a transcription of the real; it records a given reality
in all of its gritty and often-overlooked detail. Yet the photo-
graph’s image of reality is flattened, cropped, reduced in scale,
and rendered in false color (or converted to black and white).
Today it is often digitally manipulated. Photographic works
by Bas Princen, Josef Schulz, or Philipp Schaerer present con-
cise and direct images of anonymous architectural construc-
tions – warehouses, office buildings, or industrial constructions
– which occupy featureless landscapes that suggest the terrain
vague of the postindustrial city.17 The objects photographed are
abstracted and reduced to iconic, repeatable forms: models that
will appear and reappear in these context-free landscapes.
Parallel to this has been the emergence of a graphic lan-
guage that also borrows from the immediacy of the model.
OFFICE Kersten Geers David Van Severen, for example,
maintains an active collaboration with Princen. In their char-
acteristic perspectives (now widely emulated), flattened
abstract forms, minimally detailed and materially consistent,
occupy featureless grounds. Colors are assertively false, in

23 Log 50
 service of a graphic immediacy that means that these draw-
ings are not read and translated but signify directly. A con-
stellation of ideas and techniques around the concept of the
model (physical or virtual), its photographic reproduction,
and this emergent graphic sensibility begin to define the out-
lines of an aesthetic project that favors immediacy, objectivity,
legibility, and serial repetition. For these architects and pho-
tographers, the terrain vague of the postindustrial city and
these iconic architectural objects belong not to an individual
architect or author but to the discipline as a whole and to a
larger collective consciousness. Architecture, in this sense, is
always both more and less than a culture of images, and in a
society saturated with images, that tension is fundamental to
architecture’s political task today.

A Model Aesthetic
Thus a model repeatedly becomes the model for something else, a
following incarnation, serving as a kind of plan. For it is capable,
just like a plan, of generating a series of other products – other
plans, other objects, other models.
18.  Kersten Geers, “The Model as a Plan: A  – Kersten Geers18
Monument to Scientific Error,” OASE 84:
Models (2011): 64.
19.  The most obvious example is Michael Models also correlate to the logics of display. Shown in gal-
Meredith’s “44 Low-Resolution Houses”
exhibition at the Princeton University
leries or museums, the architectural model acquires the status
School of Architecture in 2018, where all of an autonomous sculptural artifact. A number of recent
the projects were exhibited in consistently
scaled white paper models. Others include exhibitions have presented works by a loosely bound group
“Inscriptions: Architecture Before Speech” of younger architects who share an affinity for the imme-
at the Harvard GSD in 2018; the exhibition
and book Possible Mediums (Barcelona: diacy and “presentness” of the model.19 It is not only that
Actar, 2018); the final issue of PRAXIS, Bad models proliferate across these exhibitions; the foreground-
Architectures (November 2019), the 2017
Chicago Architecture Biennial “Make New ing of the model as a display strategy marks a conceptual
History”; or, more distantly, the architects shift in design thinking. Among these architects (The LADG,
of the Ordos 100. See Stan Allen, “Pattern
Recognition: Scanning Inscriptions,” in WOJR, Collective–LOK, Bureau Spectacular, Studio Sean
Inscriptions: Architecture Before Speech,
ed. K. Michael Hays and Andrew Holder
Canty, MOS, JaJa Co, First Office, MALL, and Outpost Office,
(Cambridge: Harvard GSD, forthcoming). among others), there is a clear preference for mass and sur-
20.  I am indebted to James Wood, who
recently pointed this out in a seminar.
face (properties of the model) over plane and line: the ele-
ments of drawing. Detail is suppressed and the forms tend to
be clear and simple. Preferred shapes are singular and often
figural; evidence of assembly is minimized. Exhibition strat-
egies – particularly the staging and photography of models
– tend toward the scenographic, with strong lighting, simpli-
fied outlines, and a frontal presentation that recalls theater
sets.20 The spatial logics of design procedures such as Boolean
operations associated with computer modeling render this
work more plastic than geometric: the computer allows
the designer to work with form as if it were already matter

24 Log 50
 – mass that can be molded or carved, in contrast to operations
of projection or the logic of the trace. Following on that, I
would suggest that in this work, history, or the vernacular, is
modeled more than quoted, which distances it from postmod-
21.  See Robert E. Somol, “Grind Houses,” ern reference, which operated through a linguistic filter.21
in 2G 67: Johnston Marklee, (2013): 12.
Somol contrasts the use of projective
Nearly all models produced today involve computer fab-
geometry by Preston Scott Cohen in rication. At one end of the scale are robots and at the other,
the Wu House to the use of modeling in
Johnston Marklee’s Vault House: “Cohen 3D printers. These procedures, in turn, map onto traditional
invents a geometric problem in order to sculptural categories: milling, like carving, is subtractive,
expose architecture while Johnston
Marklee are primarily invested in the cul- laser cutting involves the assemblage of fabricated elements,
tural possibilities provoked by architecture.” and 3D printing mimics sculptural modeling. This suggests
Emphasis original.
22.  According to the Rhino website, that among the visual arts, sculpture has displaced painting
Arctic mode uses “ambient occlusion,” a as a primary point of reference for architects today. If there is
rendering technique that calculates the
exposure of each point on a surface to an emblematic object for the present moment, it is the printed
an ambient light source. “Arctic display model. In a printed model, the material is unified, usually
mode,” Rhinoceros 6, https://docs.mcneel.
com/rhino/6/help/en-us/options/ white, and detail is minimized, rendering it highly abstract.
view_displaymodes_arctic.htm.
The 3D print is as close as possible to a direct transcription of
the shaded design model visualized on-screen. The abstrac-
tion of the model has recently been cycled back into the soft-
ware itself: Rhino’s “Arctic” display mode, introduced with
version 6, “turns all objects white, turns the background
white, and adds soft shadows.”22 The aesthetics of the white
cube gallery space, disconnected from place and context, are
reproduced in the working space of the computer.
But it would also be a mistake to think that this is sim-
ply a recent phenomenon or one dictated exclusively by the
software and limited to the gallery. Herzog & de Meuron’s
concrete house in Leymen (completed in 1997) is a mono-
lithic construction: walls, floors, and decks are all concrete.
Windows are punched voids, and the mullions are set back to
emphasize the thickness of the concrete walls. The roof is of
necessity waterproofed, but the material is close in color to
the concrete, unifying the form. Transitions are minimized:
there are no eaves, overhangs, or reveals that would suggest
an assembly of parts. As a plastic material that can be molded
and cast, concrete tends toward this kind of unified, sculp-
tural whole. The house hovers above the ground, like an
object that could be plucked up and moved at any time. The
material of the house is assertively real – the concrete is var-
iegated and weighty, yet its materiality is undercut by the
abstraction of the form and the precision of its realization.
Like a model, it oscillates between the thing itself and what
it suggests beyond itself. The concrete house also functions
as a model in the sense that it models – that is to say, captures
in condensed and compact form – the archetypal idea of the

25 Log 50
Aires Mateus, House in Leiria, Leiria, Portugal, 2010. Photo © Fernando Guerra.
Courtesy the architects. Top: Herzog & de Meuron, House in Leymen, Leymen,
Haut-Rhin, France, 1997. Photo: Margherita Spiluttini. Architekturzentrum Wien
collection. Courtesy the architects.

26 Log 50
 house. Like Laugier’s primitive hut, the house is a prototype
for countless iterations of this selfsame form, past and pres-
ent, locating it outside of chronological, successive time.
A final example, closer to the present. The work of the
Portuguese partnership Aires Mateus is insistently stripped-
down, white, and abstract. Their 2010 house in Leiria, like
the Herzog & de Meuron concrete house, repeats the arche-
typal pitched-roof form of the house. It is even more insis-
tently monolithic than the concrete house, with no visible
detail or material transitions whatsoever, and a seamless
white roof. As photographed, it could almost be mistaken for
a 3D-printed model. This is particularly true of the void cut
into the roof that connects all the way down to the courtyard
below grade. The apparent simplicity of the exterior conceals
a spatial complexity that is revealed through experience – an
effect of modeling and not a three-dimensional translation
23.  The model undercuts the primacy of the of drawings.23
plan; going further, it neutralizes an entire
vocabulary of formal analysis that depends
Now it might be noted that simple white plaster forms
on the conventions and categories of in strong sunlight belong to the vernacular of Portuguese
orthographic drawings: frontality, rotation,
transparency, flatness, and depth. building, and this is, in part, what makes the architecture of
Aires Mateus so compelling. Part of the program is located
below grade, and the sectional interplay grounds the house
in the reality of the site. The abstraction of the model is
here linked not only to the reality of place but also to spatial
and topological richness, to local building traditions, and to
archetypal forms. It would be easy to multiply examples: Sou
Fujimoto’s stacked Tokyo Apartments, Johnston Marklee’s
Vault House, and Cecilia Puga’s Casa en Bahía Azul are all
built works that derive from a design logic based on the com-
bination and recombination of simple model-like elements,
both solid and void. All of these, and many others today, are
paradigmatic examples of the power of thinking in models –
not the literal use of any specific technique but a more gen-
eralized conceptual shift that opens new possibilities at once
radically unexpected and, at the same time, deeply embedded
in collective memory.

Stan Allen is an architect and George

Dutton ‘27 Professor of Architecture at
Princeton. His most recent publication
is Situated Objects, a book of drawings
and photographs documenting his
architectural work from 2012 to 2019.

27
A matrix of architectural drawings generated from a neural network trained on thousands of Beaux-Arts drawings, includ-
ing plans, sections, and elevations. In many cases those drawing types are hybridized or morphed into a new and ambigu-
ous arrangement through the encoding of the neural network. Project team: Andrew Witt, Gia Jung, Claire Djang, 2020.
Image courtesy the author.
Andrew Witt
Shadowplays:
Models, Drawings,
Cognitions
Model is generalization; form is special case.
– Buckminster Fuller, Synergetics 2, 1979

In its classical guise, an architectural model is both a tangible

artifact and a proxy for the imagined reality that it is designed
to resemble. With its volume, shadow, and tactility, a model
stakes a claim for an architectural idea as embodied fact. A
model looks like a building and vice versa: they share a com-
mon and specific form, conjoined as a dyad of original and
copy. They exist together in a zone between physical and ideal,
fact and fiction. A model is an intermediary between appear-
ance and imagination, anchored in the form of a specific object.
If an architectural model is a designed artifact, a 21st-
century scientific model is more mathematical or logical than
physical and spatial. As philosopher Ian Hacking notes, “A
model in physics is something you hold in your head rather
1.  Ian Hacking, Representing and than your hands.”1 Climatic, economic, and ecological models
Intervening: Introductory Topics in the
Philosophy of Natural Science (Cambridge:
are all austere mathematical systems instead of thick objects.
Cambridge University Press, 1983), 216. Visual resemblance is irrelevant to the behavior of systems
that scientific models aim to encode and make operative. If the
architectural model relies on resemblance, the scientific model
rests on the numeric language of deep and hidden structures.
When the two poles of architectural and technoscientific
models are juxtaposed, a spectrum of practices opens between
them. Architectural models evoke specific imaginations
through tangible material objects, while logical scientific
models posit the mathematical interactions of abstract entities
and phenomena. Yet separating the disparate tactics of archi-
tectural and scientific models is ever more confounding due
to the proliferation of digital instruments that surreptitiously
import the modus operandi of the exact sciences into the
practice of design. Abstract varieties of quasi-scientific digital
models are increasingly supplanting physical models, whose
function rests merely on appearance. The once tautological
connection between model, resemblance, and representation

29
 in architecture is giving way to new relationships between
epistemic abstraction and technique. What is emerging is an
increasingly scientific intuition – if not an explicit under-
standing – of model as a creative and evaluative matrix that
exceeds the scaled specification of a single building.
By attending to models that are not precisely architectural
but are on a continuum between architectural and techno-
scientific, the roles, possibilities, and futures of architectural
modeling can be critically reframed. In particular, the dichot-
omization of visual resemblance and instrumental abstrac-
tion that dominates discussion of models can be critiqued and
perhaps overcome. If the building is dislodged as the exclusive
focus of model representation, more disciplinary functions of
the model emerge, such as its capacities for cultural encapsula-
tion, propagation, and diffusion. Here I consider two types of
models – skiagraphic models of 19th-century shadow rendering
and image-based neural network models of 21st-century arti-
ficial intelligence – that abandon any pretense of resemblance
to buildings in favor of more abstract roles within the knowl-
edge culture of design. Each type of model brings visual and
mathematical rigor – geometric rigor for the skiagraphic model
and statistical rigor for the neural model – to bear on systems
of perception and representation. Their function is not merely
instrumental: both play critical roles in encoding visual practices
beyond the direct specification of buildings. In both episodes,
models are not scaled miniatures but tools for training architec-
tural perception and creation by both humans and machines.
In their philosophical account of scientific models, bio-
mathematicians Philip Gerlee and Torbjörn Lundh remind
us that the word model descends from the Latin “modulus,
a diminutive form of modus, meaning a small measuring
2.  Philip Gerlee and Torbjörn Lundh, device.”2 A model, then, is a ruler against which to gauge, to
Scientific Models: Red Atoms, White Lies
and Black Boxes in a Yellow Book (Berlin:
delimit, and to judge. Models are devices to dimension not
Springer, 2016), 1. only buildings but the culture of architecture – its practices,
3.  Bas C. van Fraassen, Scientific
Representation: Paradoxes of Perspective
conventions, styles, and processes. Yet the appearance of
(Oxford: Oxford University Press, 2008), 13. buildings is never a distant concern of the architect, and even
the relentless abstractions of scientific models can be hacked
for freshly intense and unexpected kinds of design invention.
New computational forms of neural vision open up strange
kinds of glitched, warped, and liquid transformations. In
this way, visuality mediated by calculation models confirms
philosopher of science Bas C. van Fraassen’s observation that
“distortion, infidelity, lack of resemblance in some respect,
may in general be crucial to the success of a representation.”3
These new models are engines to mutate representation itself.

30 Log 50
Jean-Jacques Lequeu, skiagraphic Modeling Shadows
construction showing the classical Architectural models are specific artifacts, but they are also
technique derived from descriptive
geometry. From Lequeu, “Architecture
evidence of disciplinary vision and concretized conventions
Civile,” 1777–1825. Drawing courtesy that invite the user to behold the idea of a building in a par-
Bibliothèque Nationale de France. ticular way. The status of architectural models in the larger
Right: Jean-Jacques Lequeu, plans,
pantheon of representations is brought into relief through
elevations, and sections of two domed
projects, assembled as a single their sometimes peculiar and even competitive relationship
drawing and rendered with precise with architectural drawings. Models rarely exist alone as the
shadows, 1775. Drawing courtesy
only representations of buildings. Instead, they are one in an
Bibliothèque Nationale de France.
entourage of other representations – drawings chief among
them – that collectively delineate the world of a project.
Modeling and drawing are conjoined practices and nowhere is
that more apparent than in the atmospheric realm of shadows.
With its illusion of solidity, the shaded drawing seems to
exceed the merely documentary qualities of technical plans
and adopt the appearance of a three-dimensional model. This
conflation between drawing and model is most discernible in
the artfully rendered drawings of the 19th-century French
Beaux-Arts architects. Buildings were drawn as if they were
models, shadows cast as if the sectional thickness were cut
away. Among the most facile hands was Jean-Jacques Lequeu,
whose remarkable drawings seem to close the gap between
drawing and model. The shadows cast in Lequeu’s interiors
evoke his contemporary Jean-Baptiste Rondelet’s monumental

31 Log 50
 sectional maquette of the Panthéon in Paris as much as they do
actual buildings. Much scholarship on the enigmatic Lequeu’s
work rightly focuses on his flamboyant imagination or meticu-
lous craft. Yet shadows held a definite priority in Lequeu’s
technique. His “Architecture Civile,” an unpublished drawing
manual in which he claimed to outline “the rules of the science
of natural shadows,” is substantially devoted to the tonal and
4.  “Les règles de la Science des Ombres geometric intricacies of rendering shade.4
naturelles,” quoted in Philippe Duboy,
Lequeu: An Architectural Enigma
In Lequeu’s drawings, we see a virtuosic manifestation
(Cambridge: MIT Press, 1987), 14. of skiagraphy, the projective science of rendering shadow. As
5.  See William Muschenheim, “Curricula
in Schools of Architecture: A Directory,”
a disciplinary practice, skiagraphy altered the common pre-
Journal of Architectural Education 18, no. 4 cedence between drawings and models. To produce a skia-
(March 1964): 56.
6.  Thomas DaCosta Kaufmann, “The
graphic drawing, the meticulous draftsperson projectively
Perspective of Shadows: The History of constructs the shadows of a complex model – a building, a
the Theory of Shadow Projection,” Journal
of the Warburg and Courtauld Institutes 38 fragment, or an entirely contrived object – with exquisite
(1975): 258. precision. In its pure form, skiagraphy was an academic exer-
cise to mold a visceral intuition of light. Between the early
19th and mid-20th centuries, skiagraphic drawings were de
rigueur in architecture schools across Europe and the United
States. The practice persisted well into the 20th century and
was taught at reputable schools of architecture, such as The
Bartlett, into the 1960s.5 As much as almost any architectural
drawing practice, skiagraphy defined a family resemblance
among architectural drawings of a certain style and from a
certain period.
What is the object of representation in a skiagraphic
drawing? The apparent focus of depiction would seem to be
the model itself. But in fact, skiagraphic models are more like
props, merely incidental to the object’s shadows, which are
the proper focus. The true test of the draftsperson’s skill is
not the rendering of the model per se but rather the render-
ing of the epiphenomenal shadows. The secondary effects of
the shadows are elevated to the primary object of attention.
Drafting these shadows entails a sophisticated and systemic
analysis of the atmospheric conditions that surround the
model as well as the occluded geometry of the model itself. In
other words, skiagraphy requires a concept of a world system
that the rendered model inhabits. Unmoored from represen-
tational obligations, the model can adopt other functions and
characteristics that ignore the conventions of building and call
attention to the systems of representation itself.
In his account of shadow projection, art historian
Thomas DaCosta Kaufmann calls practices like skiagraphy
“modeling shadows.”6 Kaufmann notes that specially con-
structed models played an essential role in the training of

32 Log 50
 Renaissance painters’ visual intuition for the construction of
light and shade: “Vasari says that his contemporaries contin-
ued to use ‘rounded’ models of clay or wax before drawing
their cartoons, in order to see shadows in sunlight; he says
that Michelangelo had used models, and that the sculptor
Jacopo Sansovino had supplied wax models for a number of
7.  Ibid., 260. painters.”7 The models Michelangelo and other painters used
8.  See Michael Hirst and Carmen Bambach
Cappel, “A Note on the Word Modello,” The
were physical props or maquettes that furnished the space of a
Art Bulletin 74, no. 1 (March 1992): 172–73. painting.8 These models were not intended as representations,
9.  See Gaspard Monge and Barnabé Brisson,
Géométrie descriptive: augmentée d’une théorie
except in the highly indirect fashion that they simulated the
des ombres et de la perspective extraite des shadows of specific figures. It was the resemblance of shad-
papiers de l’auteur (Paris: Gauthiers-Villars,
1922). This publication followed Monge’s
ows that mattered, not the resemblance of model to object.
1798 edition published by Baudouin in Paris. Models were expedients to initiate and train practitioners, to
10.  “L’essence même du dessin est
purement mathématique, car les deux seuls
introduce a disciplinary way of seeing.
modes sous lesquels il peut être envisagé, In 19th-century France, skiagraphy spanned architec-
le géométral ou le perspectif, aussi bien
en ce qui s’applique au trait qu’en ce qui tural and engineering practices and thus inevitably impinged
regarde le tracé des ombres, reposent sur on technoscientific culture. In Paris, then the European
des lois exactes, des vérités mathéma-
tiques. Cette manière de le considérer est nucleus in the professionalization of both architecture and
justifiée par la langue, puisque l’artiste et engineering, skiagraphy was an indispensable part of both
le mathématicien, chacun dans sa sphère,
emploient les mêmes mots, ligne, plan, architectural training at the École des Beaux-Arts and engi-
proportion, symétrie, équilibre, etc., en neering training at the École Polytechnique. As the birth-
leur conservant une signification identique;
elle se résume dans l’expression fort usitée place of Gaspard Monge’s descriptive geometry, France was
autrefois, la science du dessin; en soi le
dessin est donc une science.” Hervé Loilier,
a fertile soil for a quasi-scientific model of skiagraphy to take
“L’enseignement du dessin et des arts à root. Monge developed a theory of shadows that was an inte-
l’École polytechnique,” Bulletin de la SABIX
52: À la rencontre des peintres polytech- gral part of later editions of his Géométrie descriptive and led
niciens (2013): 13. Author’s translation. the way for a considerable literature of manuals, including
Lequeu’s, for training aspiring draftspeople in “les dessins des
ombres.”9 French sculptor Eugène Guillaume, who straddled
the porous boundary between the arts and engineering as both
director of the École des Beaux-Arts and professor of drawing
at the École Polytechnique, framed the pedagogical philosophy
of skiagraphic studies as a precise mathematical exercise:
The very essence of drawing is purely mathematical, since the only
two modes by which it can be envisaged, the geometrical or the per-
spectival, both that which would be applied to draw lines and to
trace shadows, rest on exact laws: the truths of mathematics. This
manner of consideration is justified by language that the artist and
the mathematician each employ in their own sphere, using the same
words of line, plan, proportion, symmetry, equilibrium, and retain-
ing the same meaning . . . drawing itself is a science.10
Drawing shadows was not an act of pure intuition and per-
ception but rather a deliberate practice of exact construction.
In a skiagraphic drawing, the raison d’être of the mod-
els was to provide an entry into a representational system and
a suitable challenge of skill for the eye, mind, and hand of

33 Log 50
the composing draftsperson. Models functioned as pretexts
for a specific mode of disciplinary training. Through them,
the architect was viscerally and autonomically sensitized to
the behaviors, moods, and subtleties of shadow. Physically,
skiagraphic models could be abstract platonic forms, odd
and awkward assemblages, mechanical contrivances, math-
ematical maquettes, or fragments of actual architectural
models. In the Beaux-Arts context, Corinthian capitals or
fragmented entablatures arranged in still-life tableaux were
favorite examples. Other skiagraphic models were improb-
able piles of odd forms, generative devices designed to induce
as much variety and difference in their shadows as possible.
Skiagraphic models were furniture in a regime of a highly
mathematical representation. Not all skiagraphic models were
even physical maquettes. They could be, and often were, more
abstract entities fancifully imagined and projectively con-
structed entirely within the drawing itself. Whether physical
objects or mathematical entities, the models’ highly contrived
forms were intended to probe the limits of representation,
beholden not to the demands of building but rather to the pri-
vate and autonomous conventions of architectural drawing.
The skiagraphic model was a parafactual entity that
existed purely as the linchpin of a specific practice of archi-
tectural seeing and drawing. The model had a definite physi-
cal presence and form, yet that form was secondary to its
role as an object of initiation. It never served to clarify or
expedite the design of a building. On the contrary, it defined
the architect, her intuition, and her vision. In that sense, it
was an entirely cultural artifact, a resolutely architectural
model that nevertheless had nothing to do with building,
only with seeing and drawing.

Beaux-Arts Deepfakes
In the 1950s, a quite distinct kind of model began to emerge in
the work of mathematical psychologists interested in visual
perception. These were not models of individual things (like
buildings) or even models of explicit systems (like skiagraphic
geometry) but models of perception itself. In his remarkable
1950 paper “Mathematical Biophysics, Cybernetics and
Significs,” mathematical psychologist Anatol Rapoport was
among the first to use the term model to describe the func-
tion of neural networks, the reticulated structures posited as
the basis of organic nervous systems. He recounts the conver-
gence between biological perception and electronic calcula-
tion: “The two programs of research, a mathematical theory

34 Log 50
 of the nervous system on the one hand and the development
of electronic computers on the other, proceeded along parallel
lines. . . . Workers from both fields soon found themselves talk-
ing to each other in a language which was a curious mixture
of psychophysiology (neurons, synapses, refractory periods,
threshold, etc.) and electronics (feedbacks, vacuum tubes,
11.  Anatol Rapoport, “Mathematical amplifiers, transformers, etc.).”11 When encoded electronically,
Biophysics, Cybernetics and Significs,”
Synthese 8, no. 3/5 (1950/1951): 189.
these new neural models took on the comportment of human
12. See Frank Rosenblatt, “The Perceptron: A or animal reflexes, reactions, and cognitions. They could be
Probabilistic Model for Information Storage
and Organization in the Brain,” Psychological
“trained” and generate new internal associations in response
Review 65, no. 6 (1958): 386–408. to serialized stimuli. In short, they could learn to sense.
13.  Ibid., 386.
One of the earliest computational neural networks was
an optical mechanism for the discretization of light and
shadow. In 1958, psychologist Frank Rosenblatt introduced
the perceptron, the mathematical framework for an intercon-
nected matrix of retinal sensors configured to detect gradi-
ents of illuminance.12 Rosenblatt attacked the assumption of
eidetic resemblance between model and modeled object in
neural representation, claiming that “the images of stimuli
may never really be recorded at all . . . the central nervous
system simply acts as an intricate switching network, where
retention takes the form of new connections, or pathways,
between centers of activity.”13 Rosenblatt’s argument is suf-
fused with multiple models: coded models, physiological
models, conceptual models. As befitted an essentially electri-
cal apparatus, the neurons of Rosenblatt’s network could be
readily and continuously tuned to inculcate particular kinds
of visual training.
In the 70 years since Rapoport’s account of neurocom-
putation, models that learn have made fitful but dramatic
advancements toward a distinctly novel fusion of perception,
computation, and creation. New tactics of deep learning like
generative adversarial networks have relentlessly pushed the
limits of computational perception and affiliated techniques
of image generation. Generative networks are models that are
not explicitly and deliberatively theorized but instead emerge
autonomically from iteratively applied statistical encodings.
Taking enormous archives of images as feedstock for train-
ing, neural models build matrices of statistical probabilities
that encode perceptual and creative behavior. On the one
hand, the process of encoding a neural network is akin to the
use of compression and encryption algorithms, distilling the
vast array of images to numeric correlations. On the other
hand, neural training is like teaching a child to recognize and
draw shapes through the reinforced repetition of countless

35 Log 50
examples of ascending complexity. Image by image, a distinct
intuition is formed through cybernetic feedback.
If classical architectural models were explicit physical
representations, neural net models are black-boxed codices
of relational connections. Though they are computationally
deterministic, neural networks are not rule sets per se. Instead
of semantically articulated rules, neural models are vast led-
gers of numeric correlations. Derived from probabilistic and
statistical associations as opposed to explicit logical rules, when
visualized in their raw form, these models appear almost as
noise even to the educated eye.
Yet the imagery produced by suitably trained models is
marvelously specific and inescapably legible. When trained on
the Beaux-Arts drawings of Lequeu and thousands of others,
a neural model begins to draw in the luxurious style of figured
volumes, filigreed details, and crepuscular shadows as con-
vincingly as any suitably trained draftsperson. What emerges
in these drawings is a statistical skiagraphy: shadow rendered
not with the constructive principles of descriptive geometry
but with the stochastic processes of neurocomputation. The
unmistakable forms of domes, colonnades, entablatures, and
other telltale elements of the Beaux-Arts vocabulary emerge
in a mannered chiaroscuro calculated from tensors of prob-
abilities. The images have impressionistic and atmospheric
qualities, as if seeing the precise lines of the building through
a light fog. There is a striking vagueness about these images;
they are more sketches than renderings. Yet that belies their
calculational precision, and what may glancingly appear as
fluid washes of watercolors are actually grayscales calibrated
by an exact science of machine learning.
Neural networks not only capture and reproduce the
specific atmospheric subtleties that elude more procedural
means of computation, they also have the capacity to blur
the line between distinct forms of architectural representa-
tion. If neural models are trained on several different genres
of drawing, they fluidly hybridize all these disparate types in
their generated images. Mutations of plan, section, and per-
spective are vivisected into strange new quasi-montages that
seamlessly blend them all together. Plan melts into section,
coalescing or dissipating through the technical intermedi-
ary of the neural model. Like skiagraphic models, the draw-
ings do not refer to specific buildings and, indeed, often depict
something that defies consistent interpretation. Instead, they
are artifacts of pure disciplinary representation. In this they
resemble another Beaux-Arts staple, the composite drawing

36 Log 50
A matrix of sections generated from a juxtaposing plan, section, and elevation in one patchwork
neural network trained on thousands of image. The neural model becomes a mixing chamber to refor-
Beaux-Arts drawings. Though derived
entirely from unsupervised training, a
mat all the myriad forms of modeling and drawing in a com-
characteristic technique of shadows mon and continuous visual language.
begins to emerge. Project team: In circumscribing and relating a corpus of images, neu-
Andrew Witt, Gia Jung, Claire Djang,
ral networks delimit specific territories of imagination. More
2020. Image courtesy the author.
than models, neural networks are maps. They statistically
interpolate disparate images and thereby plot a gradient of
interstitial architectures. Instead of modeling one form or a
discrete set of forms, they offer a model of visual invention
itself, and with it, a continuous and seamlessly variable terrain
of endless and endlessly different forms. Like the Situationist
dérive, latent walks through the neural space allow the user to
wander through the dream space of a trained intuition, each
step generating a new and surprising result. This continuity
is of a totally different order than parametric or combinato-
rial variation, where incremental changes of scale or den-
sity retain the essential topological organization of a space.

37 Log 50
 Instead, neural variation ranges across type and topology,
setting up liquid interpolations between improbable and
chimerical forms.

Modelers, Human and Machine

Despite their mathematical origins, neural models transcend
the strictly deductive and largely instrumental methods of
most contemporary scientific modeling. As models of thought
and perception, they expose the qualitative vagaries of taste
and style to exact modeling. In doing so, neural models con-
found the direct relationship between visual resemblance
and representation. Resemblance is merely a secondary effect
of a model that can learn and manipulate entire representa-
tional systems. In this paradigm, the architectural model of the
future will take on increasingly protean roles, representing not
only specific buildings but also the disciplinary intuitions and
cultural nuances of architecture itself.
At one time, a finite and specific architectural model
might have been interpreted as a representational terminus,
a definitive conclusion of a search, or the birth of an archi-
tectural idea in an embodied reality. Even today, most digital
3D models merely replicate or amplify the qualities of eidetic
resemblance characteristic of physical maquettes. In shifting
from embodiments of specific forms to generalized encodings
of representational processes, models become ever more expan-
sive and open-ended vessels of design culture. Now even imag-
inations, aspirations, and obsessions are amenable to modeling.
Once proxies for buildings that were manipulated by archi-
tects, models are on the verge of integrating the operative tastes
and judgments of the architect herself.
Inanimate models and human authors have always main-
tained unconfused and distinct places in architecture. As arti-
ficial intelligences model, incubate, and encapsulate cognition,
that careful distinction between made and maker, thought
and thinker may seem as antiquated as physical maquettes
themselves. Between the maquette and the architect there is
a new actor and mediator, the quasi-intelligent model that
embeds human intuitions and hallucinates endlessly elastic
images, drawings, and buildings. In the realm of imagination,
the gap between the neural-generated deepfake and the
human-imagined model is eroding. When models can down-
Andrew Witt is associate professor load and contain patterns of thought, they cease to be distinct
in practice at the Harvard Graduate
objects and simply become the way architecture is created.
School of Design. He has just enough
experience with chiaroscuro and
neural networks to be dangerous.

38
Erik Herrmann
Role Play

Today, one question predominates: that of information machines in

their twin aspect as “thinking machines” (machines à penser)
and “machines that make us think” (machines à faire penser).
– Abraham Moles1

In the early 1960s, German philosopher Max Bense and French

engineer Abraham Moles cofounded the field of informa-
tion aesthetics, a short-lived but vibrant postwar philosophi-
cal movement combining elements of information theory and
cybernetics, applying scientific methods and tools to artistic
production. The movement is most often associated with its
computer-produced art – drawings of randomly determined
patterns that Bense claimed represented his aesthetic theories.
Motifs from these early works continue to influence contem-
porary digital aesthetic genres, such as glitch art, but this leg-
acy only partly reflects the principles of information aesthetics.
A reexamination of Moles’s research into cybernetic systems
invites a broader interpretation of the movement: one that
expands the cultural definitions of digital culture to include
not only visual aesthetics but also new modes of exchange,
participation, and cooperation based on logic inferred from an
1.  Abraham Moles, “Cybernetics and the information society.2
Work of Art,” in A Little-Known Story about
a Movement, a Magazine and the Computer’s
These new cultural definitions are most succinctly
Arrival in Art: New Tendencies and Bit described in Moles’s 1965 essay “Cybernetics and the Work
International, 1961–1973, ed. Margit Rosen
(Cambridge: MIT Press, 2011), 217. of Art,” where he suggests that the creative capacity of the
2.  For more on Moles’s research and computer is not located in the disembodied machinations
the field of information aesthetics, see
Armin Medosch, New Tendencies: Art at of the computer’s processor (as in Bense’s research) but
the Threshold of the Information Revolution in broader creative systems and processes. Moles’s analysis
(1961–1978) (Cambridge: MIT Press, 2016).
includes a series of cybernetic organizational charts illustrat-
ing several models of creativity. Cybernetic analysis centers
on regulatory systems of control and communication with
an emphasis on causation and feedback, making it uniquely
suited to explore the relational dynamics of creative praxes.
Each model of creativity includes inputs, outputs, and the
flow of information through dynamic systems of human
and machinic interlocutors. Cybernetics is a technical field
grounded in mathematics, but Moles reasons that art’s ambig-
uous rubrics actually make it more compliant to cybernetic

39
modeling than science. With remarkable prescience, Moles
introduces five models that rehearse how future artists may
think with computers.
The five models, designated for the purposes of this essay,
are the computer as:
1. Artificial Auditor
2. Intelligence Amplifier
3. Permutational Algorithm
4. Simulator of Artistic Creation
5. Integrator of Forms
These models suggest alternative relationships between art-
ists, artworks, machines, and the external world precipitating
from machine-based art. Moles diagrams entire systems of
control in the discipline, theorizing computation’s potential
disruption of not only artistic production but also dissemina-
tion and criticism. Each model of artistic praxis centers on a
problem or object of study, specifies computational methods,
and delimits potential results. As such, these are not proce-
dures of machine-based artistic production, but comprehen-
sive models anticipating the reconfiguration of art’s processes
in an information society.
Today, over 50 years after Moles’s article, architecture
has undergone shifts brought about by near-ubiquitous com-
puting, unprecedented speed, and new modes of networked
labor, distribution, and evaluation. What was once labeled
digital architecture is no longer a discrete category, and
emerging subgenres of digital work challenge architecture’s
conventional frameworks of analysis. In the face of such
systemic changes, architecture requires methods of inquiry
inclusive of other cultural forms beyond visual aesthetics.
Moles’s five models offer novel means of reconsidering the
shifting roles of both the architect and the computer within a
milieu of contemporary digital pluralism.

The Auditors
In his first model, Moles proposes the machine as an auditor (or
spectator) that assists the “aesthetician” in filtering content and
bringing it to a collective public attention; the external world
is exhaustively digitized and easily accessible. The aesthetician
is then elevated to the role of a critic appropriating material to
generate new work – not creating or designing in a traditional
sense, but searching, filtering, sorting, and reframing.3
Stock-a-Studio’s XS to XL embodies this model by com-
bining material and digital circulation with a keen attention
to aesthetic choices hidden in plain sight. Financed with a

40 Log 50
Stock-a-Studio, XS to SL, Landscape
paver stacks in room, 2017. Image
courtesy the architect.

$200 budget, XS to XL is a collection of architectural portraits

of off-the-shelf construction materials, featuring banal pack-
aging and a muted palette of soft grays, blues, browns, and
beiges. Purchased samples are photographed and recomposed
as uncanny architectural subjects using crude digital editing.
The resulting synthetic portraits feature ambiguous lighting,
perspectival errors, spatial miscues, and digital glitches, while
ensuring that their constituent elements remain discernible.
The materials are temporally frozen in circulation, reconfig-
ured as easy-to-share images, while in real life the pristine
materials are returned to retailers.
Stock-a-Studio’s impulse to locate the artful in the mun-
dane echoes Moles’s observation that “the critic will be trans-
formed into an artist, as soon as [they] place a frame around a
small piece of asphalt, which [their] infallible gaze will have
3.  The displacement of artistic innova- appreciated as being aesthetic.”4 In this model, the designer
tion from acts of production to acts of
contextualization parallels art critic
initiates the process and defines selection criteria while the
David Joselit’s “epistemology of search.” machine influences the process through the implicit biases
See David Joselit, After Art (Princeton:
Princeton University Press, 2013). of search queries and the database’s content. Working in the
4.  Moles, 219. Auditor model, Stock-a-Studio’s work exposes digitalization
not as a smooth continuum from real to virtual but as a com-
plex and conflicted aesthetic terrain with material, political,
and social dimensions worthy of closer scrutiny. Through
purposeful acts of appropriation and contextualization, they
rehearse their work within contemporary modes of production
and circulation, interrogating the content formats and ranking
algorithms that organize and standardize digital exchanges.

41 Log 50
 The Amplifiers
Moles’s second model is the computer as a tool to amplify a
limited human mind. Here, the designer is confronted with
a situation too complex to address without machine aid, and
thus rigorously translates it into a series of tightly framed
problems calculable with well-defined algorithms. The art-
ist is an initiating source of intellect, but the computer per-
forms countless calculations on their behalf and results are
evaluated algorithmically for fitness. Of all the models Moles
proposes, this aligns best with the conventionally understood
“computational design” of the last 15 years.5
ICD/ITKE Research Pavilion 2014–15, Achim Menges’s research group at the University of
Stuttgart, 2015. Fiber layout draw- Stuttgart’s Institute for Computational Design and Construction
ing. Project team: ICD, Institute for
Computational Design and Construction
(ICD) combines biologically derived concepts with paramet-
(Professor Achim Menges); ITKE, ric and algorithmic methods in a research practice dedicated
Institute of Building Structures to further integrating the computer into design and manufac-
and Structural Design (Professor
turing in architecture. In their 2014–15 ICD/ITKE Research
Jan Knippers). Scientific develop-
ment: Moritz Dörstelmann, Valentin Pavilion, Menges and his team used agent-based modeling to
Koslowski, Marshall Prado, Gundula lay out hundreds of reinforcing fibers structuring a pneumatic
Schieber, Lauren Vasey. System devel-
bubble. In agent-based modeling, the programmer creates
opment, fabrication, and construction:
WS13/14, SoSe14, WS14/15: Hassan numerous software entities (intelligent agents) that act autono-
Abbasi, Yassmin Al-Khasawneh, mously in a software environment to achieve a user-defined
Yuliya Baranovskaya, Marta Besalu, goal. In the case of the research pavilion, these agents were pro-
Giulio Brugnaro, Elena Chiridnik,
Tobias Grun, Mark Hageman, Matthias
grammed to negotiate fitness criteria too complex to balance
Helmreich, Julian Höll, Jessica Jorge, manually. The final pavilion design features an ETFE mem-
Yohei Kanzaki, Shim Karmin, Georgi brane supported by 45 kilometers of woven carbon fiber. Each
Kazlachev, Vangel Kukov, David Leon,
undulating fiber of the pavilion’s lattice indexes the pathway of
Kantaro Makanae, Amanda Moore,
Paul Poinet, Emily Scoones, Djordje a single intelligent agent.
Stanojevic, Andrei Stoiculescu, Kenryo Architects working in the Amplifier model employ the
Takahashi and Maria Yablonina;
computer’s aid to produce, in technical terms, the most with
WS14/15: Rebecca Jaroszewski, Yavar
Khonsari, Ondrej Kyjanek, Alberto the least. Design is discretized into computable problems
Lago, Kuan-Ting Lai, Luigi Olivieri, and measurable rubrics for evaluation. Quantifiable criteria
Guiseppe Pultrone, Annie Scherer, like height, weight, width, mass, and volume are factored
Raquel Silva, Shota Tsikoliya; with the
support of: Ehsan Baharlou, Benjamin
to demonstrate efficient relationships between materials
Felbrich, Manfred Hammer, Axel and final forms – the lightest pavilion per square foot, the
Körner, Anja Mader, Michael Preisack, thinnest filament to span a distance, etc. Early work exem-
Seiichi Suzuki, Michael Tondera. © ICD/
plifying the Amplifier model was often characterized by
ITKE University of Stuttgart.
part-based construction logics and geometric complexity
that registered the intellectual labor displaced by machines.
Contemporary work echoing the Amplifier model includes
the design of efficient systems at the microarchitectural scale
of material science. Researchers like Dylan Wood (also of the
ICD) program materials to create responsive, self-actuated
surfaces that can adapt to environmental stimuli, such as
changes in humidity or temperature.

42 Log 50
 The Permutators
In the third model, Moles proposes the computer as a machine
for permutations. Permutators use generative algorithms
manipulated by dynamic parameters to produce fields of pos-
sibilities from discrete elements, and methodologically filter the
results. This method often enables novelty or innovation within
the synthetic cohesion of permutations. While Amplifiers seek
an optimal solution, Permutators generate repertoires or bod-
ies of work. Moles regarded permutational art as well suited to
a neoliberal society, given its capacity to produce “personalized
5.  See Achim Menges and Sean Ahlquist, motifs” from the same algorithm.6
eds., Computational Design Thinking
(Chichester: John Wiley & Sons, 2011).
Numerous contemporary practices demonstrate per-
6.  Moles, 220. mutational tendencies in their work, but perhaps none more
7.  Elena Manferdini, “Building Portraits,”
The Plan Journal 1, no. 1 (2016): 15.
overtly than Atelier Manferdini in Building Portraits, a series
8.  See Moles, 220. of intricate elevation studies that Elena Manferdini claims
reflect contemporary sensibilities for “diversity and hetero-
geneity.”7 Based on facade elevations of Mies van der Rohe’s
Lake Shore Drive Apartments, Manferdini’s drawings employ
scripted operations of layering and misregistration (the mis-
alignment of color layers in printmaking) to produce ambig-
uous spatial effects. The drawings are graphically dense to
the point of impenetrability, a virtuosic interplay of elements
– fields, shapes, color, lines, and pixels. A distinct sensibility
emerges in the recurring visual motifs of grids and shadows.
Across the 42 images in the series, the viewer is able to trace
moments of invention and maturing virtuosity as novel tech-
niques are discovered and less effective ones discarded. In
their best moments, these drawings deploy patterns and fields
to lull the viewer’s eye before closer scrutiny reveals repre-
sentational paradoxes.
With permutation, as in the Amplifier model, designers
translate their intent or concept into the virtual realm of code.
Content emerges through serialized, cyclical operations. As
seen in the Auditor model, designers act as disciplinarians and
rule makers, using selection criteria to tune the system. These
cycles of production and judgment require a dynamic feed-
back between algorithm, designer, input, and output. Suites
of permutational outputs imply emerging habits or behaviors,
complementing or supplanting traditional design sensibilities.

The Teachers
Moles’s fourth model entails machine-based acts of composi-
tion based on existing artistic oeuvres. Methodologically, this
model requires a dialogue between two machines. To begin, a
machine performs statistical analysis to unlock aesthetic and

43 Log 50
 structural secrets embedded in the original artworks. The first
machine then propagates replicas based on the rules and ele-
ments gleaned from the analysis, and a second machine com-
pares the replicas to the originals.8 Moles’s two-machine model
is an uncanny analogue for contemporary generative adversar-
ial networks (GANs), which emulate existing works with deep
learning models. In GANs, a “generator” algorithm produces a
series of “fakes” and a “discriminator” algorithm judges their
authenticity. If the generator fools the discriminator, the pro-
cess is deemed successful and the “fake” is considered legiti-
mate. GANs utilize backpropagation (a process of comparing
output to a desired output) to gradually improve the intelli-
gence of the system and produce novel copies, making them
ideal to approximate new work based on previous models.
In his recent research with GANs, architect Stanislas
Chaillou demonstrates their morphological potential.
Working at the Harvard GSD under advisor Andrew Witt,
Chaillou developed a “generation stack” – a series of three
GAN-based models for organizing building layouts at every
scale, from massing to furniture. Chaillou used dozens of
Atelier Manferdini, Building Portrait existing site plans, floor plans, and furniture layouts as train-
V, from “Building the Picture,” an ing sets to develop the underlying intuition of the GANs; this
exhibition at the Art Institute of
Chicago, March 7–September 20,
allows the user to simply specify a few parameters, and the
2015. Archival ink prints, 33 by 48 GANs fill in the blanks. His work concluded with a large-scale
centimeters. San Francisco Museum housing proposal featuring GAN-generated plans derived from
of Modern Art permanent collection.
four separate style-based training sets, including collections of
Drawing courtesy the architect.
baroque and Victorian plans. The resulting images read as con-
voluted collages, revealing traces of their stylistic inspiration.
Problems of architectural morphology and typology are,
by their nature, predictable and repeating; Chaillou’s method
acknowledges and even encourages this in laying out sophis-
ticated and large-scale plans. The convenience and efficiency
of the method, however, overshadows more crucial cultural
implications of GANs. GANs only “know” what they are
“told” through the training set, which is essentially canon-
ized; they have no capacity for causation or reasoning. As
Chaillou’s stylized plans show, biases from the training set
propagate in the generated work.

The Dreamers
Moles’s final model describes an integrated machine for
detecting subliminal connections between disparate works to
produce novel forms. Moles envisions a system of deep, lay-
ered analysis, capable of uncovering imperceptible charac-
teristics and hidden relationships between existing artworks.

44 Log 50
Stanislas Chaillou, GAN-generated
plans, 2019. Stills from transforming
GIF. Drawing courtesy the architect.

The most obvious contemporary corollary is Google’s

DeepDream, which uses a deep, convolutional network to
produce hypnotic images by mapping semantic characteris-
tics of one image to another. DeepDream essentially algorith-
mically replicates pareidolia: the human capacity to identify
meaningful patterns in random information, as when one sees
a face in a cloud. In this model, the designer acts as a teacher,
gradually introducing new content in response to the algo-
rithm’s often surprising results.
The firm LAMAS (Vivian Lee and James Macgillivray)
has emerged as an early pioneer of architectural design with
machine learning. Lee and Macgillivray’s Delirious Facade
project generates surreal facades from the semantic interplay
of neural networks. LAMAS trains algorithms to search for
and amplify architectural elements in an image to the point of
recognition, as in their attempt to find the iconic arched win-
dows of an old Beaux-Arts factory within the angular bays of
a postmodern facade. New architectural content emerges in
the hallucinatory traces the algorithm’s search leaves behind,
and serendipitous hybridity occurs as the algorithm searches
for latent traces of that which is not there. LAMAS’s recent
experiments enable new forms of contextualism possible
with machine learning, an understandable approach given
DeepDream’s propensity for producing images that induce
“in-between” readings of traditional architectural styles.

45 Log 50
 The negotiation of style raises a key aspect of both the
Dreamer and Teacher models: algorithms can no longer be
considered agnostic or even fully understood. It is increasingly
difficult to predict a neural network’s behavior and reveal
what the computer has done, diminishing the possibility that
future users will be able to “think” like machines or “control”
them. The affordances and tendencies of computer vision and
machine learning, which have recently faced criticism for
harmful inherent identification biases, require critical scrutiny.
With these models of production, it is incumbent upon archi-
tects to carefully consider both the constitution of the training
sets used and the protocols of machine vision deployed.

Displaced Roles
As Moles remarks, “In short, cybernetic analysis constrains
us – and this is not the least of its merits – to change our per-
9.  Moles, 218. spectives and our scales of value.”9 These five models reflect
new modes of labor, exchange, participation, and cooperation
intrinsic to digital design. Collectively, they reveal how the
unified and mythic roles of architect and computer are ren-
dered increasingly intractable in a networked society. Moles
predicted a future of machines of varied discretion – the
computer as a discrete black box, a number cruncher, a digi-
tizer, a storehouse of culture, a pair of dueling algorithms, an
analytic stack, or a global network. In each case, the computer
does not merely replace the architect; it displaces the archi-
tect’s role in new models of creative production.
In the cases of the Amplifier and Permutator models, for
example, designers marshal their work into tightly framed
problems that can be translated to the virtual realm of code
and computed algorithmically. The machine serves as a kind of
role model, where the design process emulates the computer’s
(purported) rationality, neutrality, and predictability. In the
Auditor, Teacher, and Dreamer models, designers participate
in networked exchanges of cultural forms. This work seeks
commonality, with alterations to perception elaborated by net-
worked machines, neural networks, and opaque algorithms,
effectively revealing how computers affect our judgment.
Moles’s models reflect many forms of contemporary
digital design, but they are certainly not comprehensive.
Platform economies, for example, enable highly distributed
acts of authorship and labor not considered in his essay.
Still, Moles’s cybernetic models are particularly relevant for
architects caught up in today’s euphoric moment of digi-
tal dispersion and acceleration. Digital design in architecture

46 Log 50
Lee And Macgillivray Architecture
Studio (LAMAS), Delirious Facade,
2016. Image courtesy the architects.

has moved from the assembly-line logics of the fabrication lab

to the transactional and managerial aggregate of platforms.
Architects working on today’s postdigital frontier reject uni-
fied myths of computers and interrogate the cultural or social
dimensions of machines through their work. No longer exclu-
sively pursuing formal innovation, these designers take on
myriad roles while probing the affordances, tendencies, and
10. These designers and their projects have biases of digital tools in their work.10
internalized a dire warning from Paul
Valéry, who writes that “[The Machine’s]
Take the work of architect Curtis Roth, for example,
exactitude, an essential feature, cannot whose networked paintings materialize the traces of the digi-
tolerate the imprecision or social caprice. . . .
It tends to eliminate those individuals who tal self through a managerial creative practice. Roth crowd-
are from its point of view, imprecise, and to sources information from HITs (human intelligence tasks)
reclassify the others, without respect to the
past or even the future of the species.” Moles, performed by online workers and collects digital fingerprints,
“Cybernetics and the Work of Art,” 223. including IP addresses, browser settings, server locations,
time stamps, and, in his most recent work, eye/mouse move-
ments. For his series Wrist Painting, Roth developed software
to harvest users’ mouse movements and gestures – the same
gestures used by web analytics giants like Amazon to algo-
rithmically deduce users’ eye movements. Roth collects data
from interactions of online workers with his software and
physically renders the space of the screen in large-scale paint-
ings fabricated with DIY tools and open-source software.
His recent efforts include a productive feedback loop with a
select group of online workers in order to develop a collective

47 Log 50
 intelligence akin to artistic intuition through managerial
practices. This workflow echoes tendencies from the Amplifier
and Permutator models, but establishes novel reconfigura-
tions of artistic practice in contemporary networks. Roth’s
work addresses a central concern of postdigital design – how
the architect’s role is fabricated in digital networks.
Moles’s Auditors, Amplifiers, Permutators, Teachers,
and Dreamers characterize distinct tendencies discernible in
today’s discipline, although, as Roth’s work demonstrates,
these models are often composited into novel forms of praxis.
Curtis Roth, Wrist Painting, 2020. In architecture, some are more prevalent than others. The
Image courtesy the artist. Permutator model, for instance, can be detected in many digi-
tal practices focused on seriality or versioning. The Teacher
and Dreamer models have not been broadly adopted in prac-
tice, although a burgeoning interest in machine learning
and the inclusion of these tools in consumer software have
resulted in an explosion of work in this area. Moles’s models
are not exhaustive and inevitably new models will emerge,
but what is clear is that machines of increasing discretion will
only further the need for architects to reconceptualize their
role in and relationship with digital design. Technological
developments are rapid, capricious, and unpredictable; archi-
tects can only prepare by increasing their capacities for adap-
tation and novel thinking through principles of role play
– altering behavior, developing personas, shifting perspec-
tives, realigning values, building empathy, and working col-
lectively – all while training for real-world scenarios. The
future of digital design will not be determined by machines,
but in how our relationship to machines is reconfigured in
new models of digital design. Moles’s models are a helpful
framework through which to critically reconsider the roles
we allow machines to fulfill and, by extension, the roles we
choose to play.

Erik Herrmann is assistant professor

of architecture at The Ohio State
University’s Knowlton School and codi-
rector (with Ashley Bigham) of Outpost
Office. He is also a German Chancellor
Fellow of the Alexander von Humboldt
Foundation and a MacDowell Fellow.

48
Eric Höweler
Verify in Field:
Models and Other
Useful Fictions
All models are wrong, but some are useful.
– George Box, “Robustness in the Strategy
of Scientific Model Building,” 1979

The model is a “useful fiction” – an imperfect approximation

of reality, just useful enough, at times, to serve as a means to
predict conditions and act on the physical and material world.
If statistician George Box’s statement is true for design, then
the range of types and the relative precision of models are
contingent on verification for their efficacy. No longer limited
to an anticipatory mode of representation, the contemporary
model is actually multiple models, each tailored to inform the
design process and instrumentalized to act on the world. This
action, or verification, allows the imperfect and contingent
findings in the models to feed back into the process of creating
the “real”; and on occasion, the “real,” the built structure, is
used to verify the models themselves.
The recently completed Collier Memorial forms a shal-
low arched portal at the entrance to MIT’s East Campus. It is
made of 32 solid granite blocks that form a five-way irregu-
lar vault. The memorial is undeniably physical and real, yet
it appears implausible, almost uncanny. The curvature of the
vault looks too shallow. The buttress walls appear too thin.
The compound curvature of the stone seems out of scale,
like the imprint of a large pebble in clay or a scooped-out tub
of ice cream. To us, as its designers, the memorial looks just
like the hundreds of models that preceded it, from the digital
models we orbited around in our computers to the compu-
tational models that predicted its forces and the miniature
physical block models that demonstrated its physics. But the
built memorial itself can also be understood as a model by
measuring the actual displacements of the blocks in order to
confirm the forces forecasted in the digital models – that is, to
verify the accuracy of our anticipatory models. Like the alle-
gory of the map in Jorge Luis Borges’s story “On Exactitude

49
 in Science,” in which a 1:1 scale map covers the territory, the
Collier Memorial is a 1:1 scale model of itself – both a rep-
resentation and real at once. What for Borges is absurd is for
us productive in the reordering of assumed hierarchies and
sequences of models and buildings. The design workflow and
execution of the memorial show that models today are simul-
taneously anticipatory and retroactive tools of verification in
the process of ordering and reordering the world.
The Collier Memorial was built to mark the site where
MIT Police Officer Sean Collier was killed following the
Boston Marathon bombing in 2013. The concept of a missing
figure led to the design of the memorial, which reads simul-
taneously as a solid/void figure and as a series of half arches
leaning into and supporting one another to form an extremely
shallow vault. Weighing approximately 190 tons, the blocks
thrust outward toward the buttress walls, while the rebar in
the foundation grade beam resists the thrust by pulling the
ends together. Gravity loads transfer through the blocks and
Höweler + Yoon Architecture, into the grade beam before passing into the ground through 33
Collier Memorial concept diagrams, friction piles. In modeling the physics of its geometry, the built
Cambridge, Massachusetts, 2015. All
images courtesy the architects.
memorial not only translates forces into form but is contingent
on near-zero fabrication and assembly tolerance to achieve
stability. This near-zero tolerance is embedded in the form and
detail of the memorial, which no longer merely approximates
its model but verifies the accuracy of that model in the field.
The notational convention VIF (verify in field) on archi-
tectural drawings indicates that the information on the draw-
ings is incomplete and further measurement is necessary. The
note typically identifies conditions where tolerance needs to
be accommodated to address the inevitable contingencies of
construction. VIF reveals the gap between design intent and
built reality and acknowledges the disciplinary investments in
the instruments of design: models, drawings, and prototypes
that are architecture’s means for operating on the world at a
distance. The field of construction is outside the scope of the
architect and is the space of contingencies, dependencies, and
uncertainties. VIF can be understood as a disclaimer – archi-
tecture’s “fine print” that speaks volumes about the profes-
sion: our culture of risk, the role of representation, and our
agency to act on the physical world.
VIF cleaves the realm of conception from construction,
model from building, representation from execution, and
architect from builder. Beyond a strict delineation of scope,
risk, and responsibility, the separation of model from build-
ing is significant for how we conceptualize disciplinary

50 Log 50
Höweler + Yoon Architecture, Collier practices, especially as these evolve with new technologies.
Memorial, Cambridge, Massachusetts, Traditionally, the model precedes the building, and the build-
2015. Photo: John Horner.
ing is its objective. The model is anticipatory, a projection
of a future reality; once the project is built, the real building
ostensibly supersedes the model. Yet contemporary design and
construction workflows are altering these assumed sequences
of priority and precedence, introducing new instrumentality
to models as fabrication tools and challenging an inherited
assumption about the hierarchy of buildings and models. At
the same time, postoccupancy evaluations, life cycle analyses,
and material obsolescence also question the idea of the build-
ing as the primary objective of design, where the completed
building often initiates another series of models that have to
do with monitoring user behavior and energy consumption.
While VIF, at its most mundane level, addresses the physical
and contingent conditions of construction and mitigates risk,
our use of the term to frame contemporary design processes
examines it as a kind of modality, one that develops the means
for more agency, responsibility, and engagement.
Architectural drawings are considered the instruments
of the architect, ceding the physical and material work of
construction and execution to fabricators and builders.
Contemporary design software, building information models,
project delivery methods, and digital fabrication tools have
shifted the boundaries between representation and execution

51 Log 50
 in important ways. Increasingly, digital models are consid-
ered part of the deliverables, and contractors use them as a
basis for fabrication and construction. The various models
employed in the conception and construction of the Collier
Memorial exemplify contemporary digital workflow issues
relative to modeling, simulation, and verification.
The process of design and construction required at least
four types of models. The memorial’s design began as a concept
of hollowing out a figure and tracing its edges, and quickly
developed as a geometric 3D model that precisely defined the
intersection of two objects: an extruded asterisk-like figure and
a large ovoid. The latter figure was subtracted from the former
in a Boolean operation in Rhino. The geometric model also
defined the limits of the central void and conveyed a “conspic-
uous absence” as the central concept of the memorial.
The analytic model translated geometry into material,
assigning attributes such as mass, weight, and gravitational
vectors. The primary function of the analytic model was to
Höweler + Yoon Architecture, Collier calibrate the geometry of the memorial to ensure its stability.
Memorial fabrication, 2015. Structural analysis was performed both with hand calcula-
tions and physical model simulations (rocking block tests) and
in discrete element models (3DEC), thrust network analysis
(RhinoVAULT), and finite element methods (Ansys). Though
we tend to assume the surety of structural calculations, the
various methods of structural analysis yielded different results
and significant debate among the structural engineers on the
design team. The unusual construction method – the all-com-
pression, irregular, solid stone structural vault – required the
engineering team to develop new structural analysis tools and,
eventually, to find consensus around an approach that consoli-
dated the results of the various analytic models employed.
The fabrication model set up the tool paths for sequenc-
ing machinery in sawing and milling the stone blocks. The
load transfer between blocks required a high level of precision
fabrication and a near-zero tolerance. During fabrication, we
asked ourselves, “How do you know that the stone is precise
to the fraction of a millimeter if the saw is also wearing down
with each pass?” The stubborn materiality of the granite reg-
istered against the insistent passes of the diamond-tipped saw,
causing it to wear and necessitating regular recalibration of
the saw to verify its actual dimension.
Traditional arches are built with scaffolding and false-
work, and assembly is sequenced from the buttresses to the
center, where the keystone completes the arch. The necessary
precision of the fit between stones at the interfaces of the

52 Log 50
Collier Memorial blocks required that we reverse this pro-
cess, starting with the keystone and working our way out so
that any tolerance was driven outward from the center. The
five-sided irregular keystone was set first, then the five adja-
cent ring stones were fitted to it. Each interface between key-
stone and ring stone consists of three faces to accommodate a
stepped section, meaning 15 interfaces needed to fit together
precisely. The buttresses were then constructed, arraying
from the center. This reordering of the construction sequence,
required by the need for a precise fit between the stones, hints
at a larger reordering in our contemporary design workflows,
from the making of models and buildings to the use of simu-
lations in the design process, as well as in the data collected
from completed buildings that renders all built artifacts as
another set of data points for further analysis.
During the assembly of the memorial, a team of research-
ers at MIT installed strain gauges at each joint to measure any
displacement and compare the actual movements of the stones
to the structural calculations performed before construction.
Postconstruction analysis revealed minute displacements of
blocks after the removal of the scaffolding and confirmed the
structural calculations. These measurements were used to
verify the calculations and assumptions made in the design
phase. In using the built structure to confirm the displace-
ments predicted in the structural analysis model, the built
memorial is a model itself. As such, it closes a loop that begins
and ends with a model, thereby positing a new understanding
of the architectural model with expanded authority over both
process and reality.
The translation from model to building, from concept to
construction, is never linear. Despite our embrace of digital
tools and fabrication techniques, there is still ample room for
friction, gaps for tolerance and deflection, and anticipated and
unanticipated conditions that require the feedback implied
by VIF notation. Strategies for verification, integration, and
sequencing become design drivers, where the brute physicality
of materials and the informed sequencing of methods impress
themselves upon the apparent ease of digital workflows.
Box’s statement that all models are wrong but some are
useful sheds light on our uncritical embrace of models in
architecture and our unrealistic expectations of their accuracy.
Modeling and simulating, even with sophisticated predictive
tools, will always yield answers that are technically wrong but
still useful. Predictive climate change models, for instance, can-
not be 100 percent accurate but are still necessary for planning

53 Log 50
 and policy making. Indeed, in today’s culture of online com-
merce, big data, and predictive analytics, we are increasingly
dependent on models for weather forecasts, traffic navigation,
and even COVID-19 analyses to make decisions every day. In
design, “the model” has a newfound authority in building
simulations, structural calculations, material behavior, and
energy consumption. Models have essentially been instruments
of forecasting. Now they are also used to test prior forecasts,
resulting in a retroactive process of verification, or backcasting.
Models are approximations, whether they are physical
maquettes, digital files, or simulations. As digital workflows
evolve to turn models from tools of conception into instru-
ments for execution, it is easy to overestimate the precision
implied by the digital tools. With this comes an expectation
of perfection and an intolerance for the material deviations
from the digital ideal, a condition that Francesca Hughes calls
1.  See Francesca Hughes, The Architecture the “architecture of error.”1 For Hughes, the evolution of the
of Error: Matter, Measure, and the
Misadventures of Precision (Cambridge:
instruments of measurement, design, and fabrication has cre-
MIT Press, 2014). ated expectations between models and material that recatego-
rize much of the physical world as full of errors, imprecise
and unconforming. An expanded understanding of precision
produced through the intricate interplay of conception, exe-
cution, and verification is needed to realign disciplinary think-
ing about the precedence of design models in the digital age.
In Borges’s allegory, the practice of cartography was so
advanced that it produced the ultimate map at 1:1 scale. The
absurdity of the full-scale map, the map’s eventual aban-
donment, and its ruination are a warning about the limits of
representation and disciplinary overreach. Models, like maps,
are useful when they leverage their abstraction and specificity
to become instrumental. The collapse of difference between
map and territory renders the map useless, even as the alle-
gory invites further speculation on the interactions between
the two. A survey of the design process and execution of the
Collier Memorial reveals a series of models, feedbacks, and
workflows that are illustrative of the contemporary relation-
ship between model and material. The demands of a near-zero
tolerance in the design and fabrication, and the reordering of
the construction sequence to accommodate inevitable on-site
contingencies, foreground the importance of verification in
both design and construction. The scrutiny and refinement of
the design and construction process itself suggest the devel-
Eric Höweler is a cofounder of Höweler
opment of new modes of design – what might be called “best
+ Yoon Architecture and associate pro-
fessor of architecture at the Harvard practices,” or what we might call “model behaviors.”
Graduate School of Design.

54
Mark Lee
& Sharon Johnston Models and Models
Of Models
Q: Why didn’t you make it larger so that it would loom over the
observer?
A: I was not making a monument.
Q: Then why didn’t you make it smaller so that the observer could
see over the top?
A: I was not making an object.
– Tony Smith replies to Robert Morris
about Smith’s six-foot cube

As models take on various roles, from generative to perfor-

mative, three particular models of models are integral to
contemporary practice. The conceptual model is a form of
representation that is both abstract and exploratory. The
exhibition model is a form of performance developed to com-
municate a particular narrative in a specific context. And the
full-scale model is a form of approximation – shifting in focus
from the idea to the actual qualities of space and structure
while maintaining a critical distance – refined but not final.
Scale is the central factor that determines the intent of
each of these models. Tony Smith’s responses to questions
about his six-foot steel cube simply situate the piece in rela-
tion to the scale of the human body. Whether the cube is a
monument or an object depends on its scalar relation to the
subject. Similarly, while an architectural model often serves
as a device to document an approach or intent, it is ultimately
the model’s scale that defines its role and efficacy. A small-
scale, object-like model allows for an omniscient view of the
developing concept behind a building, whereas a full-scale
model approximates a building or a fragment of a building
at its intended scale, allowing for a different level of detail
or study and an experiential engagement with the building,
while remaining disassociated from its final context. Scale is
not only a medium in service to the model’s objective but also
the very condition that defines and drives the model’s impact.

The Conceptual Model

Of the many models, the conceptual model inaugurates the
design and sets the tone for the rest of the design process, often

55
Johnston Marklee, Vault House, Oxnard, California, 2008. Foam core and
museum board model. Right: acrylic and foam massing model. Top: Hill House,
Pacific Palisades, California, 2004. Foam core and basswood model. Right:
acrylic and cardboard massing model. Photos courtesy the architects.
encapsulating the central idea behind the project. Usually small
in scale, and always open-ended, conceptual models address
problems and give shape to ideas through an in-between con-
dition – between idea and building, material and immaterial,
problem and solution. Each conceptual model takes on its own
meaning as it provides a new perspective to help define a project.
The conceptual model that demonstrates an idea has a
long history in architectural practice. From the photographs
of Le Corbusier’s Unité d’Habitation and OMA’s Jussieu librar-
ies to the videos of Kunlé Adeyemi’s clay model and Smiljan
Radić’s tape-and-plastic model for their Serpentine Gallery
projects, models are seen as manipulated by godlike hands
that push and shape the forms, often to give access to the more
complex ideas behind the concept. Nevertheless, beyond act-
ing as a tool for communication, a conceptual model is most
efficacious when operating as an exploratory device.
In our own practice, Johnston Marklee, conceptual mod-
els are not simply about ideas but rather take on multiple roles
of exploration. In the Vault House in Oxnard, California,
the conceptual model gave shape to the idea of the house as
a viewing device. Through iterations of the model, vaulted
rooms were arranged tangentially, their arched apertures
aligned to channel light and views from the beach through the
length of the house. Similarly, through the modeling process
for the Hill House in Pacific Palisades, California, the con-
cept of the box-in-box organization for the space emerged.
With the ability to easily rearrange interior volumes within
the set-back envelope of the exterior, modeled in clear acrylic,
the idea of interlocking spaces materialized, conforming to
the parameters of the site. Sometimes the model is situated
beyond the project’s literal site, engaging the canon of archi-
tectural history. In the Sale House in Venice, California, an
expansion to the 2-4-6-8 House by Morphosis, the conceptual
model was a response to the urbanism projected from their
original models and drawings of that seminal project.
The ability to conceptualize architecture through mod-
els advances the design process; through formal articulation,
the conceptual model pushes ideas forward as it allows for
the exploration of spatial relationships, site parameters, and
historical contexts. Conceptual models generate unexpected
understandings by allowing us to examine relationships
between the initial thinking behind the project and the fully
formed building concepts. Openness defines the conceptual
model as a device for problem-solving and exploration – an
undetermined, generative process.

57 Log 50
 The Exhibition Model
In 1980, Paolo Portoghesi directed the first International
Architecture Exhibition at the Venice Biennale. Aligned
with his theme, “The Presence of the Past,” the “Strada
Novissima” installation in the Corderie dell’Arsenale was, as
Portoghesi described, “with architecture, not about archi-
1.  “History of Biennale Architettura,” La tecture.”1 The installation, which comprised 20 full-scale
Biennale di Venezia, https://www.labien-
nale.org/en/history-biennale-architettura.
facades, each by a different architect, reflected classical
2.  The 16 firms commissioned for “Vertical theater by presenting a simulacrum of a street. Like these
City” were: 6a architects, Barbas Lopes
Arquitectos, Barozzi Veiga, Christ &
facades, the exhibition model becomes performative when
Gantenbein, Ensamble Studio, Éric it ceases to be viewed from above; with a shift in scale and
Lapierre Architecture, Go Hasegawa, Kéré
Architecture, Kuehn Malvezzi, MOS, display, it attains a status similar to architecture. As artistic
OFFICE Kersten Geers David Van Severen, directors of the second Chicago Architecture Biennial, in 2017,
PRODUCTORA, Sam Jacob Studio,
Sergison Bates, Serie Architects, and we referred to “Strada Novissima” as a precedent for our
Tatiana Bilbao Estudio. theme, “Make New History.” Given the performative nature
3.  The 24 firms commissioned for
“Horizontal City” were: Karamuk*Kuo of the exhibition model, the installations “Vertical City” and
Architects, UrbanLab, MAIO, First Office, “Horizontal City” were central to the biennial in that both
Sauter von Moos, fala atelier, DRDH, Besler
& Sons, Norman Kelley, Andrew Kovacs, scale and content could frame a critical reading of Chicago’s
REAL Foundation, formlessfinder, The Los architectural legacy and create an environment that approxi-
Angeles Design Group, Diego Arraigada
Arquitectos, Lütjens Padmanabhan mated a walk through the city.
Architekten, WELCOMEPROJECTS,
June14 Meyer-Grohbrügge & Chermayeff,
“Vertical City” brought together 16 architects to recon-
The Living, Adamo-Faiden, Thomas sider the 1922 brief for the design of the Chicago Tribune
Baecker Bettina Kraus, Angela Deuber
Architect, Tham & Videgård Arkitekter,
Tower. Rather than calling for a single rendered perspective
Charlap Hyman & Herrero, and Bureau – the focus of the projects in both the original competition
Spectacular.
and Stanley Tigerman’s “Late Entries to the Tribute Tower
Competition” in 1980 – we eschewed the singular, monu-
mental view and instead commissioned 16-foot-tall models
at half-inch scale. At once scaled-down towers and full-scale
columns, these enormous, slender forms collectively evoked
a hypostyle hall. Architects including Barbas Lopes, Francis
Kéré, and Tatiana Bilbao created eclectic models with diverse
conceptual foundations and alternative modes of collabora-
tive assembly.2 Less a study in the tectonics of tall buildings,
the models enveloped the viewer in new visions for a vertical
urbanism, through an in-between scale that was both tactile
and seemingly inhabitable to the viewer.
Twenty-four architects were invited to reconsider the
status of the architectural interior in “Horizontal City.”3 Set
on plinths and organized according to Mies van der Rohe’s
1947 master plan for the Illinois Institute of Technology, the
models each presented a reinterpretation of a canonical
interior. Created as a voyeuristic experience, the viewer was
drawn through peepholes and window frames deep into
alternate interior worlds. As a group, the models were united
by a uniform horizon and by the order of arrangement; the

58 Log 50
Installation view of “Vertical City” diversity of historical epochs rendered together suggested a
at Sidney R. Yates Hall, Chicago new type of diverse urbanism of the interior.
Architecture Biennial, 2017. Photo
© Tom Harris. Courtesy Chicago
Both rooms connected to the city of Chicago and beyond,
Architecture Biennial. as they produced an immediate spatial experience and refer-
enced historical narratives to establish new individual histo-
ries as a collective. Beyond these two rooms, models dotted
the landscape of the biennial and each spun its own performa-
tive narrative and engaged the specific architectural history
of Shepley, Rutan & Coolidge’s Chicago Cultural Center
(formerly the Chicago Public Library) in which the biennial
was installed. The importance of the performative model was
underlined throughout the biennial and particularly in the
model-focused installation “Super Models.” Designed and
curated by Sylvia Lavin with Erin Besler, Jessica Colangelo,
and Norman Kelley, the installation looked at the importance
of the model in collecting and presenting new ideas. Housed
in a full-scale foam model of Oswald Mathias Ungers’s pavil-
ion in the Deutsches Architekturmuseum, the replica models
were organized to allow only specific framed viewpoints of
each particular artifact. Like the siting of the pavilion in the
exhibition hall, the display framed intentional views, further
reflecting how the exhibition model is designed to perform
with a strategic purpose.

59 Log 50
 The Full-Scale Model
The shifts in the size of a model, from small to medium to
large to full-scale, necessitate acknowledging the critical
thresholds of performance that change radically from one
scale to another. Like Smith’s cube, Fumihiko Maki, draw-
ing on the teachings of José Luis Sert, recalled, “You would
only understand the relationship between architecture and
an object when you are standing in front of the building at
eye level. And you will see a building from zero to 15 meters
– then you can understand the tactility of the building, the
4.  Fumihiko Maki, “Grounded Visionaries: activity of the people, the cars, and sometimes the smell.”4
The Harvard Campaign for the GSD”
(lecture, Harvard GSD, Cambridge,
Abstract in material but literal in scale, a full-scale model
December 19, 2014). allows for an understanding of a building in relation to the
body. Open and experimental like the conceptual model, the
1:1 model allows for ideas not only to be seen but also experi-
enced. Unlike a building, the full-scale model is meant to be
fleeting, fragile, more hastily put together. A full-scale card-
board mock-up of the pillowed concrete panels we designed
for the UCLA Margo Leavin Graduate Art Studios sufficed
to give clarity to the shape and scale of the pillowed walls in
relation to the human body. Shifting to full scale helped us to
refine the detail of the facade with a new experiential under-
standing of the building.
In 2016, Johnston Marklee, in collaboration with OFFICE
Kersten Geers David Van Severen and Nuno Brandão Costa,
took part in designing a pavilion curated by Diogo Seixas
Lopes and André Tavares for the fourth Lisbon Triennale to
house “The Form of Form,” an exhibition curated by Socks
Studio. With an aggregated plan of 12 rooms taken from proj-
ects designed by each of the three practices – some built and
some unbuilt – the 1:1 structure presented a new experimen-
tal approach to the concept of a pavilion. The agglomeration
of the 12 interlinked rooms created a sequence of spaces with
unexpected relations between forms and ideologies; with
each room incorporating images related to a core element in
architecture, the exhibition defined a continuous thread of
works that were directly interrelated, be it through affinity or
opposition. Conceived as a dialogue that challenges notions of
authorship and the limits of form, the installation was both
an exhibition and a structure in itself.
“The Form of Form” articulated the temporality of the
full-scale model through materiality and construction. The
collection of rooms was constructed with a simple kit of parts
consisting of metal studs and gypsum board, and requiring
material reinforcements such as brace frames for more

60 Log 50
 complex forms, which revealed the limits and forces at work
in the different buildings. While each room was a full-scale
replica, the in-between spaces that connected the rooms pro-
vided unexpected and new, if ephemeral, spaces. No longer
approximating or modeling another space, what emerged
from this project was a set of new spaces that were imperfect,
misaligned, and different from the original.
While the scale of the pavilion achieved a building-like
presence, its temporal existence, along with its provisional
material and structural qualities, situated it as a full-scale
model. The in-between nature of this full-scale model – an
open function, not quite a building and not quite a sculpture
– allowed for a different kind of testing and understanding of
the 12 respective projects from which these rooms originated.
As we pushed against a refined, finished pavilion, we instead
created a full-scale model that produced a feedback loop and
generated new conversations with the reference buildings.

Models, Models, Models

A model is not simply a model. It never stands alone but is
always part of a larger iterative set, whether the others are
present or not. A collection of models is always greater than
the single model because it is through the collection that the
model as an idea is measured against the model as a physical
form. And the models of models, whether for the study of a
concept of a building, for pointed perspectives in an exhibi-
tion, or for engaged temporal approximation at full scale, are
ripe for renewed consideration and exploration. Scale both
facilitates and punctuates the various roles of the model in
shaping the exploration, framing the conversation, and pro-
viding a surrogate that spurs imaginative modes of occupation.

Sharon Johnston and Mark Lee are

the founding partners of the architec-
ture firm Johnston Marklee. They are
professors in practice of architecture
at the Harvard University Graduate
School of Design, where Mark Lee also
serves as the chair of the department
of architecture.

61
MOS Architects, Studio No. 4, Studio with Flat Roof, 2019. Top: Lütjens
Padmanabhan Architekten, Extension of the Rüsler school facilities,
Niederrohrdorf, Switzerland, 2017. All images courtesy the architects.
Michael Meredith
A Conversation
About Models
This began like a lot of things nowadays, with a random e-mail,
about how it would be great to do something together – MOS,
Lütjens Padmanabhan or Lütjens Padmanabhan, MOS. Long
pauses between responses. This Log issue on models, maybe? E-mails
late at night, early in the morning, maybe? No idea where or when
they are. A glowing window on my phone. Back and forth a few times.
Looking at each other’s work through social media. I knew some of
their projects, but we don’t really know each other, we still don’t,
wasn’t even sure what they looked like when we started, but it’s
obvious we share architectural values and sympathies. – MM

MICHAEL MEREDITH: I fell in love with your work

through the models, the photographs of models in particu-
lar. I haven’t seen your buildings in person, but I’m sure they
would be good. I can tell from seeing the models. There’s a
sort of economy and playfulness, an ambiguity between flat-
ness and volume, games of scale that are done really well, a
good color sensibility. The abstract models have a graphic
quality, where things feel economic in a good way, not overly
articulated, luxurious, or upper-class, but industrial, afford-
able. Assemblies of ready-mades and products. Looking at
them makes me think of other architects, like Adolf Loos or
Álvaro Siza, Venturi, Denise Scott Brown or John Hejduk. In
general, abstraction allows us to find things we want to find.
Perhaps I’m projecting my interests into your work. This can
be a problem with abstraction too. But your work oscillates
between this abstraction and blankness of the architecture and
the realism of the model, where you have weight, material,
color, and so on. Where renderings or drawings rely mainly on
the visual as representation, models inevitably have to do more
work than a drawing, they have to hold themselves together,
deal with humidity, glue, structure… I’m curious what you
think of this. Do you see the model as a thing in itself or a prop
for image making? Do you have any concerns about me pro-
jecting references and/or an architecture of quotations?

OLIVER LÜTJENS: For us the model is certainly a thing in

itself. We start building working models at the 1:50 scale in

63
Top: Lütjens Padmanabhan Architekten, Building-Trade Vocational College, Zurich, 2018. Middle: Apartment building
Zwhatt Long-House, Regensdorf, Switzerland, 2019. Bottom (left to right): Apartment building San Riemo, Munich, 2017;
Apartment building, Binningen, Switzerland, 2011.
the early stages of a project, just as we gain MM: Yeah. It’s similar. We work at half inch
enough certainty about the general size and equals one foot, close to 1:50. Prototypes, frag-
layout of a building. It’s a great scale because ments, material samples, large foam core and
you can decide how much detail you want paper models have all been hanging around
to build or leave out. For instance, it might the office ever since we started. It helps us
not be relevant to build the window frames, see what we’re doing. We’re pretty literal;
but you can judge the impact of a special roof we need to see it to see it. This comes from
detail on the whole building. Hilary, she pushed models from the begin-
We also love 1:50 scale because the mod- ning. Maybe it’s from her time at OMA. I
els have a real physical presence in our remember visiting her in Rotterdam and the
office. If a project gets built, its model is in models were huge, with lots of little blocky
our studio for around two years. Even if at model furniture floating around. Models
some point we stop working on it, we look were the primary object of study, where they
at it every day. We see it from all different tested out forms, materials, etc. They were
angles, in different light, in relation to other always a wonderful mess.
projects. Even though our models are rather I think the photograph of the model has
abstract, as you noted, they stay an important become more important than the working
authority as we detail the project. Usually model at the moment. Before, the model was
the model is always right – especially when more for the office, working through ideas,
made quickly in the heat of a deadline, with now it’s for making images, they’re theatrical.
intuition rather than thought. For us, work- Both of us stage our models. At the beginning,
ing with models is a tool for looking at archi- we used to photograph everything in the dark
tecture as a language of form. A language and say it was to create a moody environment,
that is evolving, that allows you to copy or to where the architecture was more mysterious,
quote and make things your own. on the verge of disappearing or appearing. We
I think we have a lot of common ground claimed that we were sort of anti bright white
in the way our models inform our way of V-Ray rendering and pro moody darkness.
thinking about architecture. I guess some- The reality was we didn’t have the proper
times both of our built works even look a bit lighting equipment anyway. In dim lighting
like models. I asked my wife what she thought I think the model became more and more of
of the image we selected for a Christmas a stage set, a bunch of props for image mak-
card. It looks like the others, she said. What do ing, effects, and little videos. Like every-
you mean? I asked. It’s a model, like the others. thing in our office, over time, models have
When I told her it was the actual building she become more and more technically demand-
had to look twice to see the difference! ing, requiring more precision to achieve a
Lately you have found a beautiful way sort of abstract realism or surrealism realism.
of making images, photographs of models Muteness through folded paper.
with very subtle but strange computer back- After we started photographing models,
grounds. They are super atmospheric and we quickly began making stop-motion videos
remind me of surrealist painting, particularly with them too, with narratives and then non-
Dalí. Your buildings are mute but enigmatic. narrative narratives. Nothing happens in our
They are filled with lights and stuff, and are stories. Maybe this came from my experience
surrounded by the most adorable model rocks. of being in Marfa, Texas, working with Pierre
I would be curious to hear about how you Huyghe, or our love of the Eameses, or OMA.
work with models and what they mean to you. It’s hard to piece it together after the fact.

65 Log 50
After Assemblage ended in 2000, we were all I completely agree with your generous
making more than we were theorizing. I think reading of our models and projects, with one
we and a lot of our friends wanted to be in the exception regarding the notion and nature of
world doing architecture, not just represent- abstraction. We never speak about abstraction.
ing things. Models and videos seemed closer to But, as you said, there is a lot of abstraction
the world, and for what it’s worth, the models in our models. I guess the reason we avoid
in our office happened along with a renewed that word is because we’re more focused on
interest in literature, in reading and writing, other things while we work. The abstraction
after meeting some of the McSweeney’s people is more a result of the situations that we build
and David Foster Wallace, and being around in. Switzerland is very different from the
writers. I was reading The Mezzanine by United States. We usually have tiny irregular
Nicholson Baker, short stories by Donald plots that have to be used to the maximum.
Barthelme and Lydia Davis, stuff writers rec- As a result, the building volume is essentially
ommended to me that they thought I’d like. dictated by the building law. This means that
I love the story of confusing the model the incredible unity between figure and form
photo with the real photo. I think that’s how that we admire in your work is impossible
we (architects) exist in the world. So how for us to achieve. We are literally confronted
would you describe your architectural lan- with the task of transforming an ugly vol-
guage or expression? It’s not a parade of nov- ume into architecture through the design
elty, but there is a consistency to it. If I said of the facade! That also explains the strange
it was both refined and clunky, or if I said it slippage between interior and facade that we
was playful, abstract but also not radically often experience when we struggle to find an
new looking, would you be okay with that? I adequate exterior expression.
see a careful calibration in your work. At first So that’s one reason. We also look to the
glance it looks effortless, almost ready-made, history of architecture in constructing frame-
or sort of undesigned, but is very careful and works for our buildings, engaging with things
precise when you look a little closer. There like Italian architecture from the Renaissance.
are interesting symmetry-asymmetry games We’re historical, but not vernacular.
in the facades and plans. I think we share
this desire for certain effects, but maybe our OL: Yes, our architectural expression is
architectural expression is tied to our differ- very much tied to the specific conditions of
ent building conditions. For us that’s pitched building in Switzerland. In fact, I would
roofs and wood-frame construction, cheap even say we found our means of expression
construction techniques. Yours is perhaps through building. This might sound very
more concrete and steel? Swiss, but it’s not the refined way of build-
ing like Diener & Diener, Peter Zumthor,
THOMAS PADMANABHAN: Your work or Valerio Olgiati. We never had the oppor-
has something that Oliver and I were hoping tunity to rely on reduction, the essence of
would come to Europe from America for some material, and the value of craft like the gen-
time: an architectural reflection on the pos- erations of Swiss architects before us.
sibilities of the American house, largely made Our projects are conventional pro-
with the usual elements and materials of ordi- grams, like apartments. They almost always
nary American houses. Half a century after come with the need to build economi-
Frank Gehry’s early projects, finally. This is cally. Therefore, we work our way through
how I like to see your work, and your models. the constraints and resistance of catalogue

66 Log 50
Top: MOS Architects, House No. 10, House with Courtyard, Cambridge, New York, 2019. Middle: Cabins No. 1, Hudson, New
York, ongoing. Bottom (left to right): House No. 12, A Foam House with 98 Blocks of Foam and 8 Doors, 2017; Studio No. 4,
Studio with Flat Roof, 2019.
products, standardized systems, and the need way. Thomas was speaking about the
to heavily insulate. The cheapest way to build American house. I was wondering, is your
in Switzerland is a raw construction of brick work less about the house and more about
walls and concrete slabs on which you add 20 the American landscape?
to 30 centimeters of insulation to the outside.
With a layered construction like this, the means MM: All the superstar-architect type of work
of expression are only found within the last is unavailable to us too. I love it, but it’s too
few centimeters (or sometimes millimeters). refined, too elegant, too expensive. For bet-
This fundamental condition leads us to explore ter or worse, we are stuck working with
an architecture that is relying on surface really cheap budgets. Clients looking for a
rather than mass. It is light and thin, it is liter- bargain. Maybe because of this I feel like our
ally open, for it has joints and cracks, it is fac- work is becoming more and more American.
eted, layered, and plastic. We omit the closed It feels strange to say that. I am uncomfort-
box for an open figure. able with nationalist narratives, Swiss and
I am not sure if I can describe our archi- American, etc., but at the same time there are
tectural language, maybe Thomas can. But histories and specific situations that we are
when we design a facade, for instance, it cer- all dealing with. It really is difficult for me
tainly feels like using a kind of general archi- to say American architecture, it feels back-
tectural language. We work with proportion, ward, although I say Japanese, Chinese, Swiss,
order, rhythm, and composition. We are not Belgian, Spanish, Russian, Chilean… So much
thinking in images or in types, therefore we of history and curatorial agendas are framed
have no shortcuts at hand. Like in writing, within nationalism, shows like “Contemporary
it is an arduous process of putting one thing Japanese Architecture.” That said, I can’t think
next to another. of any shows on American architecture.
I admire the ease with which you use vol- I think the school and the house you
ume in your work. It’s like a language made mentioned have a physicality that might
out of objects, an assemblage of types and come from working in models and may be
figures, such as the Krabbesholm Højskole indebted to specific narratives of vernacu-
school, where two building parts seem to lar construction, industrial construction,
nudge, or the main elevation of the House pragmatism, a sensibility of materiality
with Courtyard, which is formed by two and structure, economy, bluntness, object-
equal pieces (or one broken piece). These are ness, literalism, blankness, playfulness, stuff
moments when seemingly autonomous enti- embodied in figures like Venturi Scott Brown,
ties are looking for a form of connection. Ray and Charles Eames, Louis Kahn, Anne
It reminds me of works by Aldo Rossi and Tyng, early Frank Gehry, Charles Moore,
Ettore Sottsass. In a similar way, we construct John Hejduk, Donald Judd… The list goes on.
a volume using different planes and layers, Last year, Hilary and I taught a design stu-
although we do not think of it as a volumetric dio that revisited the MoMA Things in the
entity from the start. Making: Contemporary Architecture and the
You said that the way you were using Pragmatist Imagination conference in 2000.
models became more theatrical. I wonder It was such an interesting event. I reread
if not only the models became more the- the debate between Cornel West and Rem
atrical but also the buildings. A bit like the Koolhaas, which seems more important now
Palladian villas, which create this amazing than at the time. West argues against Rem and
sense of place in a gentle yet monumental neoliberalism, suggesting that we can’t just

68 Log 50
MOS Architects, House No. 10, House with Courtyard, Cambridge, New York, 2019. Plan on previous spread.

erase national narratives because that would TP: Pragmatism! We are so envious of the
be inherently ahistorical, while Rem takes a pragmatism in your work! It is pragmatic,
very neoliberal view and says they’re prob- which to us is different from just being practi-
lematic, preferring a globalist perspective cal. We build in Switzerland, where the weight
in which there isn’t necessarily a significant of rules makes building incredibly expensive
distinction between France and Germany, and heavy, not only in practice but also in a
for example. West’s argument is that with- conceptual way. The cost difference between a
out narratives like our national identity, race, cheap and an expensive building is minimal. So
class, history, the discipline, and so on, capital where do you find the freedom and the wiggle
and corporate power will fill their vacuum. room to develop direct, unfiltered work?
Our histories, as horrible as they might be,
are important. Without them things become MM: Yeah, it’s a very different situation, a
too generalized, or universalized. West’s different culture of construction, labor, and
argument sounds right to me. liability. We work through small projects that
We’ve been interested in capital P prag- are under $300 per square foot, very bare-
matism and the particularly American history bones projects. I don’t think this type of work
of it. Trying to reconcile scientific empiri- exists there. Over time, we’ve learned to love
cism with nonrational/scientific belief struc- things like pitched roofs, blankness, corru-
tures. We can’t remove material construction gated metal, aluminum storefronts through a
and economics from form. I love what you sort of necessity. They are just the elements of
said about the layers of construction, know- economic building, not signifiers of “archi-
ing how to manipulate that last few cen- tecture.” Models are similar. We have limited
timeters is maybe enough, and this idea of means and try to figure out how to work with
putting one thing after another, piece by what’s readily available – foam core and paper.
piece. I once heard Kurt Vonnegut (I think it Do you build the model like the building? It
was Vonnegut) talk about his writing process, seems to me that the act of layering material
which was similar, building up piece by piece. versus casting a solid model makes it more

69 Log 50
like a building. Do you think of it as a build- MM: I know what you mean about the
ing, not just as a model or representation, but kitschiness of renderings. The almost photo-
something that has its own life? What makes it realistic ones have a sort of glossy corporate
different from a photo-realistic rendering? tone, and even though this may be a sort of
pre-Warhol Clement Greenberg thinking,
TP: We really do build our models like build- the corporate renderings, representation, and
ings. When we work on a competition, we build cheesy commercial stuff still feel wrong to us.
the interior structure of the model as soon as That said, someone like WOJR does render-
we solve most of the floor plan problems. Then ings that somehow are so real that they don’t
we work on the models and the plans in paral- fall into this reading. They feel like photo-
lel. We stick parts of potential facades to the graphs. I know this sounds naive, but we are
structural model, look at the model, take the academic practicing architects, an oxymo-
parts off, and rebuild it. Over and over. ron. And our work is not always clear even
Today, computer renderings used in to us, which is why we are constantly mak-
architectural practice privilege a perspectival ing things, looking, searching, making again.
view of form and space. We are skeptical of With this in mind, I guess a basic question is
perspectives as they feel too subjective to us. why build physical models when it’s easy to
They are about one viewpoint and end up make digital ones. Maybe it’s even easier to be
being more about the composition of the completely digital nowadays. This makes me
image than the form and composition of the think about Brian Massumi’s argument in “On
work that is being represented. In perspec- the Superiority of the Analog” and Alexander
tives, proportions are distorted and complex Galloway’s response. For Massumi, the analog
tectonic ideas are lost in the dynamic char- is the system that allows for the digital. The
acter of the image. Models, and even photo- analog is primary and the digital is second-
graphs of models, retain a certain objectivity, ary. Galloway argues that analog and digital
a factual reality outside of the observer and are equivalent and inextricable. Our work has
outside of the author. There are always dif- constantly played in this collapse of the digital
ferent ways to see the work through models, and analog, the early software experiments
similar to built work. To us, model photo- with video game physics, our representation,
graphs are a way to avoid the propaganda and our writing. Thinking about models in par-
kitsch of computer rendering. Still, in model ticular, a digital-analog spectrum is a way we
photographs there is an inherent danger that can locate all practices nowadays. We all work
they can start to look like poorly executed in this space in between. Some practices are
renderings, especially when their context is more digital and some more analog. This said,
done in a more naturalistic way. Our mod- I suppose we would err on the side of the ana-
els look best when they are photographed in log over the digital because, in the end, we are
a cold, nonatmospheric setting. When the architects interested in making buildings in
context is cold, the architecture feels warm. the world – larger-scale models.
We try to give only minimal clues of its urban
and topographic context, similar to the late
portraits of Titian, where a few loose brush-
strokes are sufficient to define the landscape
or domestic space in the background.
Michael Meredith teaches and is an architect who wishes
OL: We leave it to the imagination. he had more projects.

70
Ferda Kolatan
The Chunk Model

In recent years, a new kind of drawing has emerged as a popu-

lar addition to the repertoire of architectural representations.
The chunk model, as it has come to be called, is a 3D digital
model rendered as a 2D drawing. It usually depicts fragments
of a project as tightly packed, free-floating objects over neutral
backgrounds. The compact character of these objects, paired
with a highly detailed and realistic visual language, produces
an appearance typically referred to as heavy, dense, or chunky.
However, as is often the case with single-word nomenclature,
chunk, being punchy rather than differentiated, has become
the somewhat overextended umbrella term for a wide range
of rendered models with little or no commonality – like the
generic massing model or the cutaway axonometric.
Perhaps, then, it is useful to introduce one particular ver-
sion of the chunk model, developed both in my practice SU11
Architecture + Design and with my students at the University
of Pennsylvania, and describe its intentions, properties, and
behaviors as well as how it differs from other representational
types. For us at SU11, chunkiness is a direct consequence of
our interest in an architecture that fosters odd combinations,
exposes hidden relationships, and views all its peripheral parts
as equally valid inputs for design. To articulate these ideas,
we began forging multiple building components together and
cutting unorthodox sections to reveal the material and spatial
effects created by the fusion. Our chunk model is thus rooted
in a relational ethic that encourages unscripted interactions
between all architectural elements, including the ordinary,
arbitrary, and neglected.
The merging of disparate objects into new hybrid con-
stellations also reflects larger cultural tendencies. The desire
to embrace the ambiguous, to find unexpected collaborations,
and to privilege the mixed over the pure is, conceptually and
materially, a significant marker of our times. In the history
of architecture, the pure has been conventionally associ-
ated with classical Western traditions that herald humanist
principles and are visualized by anthropocentric representa-
tions like linear perspective. Unsurprisingly, in such tradi-
tions any ambiguous design effort was generally deemed as

71
Johann Wolfgang Baumgartner, La
Poesie. Die Dicht-Kunst, 18th century.
Engraving by Jeremias Wachsmuth
after design by J.W. Baumgartner.
7 3/4 by 11 1/8 inches. © Victoria and
Albert Museum, London.

ill-conceived, confused, or simply an aberration. But one of

these historical deviants, long maligned and ridiculed as friv-
olous, became a guiding precedent in our own investigation
of architectural hybrids and strongly influenced the specific
features and overall character of the chunk model. This prec-
edent is the rocaille.
A truly peculiar artifact, the rocaille not only gave the
rococo era its name but also provided a unique model to radi-
cally alter long-standing notions of design. Composed of
various elements, some natural, some artificial, the rocaille
blurs categorical lines in pursuit of a hybrid aesthetic. Rocks,
shells, and trees freely aggregate with arches, columns, and
volutes to reject strict delineations and encourage misfit
crossovers. Furthermore, the rocaille avoids fitting into any
of the classical art forms. It is neither architecture nor sculp-
ture nor painting. And yet, this strange object could morph
into all of the above by just scaling up or shifting shapes to

72 Log 50
Yuanyi Zhou and Wenjia Go, Machine
Park, Istanbul, 2018. Chunk model
rendering for a terracing design in
Sanatkarlar Park, Istanbul. Water
irrigation infrastructure, soil enforce-
ment, handrails, lighting fixtures,
arabesque tiling, and concrete
muqarnas combine into a new hybrid
architecture. This student project was
designed in the Oddkin Architecture
Istanbul design studio instructed
by Ferda Kolatan with Michael
Zimmerman, Stuart Weitzman School
of Design, University of Pennsylvania.
Drawing courtesy the author.

adapt to vastly different contexts. This impressive versatil-

ity and transgressive behavior are possible because the rocaille
functions primarily as a semiautonomous object, simultane-
ously embodying the world and detached from it.
In considering our idea of the chunk model, a number
of compelling parallels to the rocaille become evident. The
blending of items from multiple categories into a single coher-
ent expression, the featuring of the model as a free-floating
independent object over a neutral background, the revealing
of otherwise inaccessible views by exposing the under- and
backsides of rocks, foundations, etc., the representation of the
object as at once fragmented and total, and the emphasis on a
photo-realistic and partially fictionalized language across the
model. All these aspects describe inherent properties of the
chunk model and are critical to its aesthetic and to the ways in
which it behaves.
For example, the first model here depicts a composite
garden artifact made from organic matter, ornament, infra-
structure, and architecture. But rather than enforcing such
categorical boundaries, the model uses material differences to
create crossovers that may lead to unforeseen design inno-
vations. For instance, ordinary bits like mechanical piping,
guardrails, or lampposts intertwine with residual markings
left by construction machines on the concrete walls and the
landscape. Or richly ornamented tiling patterns (native to the
cultural heritage of the project site in Istanbul) reach across

73 Log 50
 functional lines to join up with the metal tank of the garden’s
irrigation system. In each such scenario, value judgments
based on deep-seated dichotomies like mundane/precious or
natural/synthetic are challenged and replaced with a more
differentiated and collaborative approach.
The play on contradictions and the productive tensions it
creates are crucial features of hybridity in general and extend
to the modeling and rendering properties of the chunk. By
using intuitively malleable polygon meshes, typical in charac-
ter animation, in conjunction with the precision of spline sur-
faces, the models combine the seamless “sculpted” look of the
former with the engineering-like qualities of the latter. Thus,
both organic and constructed details can be articulated to the
same fine level, making them immediately present and leading
to a foregrounding of all parts on the drawing surface. This
“flattening” effect is further amplified by avoiding perspectival
falloffs, in which parts become less legible, and by contrasting
the object’s photo-realistic style with the uniform neutrality
of the image background. The overall model appears both real
Angela Huang and Phoebe Leung, and abstract at the same time, while each of its individual com-
Ventilation Tower, New York, 2016. ponents is perceived in a flat hierarchy that promotes an archi-
Chunk model rendering depicting a
new design for the Battery Tunnel
tecture of lateral connectivity over one of layered assembly.
ventilation tower in Manhattan. Ultimately, how we assess the value of a model is contin-
Existing mechanical systems are gent not only on its various architectural functions but also on
utilized and made accessible to the
its ability to draw public interest, which is particularly true in
public to create a new form of vertical
machinic garden. This student proj- our fleeting social media age. The compactly sculpted chunk
ect was designed in the Respiratory model delivers just the right kind of instantaneous visual
Object design studio instructed
impact to thrive in the realm of shareable images. And here,
by Ferda Kolatan with Michael
Zimmerman, Stuart Weitzman School once more, the rocaille is an intriguing precedent. By keenly
of Design, University of Pennsylvania. deploying the technologies of reproduction of its time – such
Drawing courtesy the author. as etching, engraving, and printing – the rocaille became
widely distributed across 18th-century Europe. From these
templates, local artisans produced new variations, ensuring
the further proliferation and refinement of rococo culture. It
is conceivable, then, that in the progressive interplay between
the visual and physical features of a model, and in its capacity
to become digitally communicable and culturally participa-
tory, there lies the real potential of the chunk model.

Ferda Kolatan is associate profes-

sor at the University of Pennsylvania
Weitzman School of Design and a
founding director of SU11 Architecture
+ Design in New York.

74
Paulette Singley
Dollhouses and
Other Bad Objects
Within the carefully policed precincts of the academy, some criti-
cal objects are promoted to the status of good objects (say, not so
long ago, dead authors), while others are tabooed (say, in the old
days, experience).
– Naomi Schor, Bad Objects: Essays Popular and Unpopular, 1995

Just what is it about the dollhouse, that miniature make-

believe object, that taunts and teases architectural propriety?
That prohibits its use in architecture and transforms it into
the discipline’s bad object? Even today, when this threshold
has been crossed by both artists and architects, the cautionary
words of a former architecture professor of mine still echo:
“Don’t make your models look like dollhouses.” In other
words, the dollhouse has been architecture’s bad object.
For literary critic and theorist Naomi Schor, a bad object
is one whose institutional status has shifted due to consensus
regarding its inappropriate inclusion in a certain field of
inquiry. While from one vantage point this makes sense for
dollhouses in architecture – eschewing overwrought details
in favor of formal focus, replacing embedded narratives with
conceptual clarity, ensuring that singularity presides over
multiplicity, avoiding the overlay of gendered stereotypes
on inert neutral material, replacing a hobby with serious
research, or privileging exterior over interior design – from
another view, transgressing these prohibitions illuminates
contemporary architectural production from the compelling
and critical position of misbehaving models. Indeed, the
dollhouse-as-model has shifted back to the status of a “good
object,” one that the institution of architecture is absorbing
for its ability to skew spatial norms and act out alternative
domestic narratives in its compartmentalized chambers.
Seen through the model buildings of artists Sam
Durant and Mike Kelley and the toylike architectural mod-
els of Jennifer Bonner of MALL and Jimenez Lai of Bureau
Spectacular, model misbehavior describes the way in which
the loophole of playacting pulls at the loose threads of mis-
chief that the dollhouse’s ludic chambers sanction. The
emergence of what I call the hyperdollhouse conjures realms

75
 where melodrama, or bad acting, finds its proper equivalent
in interior design as the formal generator for this architec-
ture, where sites of microdisobediences in furnished tableaux
upend architectural biases toward normative familial narra-
tives. Focusing on the dollhouse as a philosophical toy – its
excessive details, unfettered ornamental adherences, province
of the “petite feminine,” and domestic saturation – reveals it
to be a model site for exploring miniature architecture, where
behavioral rehearsals and reversals are child’s play.

Abandoned Houses
In his suite of six Abandoned Houses (1994), Sam Durant des-
ecrates the convention of clean architectural models with
shoddy construction techniques, postconstruction despolia-
tion, and miscreant traces of inhabitation in rough maquettes
of the midcentury Case Study Houses in Los Angeles. From
1945 to 1966, the Case Study House program, sponsored by
Arts & Architecture magazine, commissioned significant prac-
titioners of the time to design and build inexpensive, efficient
experimental model homes. This program began as affordable
housing but ultimately produced high-priced collector’s items
built in affluent areas such as Malibu, Beverly Hills, Pacific
Palisades, and Hollywood Hills.
Durant’s model of Pierre Koenig’s iconic Case Study
House #22, the Stahl House (1960), is built with foam core
– a material generally not considered museum quality
– plexiglass, and cardboard and rests on a spindly tripod of
wooden dowels. In Abandoned House #4, a replica of Richard
Neutra’s 1948 Case Study House #20, Durant knocks holes in
the walls and roof, shoots out the plexiglass floor-to-ceiling
windows, covers it with graffiti, and litters it with debris in a
1.  “Based in L.A., Durant might be seen as kind of post–Helter Skelter aesthetic.1 Another Case Study
a second-generation ‘Helter Skelter’ artist.
That was the title of a hugely influential
House glows from within, illuminated by a blue light emanat-
1992 MOCA survey, which catapulted ing from a miniature television set that plays soap operas and
a dozen of the city’s painters, sculptors
and installation artists to the forefront of hyped-up talk shows like Jerry Springer and Rush Limbaugh.
international consciousness.” Christopher As writer and critic Jeffrey Kastner concludes, “The discon-
Knight, “In a distinctly American grove,”
Los Angeles Times, November 2, 2002, nection between the high-minded ideals of the architects and
https://www.latimes.com/archives/la- Durant’s apocalyptic vision provides a metaphor for the
xpm-2002-nov-02-et-knight2-story.html.
2.  Jeffrey Kastner, “Sam Durant,” Frieze, inevitable clash between all such noble certainties – whether
June 8, 1995, https://frieze.com/article/ social, political or aesthetic – and the accelerating vagaries of
sam-durant.
modern life.”2 That Durant has inhabited these models with
furniture and ruined them with vandalism transforms them
into hyperdollhouses that operate as bad objects.
Durant’s critique of the utopian failures of modern domes-
tic architecture in Los Angeles and beyond undermines the

76 Log 50
Sam Durant, Abandoned House #4,
1995. Foam core, cardboard, plexi-
glass, tape, spray enamel, wood, and
metal, 25 1/2 by 41 by 4 1/2 inches.
Photo: Brian Forrest. © Sam Durant.
Courtesy the artist and Blum & Poe,
Los Angeles/New York/Tokyo.

status of these houses as hygienic examples of modernism’s

belief in architecture to transform human behavior. His claim
is that human behavior will inevitably transform the archi-
tecture, unearthing, Kastner writes, latent domestic dramas
of “the American ‘home’ as a site of economic, sexual and
3.  Ibid. spiritual deviance” in the process.3 According to art critic David
4.  David Joselit, “Sam Durant at Museum
of Contemporary Art, Los Angeles,”
Joselit, “Durant’s art pivots on the double meaning – physical
Artforum 41, no. 5 (January 2003): 132. and conceptual – of modeling or remodeling,” particularly the
Emphasis original.
process of “‘retrospective possession’ (à la Invasion of the Body
Snatchers),” which can be seen in the way Durant inhabits the
architectural model as if it were a host to be reprogrammed
by an alien tenant.4 Ultimately the inhabited, versus empty,
architectural model leads to the insertion of messy interiors
that besmirch modernism’s abstract purity. Durant’s miniature
television sets, broken walls, caved-in ceilings, and bullet holes
are the kinds of details necessary to complete the domestic
rupture these architectural models script.
While Durant’s series makes evident the latent class
distinctions and persistent stylistic tyranny that describe the
Stahl House and numerous other midcentury residences in
greater Los Angeles, it also allows us to explore the architec-
tural model as a site of sacrilege, if not the sacrifice of modern
architecture’s hegemony. We gaze into these spaces as if
detectives inspecting the scene of a crime. To detect, writer
Jill Lepore explains, “is, etymologically, to remove the roof of

77 Log 50
 a house,” an act that in the generally slack construction of
5.  Jill Lepore, “The Prism: Privacy in these models appears to be inherent.5 Durant introduces and
an age of publicity,” New Yorker, June
24, 2013, https://www.newyorker.com/
establishes the model’s potential to depict the abject space of
magazine/2013/06/24/the-prism. trauma, in which the traces of human inhabitation override
6.  Sam Durant, “The Manifest Destiny of
Borrowed Scenery,” interview by Andrew
the architecture’s formal relevance. His models catalyze the
Maerkle, part 1, ART iT, September 22, idea of hyperdollhouses, insofar as they are inhabited by
2017, https://www.art-it.asia/en/u/
admin_ed_itv_e/plkue6rgtxgn8sqtlufs. microdetails that materialize the scuff marks, smears, and
7.  See Mike Kelley, “ Mike Kelley” (lecture, streaks of real life. “It was right before midcentury modern
SCI-Arc, Los Angeles, December 6, 1995),
available at SCI-Arc Media Archive, started to gain a lot of traction in the mass media, and the
https://channel.sciarc.edu/browse/ houses were being ‘remodeled’ as they were sold to new
mike-kelley-december-6-1995.
8.  “Mike Kelley, Educational Complex, owners or passed to the next generation,” Durant recalls.
1995,” Whitney Museum of American “Owners were putting on colonnaded porches or Spanish
Art, https://whitney.org/collection/
works/10293. Revival additions, trying to make the houses look anything
9.  Anthony Vidler, Warped Space: Art, but modernist.”6 These remodels, he explains, operate as
Architecture, and Anxiety in Modern Culture
(Cambridge: MIT Press, 2000), 166. “forms of critique – some of it class based, some psychologi-
cal – of the original designs, which were to some degree
forms of social engineering.” That is, engineered behaviors.

Bad Acting
Where “model behavior” suggests the desire for paradig-
matic codes of conduct, apprehending instead the power of
“model misbehavior,” or prescribed disobedience, becomes
both a challenge and an opportunity for design practices. In
their simulated naughtiness, “bad” models, such as those
Durant produces, offer an alternative path for the produc-
tion and representation of architecture’s three-dimensional
explorations.
“Bad acting” is a generative concept that artist Mike
Kelley applied to painting, but one that architecture might
consider in the production of physical models.7 In melo-
drama and soap opera, bad acting is the dramatic risk of
sanctioned misbehavior, the chance to overact, overreact,
and thus subvert theatrical conventions and narrative
apparatuses without actually breaking any rules, all while
drawing attention to normative, repressive familial struc-
tures. Kelley explored the model as an avenue for sociospatial
critique of education as melodrama and as a disciplinary
apparatus. In Educational Complex (1995), a roughly 16-by-
8-foot tabletop installation made of painted foam core,
fiberglass, plywood, wood, and plexiglass, with a mattress
underneath, Kelley displays scale models of every school he
attended and the house he grew up in. According to Kelley, it
“was done directly in response to the rising infatuation of the
public with issues of Repressed Memory Syndrome and child
abuse . . . and the popularization of . . . therapy, which was

78 Log 50
Mike Kelley, Educational Complex, predicated on the idea that certain traumatic events . . . are
1995. Acrylic, latex, foam core, fiber- repressed and only removed later through therapy.”8
glass, wood, 57 3/4 by 192 3/16 by
96 1/8 inches. © The Mike Kelley
While the stark white buildings and lack of accessorizing
Foundation for the Arts, 2021. All details position this project in line with the clean, museum-
rights reserved / licensed by VAGA at quality depiction of “good” models, a seeming malevolence
Artists Rights Society (ARS), New York.
infuses Educational Complex with the dank odor of locked
mechanical rooms and basement closets. Kelley transforms the
white model from a mode of abstraction into one that presents
a clinical diagnosis of childhood repression through a cluster
of buildings that constitute a frightening meta-agglomeration
of where we go or have gone to school. A cold, unfeeling,
dystopian campus infiltrates childhood memories of school
while simultaneously bending the language and forms of
modern architecture’s social mission back onto itself. Evoking
the unheimlich or uncanny as part of the bad acting model,
historian Anthony Vidler sees in Educational Complex a “nostal-
gia for the ‘homely,’” an architecture that forces repressed
educational trauma to surge upward as one lies on the mattress
provided beneath the model to view the sublevel, which
represents the basement of the California Institute of the Arts
building where Kelley attended graduate school.9
Kelley’s work also enters the genre of melodrama, syn-
onyms for which may be located in bathos, dramatization,
exaggeration, maudlinism, nostalgia, camp, and sentimen-
talism. Melodrama operates in the realm of excessive, almost

79 Log 50
 humorous, theatricality while delivering a set of specific pre-
scriptions that insert a sociopolitical critique into a generally
quotidian plot. The filmmaker Rainer Werner Fassbinder,
as writer Martin Hargreaves observes, stresses melodrama’s
political dimension, “denaturalizing how desire flows within
the frame, delighting in how the scenery flirts with us as much
10.  Martin Hargreaves, “Than Hussein as the protagonists.”10 Similarly, feminist film theorist Laura
Clarks’s Politics of Melodrama,” Frieze,
April 20, 2018, https://frieze.com/article/
Mulvey concludes that “as a safety valve for ideological contra-
hussein-clarkss-politics-melodrama. dictions centered on sex and the family,” melodrama “acquires
11.  Laura Mulvey, “Notes on Sirk and
Melodrama,” in Visual and Other Pleasures
a wider aesthetic and political significance,” which alludes
(Basingstoke: Palgrave, 1989), 39. to “the workings of patriarchy, and the mold of feminine
12.  Susan Stewart, On Longing: Narratives of
the Miniature, the Gigantic, the Souvenir, the
unconscious it produces.”11 Kelley’s work offers a melodra-
Collection (Durham: Duke University Press, matic interpretive mirror in which we can see ourselves in the
1993), 61.
13.  Ibid.
reflection of exaggeration, sentimentality, and theatricality. In
14.  Nicole Cooley, “Dollhouses Weren’t parallel to this stratagem, misbehaving architectural models
Invented for Play,” Atlantic, July 22, 2016,
https://www.theatlantic.com/technology/
may seem to disinter latent desires to act out on the interior of
archive/2016/07/dollhouses-werent- a small room, as well as allow for the monsters of childhood
invented-for-play/492581/.
nightmares to explore the nooks and crannies of our subcon-
scious within the remarkable verisimilitude of scaled space.

Dollhouses
As transitional objects between childhood and adulthood,
conscious and unconscious thought, desire and repression,
or architect and architecture, dollhouses have the capacity to
project fantasy into the quotidian and vice versa. For liter-
ary critic Susan Stewart, “Transcendence and the interior-
ity of history and narrative are the dominant characteristics
of the most consummate of miniatures – the dollhouse.”12
As dollhouses contain miniature representations of interior-
ity, bookcases with real books, a dollhouse of the dollhouse,
an oven that hinges open, she writes, they become a “mate-
rialized secret” that offers the “promise of an infinitely pro-
found interiority.”13 Perhaps it is this very interiority, which
renders the child an interior designer and feminizes the act
of home decoration, privileging inside over out, that so frus-
trates normative architectural production. Or perhaps it is
the interior’s transformative value in providing space for the
contravention of societal norms that so threatens architec-
ture’s exterior formal propriety and provides the detective
with continuous employment.
Originally, dollhouses were not intended for children
or for play. Nicole Cooley writes that dollhouses, first built
in the 17th century, primarily in Germany, Holland, and
England, “were closely associated with wealth and served as
markers of social class and status.”14 The German Dockenhaus,

80 Log 50
Edwin Lutyens, Queen Mary’s Dolls’
House, 1923. Photo: Topical Press
Agency / Stringer. © Getty Images.

or “miniature house,” and the Dutch “cabinet house,” which

opened “like a china cabinet on hinges that can be closed and
locked,” were similar to a Wunderkammer. Also emerging in
the 17th century were Nuremberg kitchens, toy-sized metal
houses designed for purely utilitarian purposes that “might
15.  Ibid. contain a hearth, cooking pots, a straw broom.”15 These
16.  John Martin Robinson, Queen Mary’s
Dolls’ House: Official Guidebook (London:
were used to teach girls how to set up and control a home, to
Royal Collection Enterprises, 2002), 5. understand the rules of the oikos, and to manage both house-
hold objects and servants. A precursor of Margarete Schütte-
Lihotzky’s Frankfurt kitchen (1926), the Nuremberg kitchen
defined efficiency, the laws of the household, and the man-
agement of behaviors in a distinctly feminine realm.
One of the most exceptional examples of a dollhouse, that
of Queen Mary, wife of King George V of England, is consid-
ered “a national monument in the form of a miniature build-
ing.”16 Completed in 1924 according to a design by Edwin
Lutyens and with a hidden garden by Gertrude Jekyll, Queen
Mary’s Dolls’ House replicated the then typical English town
house at a scale of one inch to one foot. The house was fitted
with running water and a flushable commode, a library with
leather-bound manuscripts by authors such as Arthur Conan
Doyle (who wrote the short story “How Watson Learned the
Trick” for the house), a bedroom with damask wallpaper and
a horsehair mattress, a longcase clock designed by Lutyens

81 Log 50
 and made by Cartier with movements copied from the 17th-
century clockmaker Thomas Tompion, chinoiserie murals of
fairy tales by Edmund Dulac, a functional Winsor & Newton
watercolor paint box made of silver, and much more. It was
displayed in the 1924 British Empire Exhibition at Wembley,
an international trade fair, where it served as a showcase for
British workmanship and an advertisement for the products,
housed within, of Great Britain and her colonies.
Given the ever-evolving guidelines for raising children,
by the 1970s, the makers of dollhouses and toys were being
encouraged to provide gender-neutral experiences for both
boys and girls. In 1977, Carrie Carmichael, in her ground-
breaking book Non-Sexist Childraising, encouraged role
reversal between boys and girls when playing with toys as a
strategy for rethinking traditional gender roles in the domes-
17.  See Carrie Carmichael, Non-Sexist tic sphere.17 In a world before Mattel’s “gender-neutral”
Childraising (Boston: Beacon Press, 1977).
18.  “Miniature Manifesto Part 2: Stop
dolls, the dollhouse was a generally safe theater for cross-
Quoting Susan Stewart,” in The Wonder dressing and exploring alternative approaches to gender-
of Miniature Worlds, a blog by Louise
Krasniewicz, June 1, 2015, https://thewon-
specific play as well as personal sexual orientation and fantasy.
derofminiatures.home.blog/2015/06/01/ For anthropologist Louise Krasniewicz, miniatures are “not
miniature-manifesto-part-2-stop-quoting-
susan-stewart/. an escape from the real world but a way to engage, confront,
19.  Stewart, 63. question, critique, or consider it.”18 Similarly, for Stewart,
“worlds of inversion, of contamination and crudeness, are
controlled within the dollhouse by an absolute manipulation
and control of the boundaries of time and space.”19 In other
words, playing with scale figures in an architectural model
allows for experimentation with the practices of everyday

82 Log 50
Jennifer Bonner, Haus Gables, Atlanta,
2018. Photo: NAARO. Opposite page:
Jennifer Bonner, The Dollhaus, 2017.
1:12 model, Baltic Birch plywood,
wood glue, card stock, paper prints,
brass hardware, 32 by 55 by 18 inches.
Photo: Adam DeTour. Courtesy the
architect.

life, for exploring diverse patterns of gender identity, and

for modeling behavioral alternatives to the bourgeois values
the dollhouse is thought to endorse. If the dollhouse does not
come with closets per se, it provides deeply nested spaces for
hiding even the most fragile secrets.
It also allows for therapeutic exercise, the modeling of
behavior through a scale model. For child therapists, psycho-
therapist Susan Scheftel writes, a dollhouse “offers the pos-
sibility for a child to stage dramas on an acceptable play stage:
a safe and small space within which to create narratives that
somehow parallel and reveal much about their own inner and
20.  Susan Scheftel, “Welcome to the outer lives.”20 She argues that “there is little question that a
Dollhouse: The Therapeutic Power of
Miniature Worlds,” Psychology Today,
child who is acting out small dramas in a dollhouse is working
September 29, 2015, https://www. through internal issues about self and family in a productive
psychologytoday.com/us/blog/evolving-
minds/201509/welcome-the-dollhouse. and symbolic way.”21 Given the potential for play therapy as
Thanks to Dr. Steven Brawer for pointing an intrinsic component of the dollhouse, and even an extrin-
me to this essay.
21.  Ibid. sic manifestation of Kelley’s Educational Complex (a play on
the word complex), then arguably the architectural model has
the inherent potential to model behavior in anticipation of the
full-scale built work or as an exploration of spatial melodrama.

83 Log 50
Bureau Spectacular, Archi-Voltron,
2018. 1:1 cat tower or 1:50 architec-
tural model, plywood, wool, paint, for
Adidas + Dezeen. Photo: Injee Unshin.
Courtesy the architects.

Faking It
For Jennifer Bonner, dollhouses “defer to the imagination a
22.  Jennifer Bonner, “A Dollhouse with deep possibility for the reinvention of lifestyles.”22 A natu-
a Grilled Cheese & Other Ideas for
Architecture,” (lecture, SCI-Arc, Los
ral landing field for any discussion on contemporary archi-
Angeles, November 6, 2019), available at tecture’s hyperdollhouse is Bonner’s The Dollhaus, a model
SCI-Arc Media Archive, https://channel.
sciarc.edu/browse/jennifer-bonner-a- based on her Haus Gables that teases out the hidden prohibi-
dollhouse-with-grilled-cheese-other- tions and inhibitions that bad models encourage.23 Six gables
ideas-for-architecture-november-6-2019.
23.  Ibid. form a prominent silhouette on the exterior and set up the
guiding principle for organizing the interior, thus devel-
oping what Bonner calls the “Roofplan,” in contradistinc-
tion to Adolf Loos’s Raumplan and Le Corbusier’s Free Plan.
Constructed in Baltic Birch plywood, with realistic finishes
and a facade that opens up on brass hinges, The Dollhaus – or
model – operates as a hyperdollhouse, exploring a childlike
view of suburban architecture through banal and ordinary
acts of substituting faux for real materials. It embraces the
domestic nostalgia that such research invokes – the memory
of remodeling a house only to find layers of vintage wallpaper
lying in wait underneath wood-grained vinyl panels.
Bonner derives inspiration from artists William
Christenberry, who photographed and captured in miniature
sculpture the patina of aging barns in Alabama, and Beverly
Buchanan, who modeled shacks at multiple scales with bright
colors painted over wood scraps. Bonner similarly explores

84 Log 50
Bureau Spectacular, The Tower of
Twelve Stories, 2016. 1:1 section
model, steel, plywood, structural tim-
ber, paint, Coachella Valley Music and
Arts Festival, Indio, California. Photo:
Jeff Frost. Courtesy the architects.

the intricate fidelity of real-faux finishes, a practice inspired

by a tradition in the American South that embraces these
affordable substitutions for expensive materials. The model’s
interior is clad with paper card stock printed with patterns
of the actual materials to be used. She applied these finishes
as “stickers” – a construction idea that carried over into the
actual built house – and had the model photographed in “syr-
24.  Ibid.; phone conversation with Bonner, upy pink and blue lights.”24 Inverting expectations of high
May 21, 2020.
and low value as a semiotic system of material signs in both
The Dollhaus and the finished residence, Bonner covered the
surfaces in a bedroom and adjacent bathroom with a car-
toonishly drawn vinyl instead of real Italian marble; substi-
tuted a thin tile for expensive poured-in-place terrazzo in the
kitchen; and in a reversal of strategies, finished a stairwell
with a tile that looks like the less expensive oriented strand

85 Log 50
Right and opposite page: Bureau
Spectacular, Another Raumplan, 2017.
1:20 architectural model, plywood,
one-way mirror, rubber, wool, Chicago
Architecture Biennial. Photos: Injee
Unshin. Courtesy the architects.

board. While the interior faux finishes seem to indicate spatial

divisions, they do not correspond to the actual boundaries of
the physical environments. In the actual house, which Bonner
refers to as a 1:1 model, the facade is skinned with fake bricks
made of stucco and painted with the retroreflective white paint
used for the lines on highways. Where actively embracing the
dollhouse as a conceptual argument is a significant stance,
translating it into an inhabitable 1:1 “model” transcends repre-
sentation and engages in Lewis Carroll’s world of scalar shift-
ing and life-size melodrama.
The tackiness of literally tacked-on materials that
describes these ultimately elegant surfaces revels in the per-
formance value of bad acting on a stage set for melodrama.
They flip the ordinary into an extraordinary value, in a world
where finish materials slip away from rather than reinforce
the bounded spaces they are thought to define. This precise
slippage ties to Bonner’s concept of “faking it,” breaking the
boundaries of good behavior. Instead of lining up all of the
joints and surfaces with the spaces they index, she produces
a world of misleading metonymy, where the sign no longer
points to the whole it is thought to signify.

Players on a Stage
From Haus Gables to The Tower of Twelve Stories, from The
Dollhaus to Another Raumplan, Bonner and Jimenez Lai both
explore the hyperdollhouse and distort Loos’s paradigmatic
Raumplan. While Lai develops cartoon narratives as generative
forms in his work with Bureau Spectacular, the finished mod-
els also suggest the narrative quality of the dollhouse (as well
as references to John Hejduk’s later figural work). In Archi-
Voltron, Another Primitive Hut, The Tower of Twelve Stories,

86 Log 50
and Another Raumplan, Lai merges performance and play in
metaphoric theaters for melodrama. Vis-à-vis the dollhouse,
these atypical plans are frontally oriented models filled with
the accoutrements of play, now on slippery sloped surfaces.
Accommodating three concurrent scales – urban, human,
and cat – Lai’s Archi-Voltron model is simultaneously a scaled
representation of a mixed-use development of high-rise build-
ings, a grouping of five pieces of furniture that can be pulled
apart from the model, and a playhouse for a cat. In the tradi-
tion of Filippo Lorenzin and Marianna Benetti’s Gerbil Art
Gallery and Aleia Murawski and Sam Copeland’s snail diner
(replete with a miniature cheeseburger and French fries),
the cat tower evinces the hyperdollhouse model insofar as its
1:100 figures give it the sense of a giant architectural complex,
while its life-size carpets and rugs throw the entire schema
into scalar oscillation. In Another Primitive Hut, an indoor
tree house, Lai transforms human figures into scale models,
merging Marc-Antoine Laugier’s 18th-century paradigm of
the so-called primitive hut with the pilotis and sculptural
forms of Le Corbusier’s Villa Savoye to produce a pavilion
for social gathering in the former Broad Foundation space in
Santa Monica. Given the difficulty in determining whether it
is a scale model or a full-scale installation, in photographs of
this project, human inhabitants are transformed into minia-
ture dolls.
Conceived as a 1:1 model for the 2019 Coachella Valley
Music and Arts Festival, Lai’s 52-foot-tall Tower of Twelve
Stories expresses the curvilinear forms of his comic strip
speech bubbles. When aggregated vertically, these individual
forms are reconfigured as white, vacuum-molded furniture
– a piano, a refrigerator, a swimming pool, a niche – that

87 Log 50
 has been pulled out of the dollhouse, enlarged, stacked, and
displaced to the California desert. When Lai and his cohorts
inhabit the spaces with chairs and carpeting, the effect of the
primitive hut project reoccurs and humans are transformed
into dolls, such is the plastic quality of the installation. For
Another Raumplan, Lai and Joanna Grant reassemble Loos’s
Villa Müller as a series of dioramas and cover it in fur refer-
encing the white angora carpet in Lina Loos’s bedroom (and
a nod to Lai’s work using fur and carpets as off-scale material
in a scale model). Learning from Beatriz Colomina’s essay
“The Split Wall: Domestic Voyeurism,” in Sexuality and Space,
they explore the spectator-spectacle relationship by stag-
ing a series of dioramas, with the living room performing as
the dominant character, allowing visitors to control whether
spaces corresponding to other characters – the kitchen, cor-
ridor, and bedrooms – are visible from the opposite side of
the model.
The playfulness and theatricality of Lai’s work betray
an analytical position toward inhabitation, indeed a kind of
behavioral anxiety, through which curved contours, spectato-
rial interaction, and scripted knickknacks align to encourage
alternative modes of living. Summoning a close ally of melo-
dramatic architecture, Lai cites Morris Lapidus as inspiration
for his work: “If you create the stage setting and it’s grand,
25.  Jimenez Lai, “Looking to Introduce everyone who enters will play their part.”25 That is, they will
Something Inconvenient,” interview by
Stewart Hicks, MAS Context 32: Character
behave accordingly. Lapidus’s fancy hotel lobbies featured
(2020): 184; and phone conversation, May staircases, as Lai observes, “where the sectional relationship
21, 2020.
26.  Ibid.
would give someone a higher ground, someone else a lower
27.  “Bureau Spectacular, insideout- ground, and there was bound to be a lovestruck moment
sidebetweenbeyond,” SFMOMA,
https://www.sfmoma.org/exhibition/
with that kind of sectional relationship.”26 In his oscillation
bureau-spectacular/. between irony and sincerity, Lai sees himself as a social engi-
28.  Umberto Eco, Travels in Hyperreality:
Essays, trans. William Weaver (London:
neer who deploys architectural form to transform every-
Pan; Secker & Warburg, 1987), 10. day people into players on a stage that subverts typical plans
29.  Ibid., 56. Emphasis mine.
in order to foster atypical performances. Just as his models
misbehave, Lai “views architecture as a medium capable of
rewriting cultural narratives” and seeks to make people ques-
tion their own behavior patterns.27
In his 1986 book Travels in Hyperreality, Umberto Eco
imagines what Superman’s Fortress of Solitude contained. Like
a Wunderkammer, it holds wax representations of events that
involved him, including a miniature city of Kandor, preserved
under a glass bell. Coincidentally, Kelley depicts Kandor –
which the villain Braniac shrank to toy-size so that Superman
had to protect it and its live inhabitants in a jar – in a series of
colorful acrylic sculptures kept under glass. Eco understands

88 Log 50
Mike Kelley, “Kandors 1999–2011,” that for Americans, “the ‘completely real’ becomes identified
Hauser & Wirth, Los Angeles, October with the ‘completely fake.’” He refers to the “‘crèche-ifica-
21, 2017–January 21, 2018. Photo:
Fredrik Nilsen. Courtesy Hauser &
tion’ of the bourgeois universe,” alluding to those hyperreal
Wirth. © The Mike Kelley Foundation miniatures or presepi, the nativity scenes from Naples with
for the Arts, 2021. All rights reserved minutely detailed hyperreal spaces.28 Here, Eco’s description
/ licensed by VAGA at Artists Rights
of hyperreality is something akin to the hypermodel.
Society (ARS), New York.
Once taboo in architectural production, dollhouses today
straddle the margin that separates good and bad objects.
They offer both the milieu and the space for the performance
of melodrama, or bad acting, often in a decidedly feminine
modality. Such acting out may be a form of therapeutic
research or a way of loosening the social patterns and mores
surrounding inhabitation. Indeed, its marginal status is pre-
cisely what makes the dollhouse a powerful tool, one whose
secret offers the promise of an “infinitely profound interior-
ity.” In an observation that invokes Bonner’s design strategy,
Eco writes, “The industry of the Absolute Fake gives a sem-
blance of truth to the myth of immortality through the play
of imitations and copies, and it achieves the presence of the
Paulette Singley is an architec-
divine in the presence of the natural,” and where the natural
tural historian whose work expands
the disciplinary limits of architec- is “cultivated.”29 As a “Fortress of Solitude,” the hyperreal
ture across diverse subject mat- dollhouse is a model that misbehaves as, and offers sanctuary
ter such as food, film, and fashion.
for, bad objects.
Her most recent book is How to
Read Architecture: An Introduction to
Interpreting the Built Environment.

89
Observations on the Chicago Model

The Chicago Model is the second mainly procured on eBay, which Office Kovacs, Proposal for Collective
Living II (Homage to Sir John Soane),
in a series of three “proposals for he repurposes for the realiza-
2017, presented at the Chicago
collective living” by Los Angeles tion of his hypermaquettes. Just Architecture Biennial. Image courtesy
designer and educator Andrew as Soane’s designs were inspired the architect.
Kovacs. The model is an homage by his collection, the images and Clearly, the making of the
to John Soane, the compulsive objects hoarded by Kovacs become Chicago Model depended on a
British collector of paintings, the fragments of his 2D and 3D condition of overwhelming pres-
sculptures, architectural frag- assemblages. The primary differ- ence, immediate accessibility,
ments, and models of buildings ence is that Soane arranged his and the flattening of commodi-
and ruins, which he exhibited in references according to an order- ties in cybercapitalism. In addi-
the theatrical space of his house ing principle, while Kovacs’s tion, the way in which the model
museum. Kovacs, too, is a collec- messy compositions show no engages the observer, forcing the
tor. In 2010, he began curat- mediation but rather are the eye to wander restlessly among its
ing an archive of architectural merz of a postdigital bricoleur. disparate elements, simulates the
images, ranging from plans to In this sense, the Chicago Model state of distraction that charac-
product advertising, which he is not a faithful rendition of a terizes the experience of the web
scans from books and magazines proposed structure; exchanging and social networks. The Chicago
and then publishes in the digi- or rearranging its elements would Model is thus not only a natural
tal space of his blog, Archive not alter its purpose: to explore byproduct but also a plastic illus-
of Affinities. He also col- ideas about density and diversity tration of the technological envi-
lects models, miniature figures, in urban space in the hope that ronment in which we live today.
plastic toys, and other gizmos, one day these ideas will be built. – Davide Tommaso Ferrando

90
Tom Wiscombe
& Marrikka Trotter The Inner
Life of Models
Bruce Nauman described his Model for Trench and Four Buried
Passages (1977) as a concept model for a subterranean space at
an unknown, much larger scale. When confronted with this
piece, the viewer is forced to speculate what it might be like to
encounter it on a vast, almost planetary scale and what its pur-
pose might be. Is it a supercollider? A mausoleum? Originally
shown in Nauman’s studio, the sculpture raised the surreal
possibility that everything in the room could also be a model
1.  See Ian Wallace and Russell Keziere, of some other giant thing.1 Taken as a precedent, Nauman’s
“In Conversation with Bruce Nauman,”
Vanguard 8, no. 1 (February 1979): 16.
piece illustrates the power of scale, which, more than anything
else, tells us what a particular construct is for. Architecture
often defaults to mirroring the medium scale of human expe-
rience, crystallizing our routine sense of our position as pri-
mary and special. Upending our expectations of scale expands
our vision of reality to include things that exist equally but
at radically different registers, like galaxies, mountains, and
coronaviruses. Like Nauman’s sculpture, architecture today
should embrace the potential of the miniature and the gigantic
to create doubt and wonder in everyday reality.
In buildings, we always associate the scale of things with
the number of pieces with which they appear to be made. One
of the reasons the CCTV Headquarters by OMA seems charm-
ing and toylike is that the overscaled channel pattern on its
envelope overpowers the regularizing grid of the glass cur-
tain wall, making it stand out against the surrounding, con-
ventionally clad buildings. While one might associate these
channel patterns with structure hidden behind them, their
true function is aesthetic. The result is a building that appears
smaller than it is, which makes the entire city seem out of
scale by comparison. Or consider Frank Lloyd Wright’s Ennis
House, which is subdivided into small, intricate textile blocks
that make it seem larger than it actually is, as if it were a
Mesoamerican temple rather than a private residence. This
effect is emphasized by the blocks themselves, which run from
the inhabitable building above to its retaining wall below
without a break, making it difficult to gauge the true extent of
the interior. In the world of special effects, a related technique
known as “greebling” involves the almost maniacal addition

91
Bruce Nauman, Model for Trench and of fine surface detail and subdivisions to small models so that
Four Buried Passages, 1977. Plaster, they believably resemble large technical objects like spacecraft.
fiverglass, and wire, 65 by 360 by
192 inches. © 2020 Bruce Nauman
When medieval architects began designing the reliquaries,
/ Artists Rights Society (ARS), New baldachins, and tabernacles that previously had been the pur-
York. Installation view of the exhibition view of other guilds, these structures quickly became vehicles
“Bruce Nauman: Disappearing Acts,”
for mixed-scale and mixed-material speculations. Tiny vaults
Museum of Modern Art, New York City,
October 21, 2018–February 25, 2019. and oriels were combined with out-of-scale jewels; miniature
© The Museum of Modern Art/Licensed castles were topped with crystal turrets; structures were clad
by SCALA / Art Resource, New York.
in too-big walrus plaques and crowned with oversize golden
eagles. Freed from the structural logistics and construction
techniques that limited larger works, these microarchitec-
tures, as art historian François Bucher calls them, ceased to
function as replicas of full-scale architecture and instead
became lively collections of multiscalar objects in the general
2.  See François Bucher, “Micro- form of a known type.2 Composite entities such as these are
Architecture as the ‘Idea’ of Gothic Theory
and Style,” Gesta 15, no. 1/2: Essays in
never resolved in a unified object, as in classical theories of
Honor of Summer McKnight Crosby part to whole, but remain bundles of parts that continually
(1976): 71–89.
catch attention, one by one.
Another example of the power of scale is the diorama,
which is a quasi-architectural form of representation. The
term diorama originally referred to a device for peering into
another world: a viewing box containing immersive, double-
sided paintings that appeared to change before the observer’s
eyes through the use of lights and mirrors. This portal-
like quality continues today in historical and architectural

92 Log 50
Thomas Demand, Kontrollraum / dioramas, which are miniature, bounded scenes with a wealth
Control Room, 2011. C-print mounted of realistic detail. Unlike many other model formats, dioramas
on Diasec. © Thomas Demand / Artists
Rights Society (ARS), New York. Photo
typically include some thickness of the ground, as if they
courtesy Matthew Marks Gallery. were cored out of the earth, turning the ground into an object
in the scene alongside other objects like mountains, buildings,
lakes, and animals. The delineation of strict boundaries
focuses our attention on specific objects or depictions of events
rather than presenting the world as a continuum. Dioramas
enchant us because they often capture very large phenomena,
like weather and war, as specific collections of transient little
things, such as snow, weapons, campfires, or horses. The real-
ism of dioramas, and the care given to each element in a spe-
cific scene, stands in contrast to abstract architectural models
that dumb down the world to scale-independent geometry. In
a similar fashion, the alluring effects of dioramas are com-
pounded in the contained format of snow globes, where alter-
nate scale relationships, aesthetics, and even laws of physics
become possible. For example, the city of Tokyo has a particu-
lar scale, unique weather patterns, and relationships to topog-
raphy, population, and infrastructure. The Tokyo captured
in a snow globe is entirely different because its relational net-
work is broken. Buildings can be differentially scaled, proxi-
mate elements can be eliminated and replaced with distant
features, and imaginary inhabitants (like kaiju) can be incor-
porated, all while the essence of the city is maintained.

93 Log 50
Revell Convair F-102 Delta Dagger
model kit box, 1958. © Revell GmBH /
QUANTUM.

Model Ontology
The cover of the box for Revell’s Convair F-102A Delta
Dagger aircraft model kit features airmen and mechanics
animatedly assembling the airplane – “With 14 Airstrip
Accessories!” – as if it were taking place at full scale. But in
the background, a large child overlooks the scene, alerting the
viewer to the fact that this activity is unfolding on his dining
room table. These two scales, existing side by side, represent a
new speculative reality full of enchanted things.
A model is never an exact replica. It is always driven by
deliberate choices about its materiality, tectonics, scale, and
level of detail. When Thomas Demand modeled the earth-
quake- and tsunami-damaged control room of the Fukushima
Daiichi nuclear power plant, his use of paper and reduction
of detail produced the sense that something was off. Demand
claims that his models are at a 1:1 scale, but there is no way
to verify this because there are no people in his work, and he
allegedly destroys his models after they are photographed. His
models thus take on a life of their own, projecting an intrigu-
ing alternate reality by creating doubt. Demand’s model
reality exists alongside the familiar factual domain of photo-
journalism’s recording of global events, making us look more
closely at our world.
The contours of a model ontology thus begin to come
into view. Model ontology is flat. It positions models along-
side the things they represent and affords them an inner life at
their respective scales. In architecture, granting models their
own ontology could have strange consequences for full-scale
objects, which might suddenly take on qualities of their small-
scale or low-res counterparts in the same way that Nauman’s
sculpture turns the gallery into a vitrine.

94 Log 50
 Architectural models have often been limited to pre-
senting proof for other things. Historians believe that the
ancient Greeks used models, or paradeigmata, as contractual
parts of construction documentation, a function they contin-
ued to serve through the end of the Middle Ages. Vitruvius
distinguished between conceptual design work, which was
inscribed on paper, and the techne, or craft, of model build-
ing. In the Renaissance, Alberti used models as demonstra-
tions of structural solidity and three-dimensional massing,
taking care to strip them of ornamentation and detail that
3.  See Massimo Scolari, Oblique Drawing: might destabilize their representational function.3 Gaudi’s
A History of Anti-Perspective, trans. Jenny
Condie Palandri (Cambridge: MIT Press,
hanging chain model of catenary arches for Sagrada Família,
2012), 137–38. reduced to scaleless diagrams of forces, are one of the best
examples of how models have been instrumentalized to prove
real-world efficiencies. This is consistent with the broader
architectural tendency to confine models to representation,
validation, and iteration, undermining them as objects in
their own right. The particular aesthetic value of models –
which derives from their alien smallness and inconsistent
level of detail, as well as their aberrant subdivisions and
methods of assembly – is also overlooked. As Bucher’s micro-
architectures and Revell’s box both show, models simultane-
ously point to an object we know (a cathedral, an airplane)
and embody an alluring parallel reality. In a time when
architecture is being asked to ground itself in the familiar
and the everyday, model ontology challenges us to question
our most basic ideas about what exists, how things exist, and
what we can know.
Models are not typically built out of the same parts as
the full-scale entities they represent; their components have
strange shapes, massive size variation, and surprising meth-
ods of assembly that are particular to model reality. Moreover,
model parts comprise sets that possess attributes different from
those of the built object, something that is particularly true
in the case of plastic model kits. These kits and their parts are
constrained by the techniques of injection molding and the
economics of mass production. Parts never appear as layers of
construction – for instance, from structure to envelope to orna-
ment – but rather as fully formed chunks with various inte-
grated features. They do not break down into obvious modular
units nor begin and end at mass inflections or pieces of material,
as we often assume architecture must do. Rather, they follow a
mysterious figuration that wraps around forms in unexpected
ways. Model kit parts are connected strangely too: by friction
fitting, snapping together, or sleeving into one another – all

95 Log 50
 valuable expansions of our traditional understanding of what
architectural parts are and how they come together.
Moreover, model kits often contain a collection of related
materials, including backdrops, bases, decals, assembly manu-
als, spare parts, and posters. Taken together with the boxes
they are displayed on, they comprise a whole world around
the model. This is a new way to think about context: not as a
fabric or continuum into which architecture must be fused,
but as a curated collection of things that stages the architec-
tural act. Rather than emerging from a site, a building might
land with its own ground or backdrop; it might arrive with
its own assembly logic that does not resemble conventional
construction documents; it might even appear to remain pro-
Lockheed SR-71 Blackbird model kit visional, as if its parts could be reconfigured or its entourage
parts. Photo courtesy Italeri. could expand.
In our office, we make plastic model kits to resist seeing
architecture as either monolithic or broken into the mil-
lion little elements of everyday construction. These kits are
speculative environments of parts that can be arranged,
rotated, scaled, duplicated, and even thrown away in a play-
ful design process. Importantly, our model kits also contain
curated elements of a building’s context, its historical prec-
edents, and its entourage, along with alternate versions of
the design, detailed chunks at other scales, and spare parts.
These are not single-purpose kits that produce a fixed out-
come. Instead, as true aficionados of model kits know, the
goal isn’t just mechanical assembly but rather lively engage-
ment. Our model kit for the Shenzhen Museum of Science
lays out giant pieces of the building – all facing forward,
capturing their strange totemic silhouettes – as well as a full
extra set of mirror-image components that heightens their
resemblance to the features of human faces: winking eyes,
smiling lips, etc. Included in the set is an extensive entourage
of the building’s imagined contents (such as Apollo rockets,
Chang’e lunar landers, Tyrannosaurus rexes, orcas, volca-
noes, and Lockheed SR-71 Blackbirds) as well as a folded
technical illustration that at first appears to be an assem-
bly diagram but is actually a construction sequencing plan
for the project at full scale. Our model kits are therefore not
reductive kits of parts but rather a flat ontology of carefully
arranged things that highlight defining features and consti-
tute a larger world. They both anticipate a particular build-
ing’s design and add to the overall menagerie of forms from
which all our work draws.

96 Log 50
Tom Wiscombe Architecture, Shenzhen
Museum of Science, Shenzhen, 2018.
© Tom Wiscombe Architecture.

Collections and Supercomponents

The special edition Lamborghini Countach LP400 model kit
by Aoshima opens to reveal its own distinct world. Multiple
subkits are nestled inside, each with its own packaging and
graphic sensibility; the kit even includes laminated cards fea-
turing related objects from entirely different car companies,
such as the Maserati Boomerang. Kits like this demonstrate
that model ontology is based on a collector’s sensibility, where
each part can be removed or regarded independently while its
membership in the set is retained. Collections enable play-
fulness and allow one to defer dealing with intricate, unified
things that have already congealed.
Before systems of knowledge were hardened into verti-
cal taxonomies in the 18th century, a more horizontal way
of making connections was the Wunderkammer, or cabinet
of curiosities. Popular in Europe in the 17th century, such
rooms housed scholarly collections of artworks, minerals,
antlers, feathers, and other natural objects, as well as antique
fragments, all arranged not by type but in moveable, aesthetic
juxtapositions based on size or color, mystical properties,
and affinities. Historian Adalgisa Lugli notes that cabinets of
curiosities were places where “heterogeneous materials con-
nected by a circular system of thought” could be encountered
individually, without an observer “having to solve or face the
4.  Adalgisa Lugli, “Inquiry as collection: problem of continuity.”4 Likewise, viewing architecture as a
The Athanasius Kircher Museum in
Rome,” Res: Anthropology and aesthetics 12
collection of parts makes each component register as some-
(Autumn 1986): 116. thing separate: something that can be pulled out, regarded,
and operated on before being returned to the group. For

97 Log 50
example, in our Dark Chalet project in Eden, Utah, a massive
central fireplace element and the building’s chunky outer shell
were designed and versioned independently before being fit
together. Its model kit contains all of the versions of the shell
and fireplace, which could be otherwise assembled or used in
other projects at different scales. In the full-scale building,
each part remains seemingly independent via deep joints, hid-
den connections, and other architectural sleights of hand.
Unlike the sculptor, who treats matter as raw and undif-
ferentiated material that can be operated on at will, the col-
lector respects existing things and incorporates them into
compositions without damaging them. Collections ask to be
operated on in “chunk mode.” Parts are always at the forefront
and can easily be removed, added, or edited. We like to approx-
imate the immediacy of physical assembly in the computer,
availing ourselves of the digital realm where objects float and
spin, as if without gravity, and remain endlessly available for
discrete transformations. Around our office, you won’t hear
discussions about pulling points or making smooth gradients,
but rather imperatives – “Rotate! Scale! Copy!,” “Throw it
down!” – that suggest the fast, analog pleasure of throwing
jacks or dice onto a table.
For us, the model kit is not a generic pile of sticks and
cardboard and glue but rather a preassembled set that already
contains intuitions and inclinations toward a project. For the
Vilnius Concert Hall kit, we began by designing and assem-
bling a collection of spires that resonated with features of the
medieval city without mimicking them directly. These were
introduced into the kit in groups of duplicates and triplicates
to create a “copy-scale” aesthetic, rather than total hetero-
geneity. Another component in the kit is a disklike container
that functions as the basic spatial enclosure. Treated almost as
a game board, the disk anticipates spires being snapped in or
moved around. The diverse contents of this container, includ-
ing helical stair objects, theater blocks, cafes, and other sub-
enclosures, are laid out in the kit alongside long panoramic
windows and a diorama-like “ground object” into which to
click parts. Also included are spired buildings from the old
city center to which our project signals.
When entire buildings can be understood as parts of a
collection, our architectural sense of what constitutes a part
is productively expanded. One of the biggest problems with
subdivided surfaces is that little pieces of material, like panels,
bricks, and sticks, always tie back to the size of the human
hand, or to an object that two humans can carry. We call

98 Log 50
Tom Wiscombe Architecture, Vilnius this the medium scale of architecture, and find it to be almost
Concert Hall, Vilnius, Lithuania, 2019. anachronistic in a time when the extremely large (like the
© Tom Wiscombe Architecture.
blue dunes of Mars) and the extremely small (like Higgs
bosons) form part of our culture and understanding of the
universe. For architecture to control scale the way models do,
we must not only suppress the automatic aesthetic of known
construction systems – what we call articulation for free – but
also invent new and alien ways to break the building into
unexpected components that introduce unexpected scales.
This is something our office developed accidentally in the pro-
cess of attempting to print very large models with very small
3D printers. Forced to break masses into parts that would fit
print beds, we found that either ending parts at mass edges or
slicing the model into quadrants damaged the reading of the
massing. Surprisingly, we discovered that creating parts with
independent figurations not only kept intact the larger mass
reading but also added a simultaneous sense of the building
5.  Conveniently, this model construction as a collection.5 We call these parts supercomponents for their
logic anticipates what is already going on
in the international fabrication industry.
strange dual quality of existing both as whole objects – with
Parts, sometimes at huge scales and fully their respective aesthetic and details – and as parts of some-
preassembled, can be produced in a global
marketplace that includes aerospace, naval, thing else.
infrastructural, and architectural fabrica- Our Sunset Spectacular project in West Hollywood is
tors, and shipped anywhere in the world.
composed of puzzle-like supercomponents that are far fewer
in number than what one might expect for a building of that
size. This visually shrinks it against the background of its
context, making it appear toylike among all of its neighbor-
ing, serious, medium-scaled buildings, as if the analog world

99 Log 50
 had been invaded by a cartoon. The effect is not unlike The
Amazing World of Gumball, in which a cast of animated char-
acters is placed in the real, everyday world. The two reali-
ties – one cartoon, one video – remain discrete and cannot
fuse even though they engage with each other. For example,
the cartoon character Gumball reaches into a living room,
grabs the “universal remote,” and then uses it in the animated
world to actually control the universe. Something at a dif-
ferent scale and resolution than the world around it has an
incredible power to unsettle the reality we think we know.
Along these lines, the Dark Chalet appears to be assem-
bled out of a patchwork of large matte black and gloss black
figures. An exaggerated joint, which we call a metaseam,
separates these figures even further, as if they could be dis-
assembled at any time. Unlike Sunset Spectacular, which
is actually constructed with supercomponents, this project
only appears to be. Its true structural and material underpin-
6.  The more truthful solution of actually nings are concealed.6 As with models, where “truth to mate-
building the Dark Chalet was explored
during bidding with a Dutch submarine
rials” is forfeited for magical effects, architectural realism
builder who came in only 10 percent above has no relationship to the materialist project, with its com-
the selected bidder, revealing how close the
construction industry is to converting to mitment to honestly revealing the way things are made. As
chunk logic. While this may be convenient, with models, where glue and friction hold improbably scaled
it is neither necessary nor sufficient.
things together, architecture that withholds its techniques
of production is alluring. Even if someone has no specific
technical knowledge, they recognize when something force-
fully deviates from the norm. Consciously or subconsciously,
one constantly tries to imagine the alternate set of practices,
tools, and sequences that could have been employed. For
example, the ancient Incan wall known as Hatun Rumiyoc
in Cuzco, Peru, demonstrates an entirely alternative form of
masonry construction. Each overscaled stone is carved into
a unique figure with up to 12 angles, so that it precisely fits
or notches into its neighboring stones rather than abutting
them. It begs the question, How did they do that? Its magic
lies less in the mechanics of the thing and more in the state
of not knowing that it creates.

Godzillas and Live Modeling

The flatness of model ontology necessarily leads to a reas-
sessment of drawing. Since the Renaissance, architecture has
been distinguished from material practice by what Alberti
called disegno: abstract composition that happens on paper.
The humanist orthographic sequence of plan, section, and
elevation represents a 2D, codified mode of both producing
and understanding architecture that is separate from the

100 Log 50
 supposedly downstream work of model building. Flat clip-
ping planes create a gap between drawing and reality that is
still often considered the primary space of invention, even
though we live in a world of MRIs and lidar scans that can
comprehend an entire object without damaging it. On the
other hand, certain drawings want to be models. Our office
prefers the techniques used in cutaways and aerospace tech-
nical illustrations, which load drawings with ever more
detail and realism rather than abstracting them. Cutaways
slice along free-form lines unrelated to anatomical or con-
structional divisions to reveal the most important inner com-
ponents of a thing while retaining its most crucial exterior
features, such as its silhouette. In the case of kaiju cutaways,
for example, fictional monsters like Godzilla are shown
in action, breathing fire, even as their inner workings are
revealed. The resolution and detail of the monster’s exte-
rior skin and spiny plates are identical to the resolution and
detail of his internal organs. Technical notes identify com-
Godzilla anatomical cutaway illustra- ponents and superficial features as if these creatures were
tion by Shogo Endo. From Shoji Otomo, objects of science. Even cutaways of real objects, like those
An Anatomical Guide to Monsters, 1967.
published by the aerospace industry, are speculative; the
cutaways of an SR-71 Blackbird spy plane, for example, are
produced by technical illustrators who must deduce the inner
workings from its mysterious exterior panelization without
access to classified schematics. Deploying technical artistry
to produce an alternate reality, the cutaway drawing is only
as accurate as it needs to be. Such selective accuracy in cut-
ting and revealing is more empathetic with the object and its
inner life, giving us access to it on its own terms rather than
through the lens of the humanist tradition.
In our practice, cutaway drawings – what we call
Godzillas – are central to the design of the objects they depict.
Containing crucial aesthetic, organizational, and constructive
aspects of the building in one composition, these drawings
seem like a work space rather than a finished representation.
The Godzillas we produced for Sunset Spectacular are so
packed with technical information that they replace most
conventional construction documents while also resembling
assembly diagrams for models or toys. They allow one to
see the exterior features of the building’s three outer planes
(including their supercomponent breakdown, patchy perfo-
ration, and embedded electrical infrastructure) at the same
time and in the same resolution as one sees the internal “tes-
seract” object, differentiated by access hatches and stiffen-
ing ribs. Chunks of the building seem suspended in midair, as

101 Log 50
 if caught midrotation. This is partially an effect of a work-
ing method we call live modeling: the design of a single object
engaged and edited through many windows, at diverse scales,
and with varied modes of representation. This is less aligned
with the coordination efficiency afforded by federated BIM
models – useful but not disruptive – and more connected
with an interest in expressing the inexhaustible qualities of
objects. Live modeling allows for continuous play and roving,
as opposed to a hierarchical design process that develops
via constant, increasing refinement. A historical precedent
here might be Piranesi’s trompe l’oeil engravings in his Della
Magnificenza e d’Architettura de’ Romani, which feature open
books, notes, drawings on scrolls, and chunks of ornament
laid out as if on a table. Together, these invite the viewer into
an immersive experience of discrete architectural parts, pre-
sented as a collection.

Models and the Cosmic Perspective

Models engage us because they present something familiar at
a new scale and with different parts. Where architects often
see models as mere representations of larger, more important
things, we see opportunity in mixing model reality, with its
playfulness, chunkiness, and alien modes of assembly, into
everyday reality. Mixing scales destabilizes the entrenched
medium scale of human affairs – a welcome shift. We humans
too often assume we have exclusive control of our world, as if
we were the apex of the universe. Everything we see seems to
be “for us.” This is an illusion that must be broken if we are
to exchange anthropocentric egoism for a humbler, more cos-
mic imagination. When architects recognize the inner lives of
models, with their elastic relations to the everyday real, they
access architecture’s potential to be simultaneously part of
our world and a challenge to it.

Tom Wiscombe is the principal of Tom

Wiscombe Architecture in Los Angeles
and chair of the undergraduate pro-
gram at SCI-Arc. Marrikka Trotter
runs the history and theory depart-
ment at SCI-Arc.

102
Jesse Reiser
Voodoo

In 1963, an F-101B Voodoo model kit flew into my hands, and

it forever transported me into the strange world of model
making. The reporter Elizabeth Sverbeyeff was at our newly
renovated house to interview my mother for the New York
Times Magazine, and she presented five-year-old me with the
1.  See George O’Brien, “New Light for an kit as a gift and means of distraction during the photo shoot.1
Old House,” New York Times Magazine,
January 12, 1964.
Distracting it was indeed, and when the Times returned to
2.  See Barbara Plumb, “And a Corner our house three years later to document my mother’s second-
for Mother,” New York Times Magazine,
October 16, 1966.
floor renovation, they found (and photographed) a small fleet
3.  See Edward Eigen, “Consequences,” of dogfighting aircraft hanging from the beams in my room.2
Log 46 (Summer 2019): 45–50.
4.  Grand Prix, The Blue Max, and The Battle
With my aircraft flying in the foreground and my mother’s
of Britain occurred at the high point of the architecture in the background, the photograph presciently
counterculture and were highly influential
to a generation that was not old enough to
captures the confluence of my then newfound hobby and later
participate in it. These 1960s films of heroic professional career, documented in a manner so clear and
striving, produced by an older generation,
were highly influential to me, as they were vivid that it could only be available to hindsight.3
to another boy my age, Peter Jackson. That my first model airplane was brought to me, and not
Jackson would later use his fortune to fund
his own model-making company, Wingnut I to it, evidences something all too well-known to those of my
Wings, and other, less lucrative films generation: my interest in models is not unique. Typical of
and enterprises around both world wars,
including the creation of a replica World the boomer generation, whose parents had lived (and fought
War I air force in New Zealand. and died) during World War II, and resulting from Cold War
anxiety and the then nascent commercialization of air travel,
model airplanes gained immense popularity in the 1960s, as
my contemporaries can attest.4 Interestingly, my (our) fasci-
nation with these models has not diminished as I (we) have
grown older, but rather entered a golden age. Serendipitously,
model kits today have reached a high point in terms of accu-
racy precisely when many from my generation have dispos-
able income to purchase now very expensive kits. No doubt
the model companies suspect all too well, as do I, that this
moment will not last; the demand for models will plateau as
my generation disappears. I, for one, continue my practice
nearly every evening, relishing the foundational role these
models have played in my own perspective toward the world.
For example, when eight-year-old me constructed a
Supermarine Mark 1A Spitfire, I thought I had understood
the plane completely through the intricacies of the model.
Like artistic anatomy, building the Spitfire was an exercise in
close attention. The precise form, contours, and transitions

103
F.E.2b cockpit, 1:72 scale. Since there of the Spitfire would be forever held in my memory and
are no existing model kits at this could be summoned at will from any aspect. Naturally, when
scale, this model was scratch built
from plastic and metal sheets, tubes,
I finally saw the actual plane at the Museum of Science and
and wires, based on photographs and Industry in Chicago, it was to my shock and dismay that it
technical drawings. Photo courtesy the did not conform to the vision I had constructed in my mind.
author. Right: F.E.2b cockpit, 1:1 scale.
The ripples, dents, seams, and overall crudeness of the full-
Photo courtesy Their Flying Machines.
size aircraft made the whole experience seem like a sham; I
was immensely disappointed. Finally, perhaps to appease my
own cognitive dissonance, I had to invert the situation: the
museum Spitfire was just a shabby copy, a poor replica of the
plastic model I so loved.
That Spitfire model is long gone, tragically having
crashed against my bedroom wall not too long after my visit
to Chicago. Following the old adage, “if it looks right, it will
fly right,” I thought surely my model would soar. Shortly
upon takeoff, reality crashed into my room. Yet since the
wreckage appeared as just that, and not as the broken plastic
pieces I had meticulously glued together, my imagination
clearly escaped intact.
“Immersion” would be a gross understatement of my
condition; I was incontrovertibly obsessed. After a while,
even my parents became concerned about my hobby, wor-
ried that the toxic model glue I was breathing would damage
my brain. My mother and father must have been right to a
certain degree, for as the years progressed and my fleet grew
larger and more complex, I began to develop an extended

104 Log 50
The completed 1:72 scale F.E.2b
model. Without any scalar references,
the model takes on the strange qual-
ity of an original, absent any defects
and conforming all too perfectly to the
mind’s eye. Photo courtesy the author.

form of attention, one that has continued to serve me well. I

can testify that years of obsessive viewing and making, years
of exercising certain muscles to the extreme, can change one’s
outlook toward the world.
In Marcel Duchamp’s To Be Looked at (from the Other
Side of the Glass) with One Eye, Close to, for Almost an Hour,
narrative and content fade away when following the work’s
instructions. With the passing seconds, minutes, and (almost)
hour, one’s world becomes relentlessly visual, one’s mind
attentive almost to the point of distraction. I must admit that
I have never spent more than five minutes with the work, but
I believe from experience (from seeing the work and making
my models) that it takes more than that to enter into the
mindset Duchamp sought to induce. Likewise, I believe that
if practiced steadily over years and decades, it can be entered
into much more quickly, like a reflex of the mind and eye,
transporting one to a state of close and deep looking. It can
become a reflex of the hand and body as well, an internalized
(dare I write) will to form and mastery, a desire to make and
remake everything in sight and at hand.
At this stage of looking and making, seemingly incon-
sequential details accumulate and latent characteristics in
the models begin to emerge. The models too begin to func-
tion well beyond the realm of representation, opening into
an expanded taxonomy of abstract, concrete, analytical, and
hyperrealistic species. Once they are no longer purely rep-
resentational, models become generative in ways previously
unforeseen, and they exercise muscles, impress upon the
mind and body ways of being in the world that can inform
other activities in completely different contexts. Beginning
with a cache of wax blocks abandoned in a garage at the
Cranbrook Academy of Art by sculptor Carl Milles, model
making has thus deeply informed our architectural practice

105 Log 50
 for over 30 years. We continue to melt down and recycle wax
models made decades prior, imparting upon each of them the
same attentiveness developed through the aircraft models.
One of the first, in 1996, was for the Kansai National Diet
Library in Japan, in which the plasticity of the wax allowed
us to expediently model and develop complex geometries in
the building’s floor slabs. Poised between the plasticity of
clay and the rigidity of cardboard or plastic, wax – which we
could carve, fold, and fuse together – allowed us to quickly
sketch out our ideas in three dimensions, free of the com-
puter model’s distortion in the perspective window. We saw
our wax modeling as an extension of John Hejduk’s reliance
on two dimensions, mobilizing it in another way and persis-
tently evaluating our models in two dimensions from every
possible aspect. When we took those wax floor-slab models
to our engineer, Ysrael Seinuk, he only had to point to the
curvature of their drooping as evidence that the floor plates
were too shallow for support. Instantly, the models went from
My bedroom as published in the New being purely representational to presenting actual structural
York Times Magazine in 1966, with my behaviors, as Seinuk pointed out. This dual functioning of the
model aircraft hanging in the fore-
ground. To include my aircraft in his
model, to geometrically represent and behaviorally present a
shot, Hans Namuth instructed the potential architecture to us, further served to reinforce our
camera crew to cut and lower the orig- use of wax, ensuring that the behavioral consequences of our
inal fishing line. When I returned to my
geometric manipulation were never far from view.
room at the end of the shoot, I discov-
ered much to my horror that my air- So when I had viewed the remnants of my Spitfire
craft had fallen and smashed against model as the wreckage of some small-scale aircraft disaster,
the floor. This traumatic experience
it seems I was not too far off. As rigorously constructed
set the stage for my future encoun-
ters with image culture. Photo: Hans material actualizations, both the Spitfire model and the
Namuth. Courtesy the author. Kansai wax model held the capacity to present actual
behavioral consequences, once they were subjected to spe-
cific environmental conditions. Scaled-down wind tunnel
tests, frequently used by structural engineers, are but one
example of this capacity. (Lest we forget that my Spitfire’s
wreckage was but a consequence of its inability to fly!) Thus
I am tempted to say that there is much that models can’t do,
but I will simply say instead that there is much that depends
on our own looking and imagining, or rather, the model’s
representational functioning.
Naturally, we learned this the hard way. The most vivid
example is our AEON project in Dubai, for which we worked
through dozens of wax models to design an undulating mul-
ticore tower. We manipulated the wax to achieve contrap-
posto, believing that if our plastic wax model could hold up,
then surely it should be possible to build in steel. Again, our
engineer could not help but highlight an obvious reality: wax

106 Log 50
is cheap and steel is expensive. Each manipulation of our
wax tower model had profound structural and budgetary
consequences when scaled up to an 84-story tower; adjusting
the wax model even a few millimeters could translate into
massive changes in how forces flowed through the building,
leading to a complete reworking of its structural design
and millions of dollars in increased costs. Cold and blunt as
always, analysis showed us the limits of our imagination, or
more simply, of the models that we had made.
A reactionary tendency, which we have faced from engi-
neers, would be to rationalize the geometry of our designs, to
strip away the subtle and irrational undulations of our wax
study models toward something more regulated and easily
defined. Though necessary and inevitable to some degree,
these processes often breed monsters, a fate too familiar to
the seasoned aircraft model makers of my generation. The
increased levels of both precision and accuracy in today’s
model kits, which employ the latest 3D software, mold making,
and printing technologies, have created molds with unbeliev-
ably fine levels of formal complexity and resolution, espe-
cially in hollow castings with undercuts. Despite this, many
of today’s model manufacturers, particularly those from Asian
countries, use those same technologies to analyze and overly
rationalize the geometry of antique aircraft, though their
crafted forms and subtle deviations run directly counter to
such techniques. While these manufacturers remain oblivious
to the consequences, the models produced from such proce-
dures are often laughably unfaithful to the original aircraft, at
least to the discerning eyes of their makers.
Here we can consider a little-discussed aftereffect of
Western colonialism and its Eastern variety, namely the
proliferation and persistence of polytechnic universities in
Asia and the self-evident effectiveness of that mindset in the
realm of economic production and national security. Though
effective in these regards, the technocratic thinking concur-
rent with this training undoubtedly fails as a means of aes-
thetic evaluation. The linearity, cleanliness, rationality, and
explicability of procedures are often blind to the monsters
it produces. For example, not unlike the notoriously messy
geometry of Hasegawa’s Lancaster model cockpit canopy
(far removed from the crafted coachwork production of the
original), Skidmore, Owings & Merrill’s Time Warner Center
emerged with horror from a defensible “reason for every
form” based on a mechanical understanding of context, yet
with little care for the consequences that can only be felt. The

107 Log 50
Reiser + Umemoto RUR Architecture,
Water Garden, Ohio, 1997. Laminated
object model coated with sintered
aluminum. We sought to extend the
formal specificity of our study models
to the built work itself. Vegetation and
water are perturbed and excited by a
grooved, laminar concrete slab below,
producing a garden above that is pro-
digious and only partially manageable.
Photo courtesy the author.

errors in these instances reveal the failures of a technocratic

predisposition, of how rational yet erroneous thinking can
telegraph and snowball toward unimaginable monstrosity.
Consider this against the distorted, inaccurate, yet profoundly
real aircraft model kits of Italian manufacturers, who nev-
ertheless capture and manipulate less the geometry than the
true sense of their source material. In other words, they get it.
Thus we fear not the sleep of reason but rather its conquest,
and we continue to push forward our crafted wax models in
all their irrationality and feeling.
As we see it, irrationality is relative to time and circum-
stance, and it seems now that the relationship between the
physical model, physical making, and building construction
has fundamentally changed. Through a mix of computation
and new fabrication technologies, craft has migrated from the
field and toward architects and their models. Concurrently,
craft can be amplified and weaponized toward the architect’s
will. The traditional battle between the wills of an architect,
materials, and a community is played out at full scale.
Accordingly, our models hold a comfortably central
position in our office, presenting us with actual behaviors
while allowing us to project our own motivations into them.
Like actors on a stage, they occupy a reality that is both real
and imagined, contingent on yet independent of our look-
ing, made both by and for us yet indifferent to the moods and
desires that we nevertheless read out of them. Even when
performances end and actors leave their roles, our percep-
tions of them remain colored by these controlled and staged
encounters. Just as the Spitfire in the Museum of Science and
Industry remains for me a shabby, scaled-up model, even
our constructed work presents to us as built (though not

108 Log 50
Reiser + Umemoto RUR Architecture,
Adidas World of Sports competition
entry, Herzogenaurach, Germany,
2014. Study model, 3D-printed plastic.
Photo courtesy the author.

shabby!) models, like those of paper and wax that continue to

multiply and evolve in our studio.
The tendency to see the working model in the constructed
building is unique neither to our work nor to models. As the
elevation drawing is both presented and inscribed on the
facade of Alberti’s Palazzo Rucellai, the instruments of design
have forever haunted, and will continue to haunt, architec-
ture in ways both explicit and latent. Yet architecture is not
reducible to the drawings and models from which it develops,
and to assume otherwise is only to reenact an oft-cited error.
At some point, architecture’s inevitable material specificity
must either be engaged or disregarded to the benefit or detri-
ment of design.
In hindsight, for instance, Toyo Ito’s competition model
for the Sendai Mediatheque was so hauntingly beautiful as
to foreshadow the disappointment of the final, constructed
building. Appearing in the competition model as a translu-
cent, floating apparition, like kelp gently swaying against
deep ocean currents, the final building retains little of that
aesthetic promise. (It seems to me that the floor slabs and
members could have been much thinner if the kelp-like
cores worked in tension and hung off a superstructure, as in
our Kansai library proposal.) At the same time, can anyone
really admit surprise that the ghostliness of the model, made
of acrylic sheets, did not translate so well to steel and con-
crete? If it is too painful to admit, as it was for eight-year-old
me when first facing the crudeness of the full-scale Spitfire,
then one can likewise reverse the situation for some sort of
consolation: the building is but one manifestation of a proj-
ect that found its most influence and promise in the model.
The project does not die with the completion of the building

109 Log 50
Reiser + Umemoto RUR Architecture, but rather extends through and beyond it, finding life and
Taipei Pop Music Center, Taipei, continuing to evolve elsewhere through the work of other
2010. Study models (2003–19) exhib-
ited in “Building Beyond Place: RUR
designers. Yogi Berra said it best: “It ain’t over till it’s over.”
Engages Taiwan’s Architectural Of course, those “other designers” include ourselves, and
Cosmopolitanism,” TAAC Tribeca / Ito’s mediatheque blew us away. We saw it as a game-chang-
E. Tay Gallery, New York City, 2019.
ing project, reading in it the seeds of Vladimir Shukhov’s
Materials vary (microcrystalline wax,
paper, brass, plastic). Photo: Injinash Radio Tower taken to some kind of fantastical extreme.
Unshin. Courtesy the author. The irony, at least when considering the shortcomings of
the completed building, is that it was perhaps too much like
Shukhov’s tower; the literal transcription of his original
structural diagram for the mediatheque’s cores birthed the
leaden, seemingly oversize structure that forsakes Ito’s origi-
nal model. As poetry stands opposed to literal translation,
here it seems a similar error was committed. If Ito’s project
was to reside more in the building and less in the model, then
the realities of full-scale building clearly entered the design
process far too late. To this day, my colleagues and I tend to
reference the model when discussing the mediatheque, disre-
garding the constructed building as some strange aberration
from an astounding original vision, like The Godfather, Part
III, or any other kind of turgid sequel.
Not unlike the better Godfather films, what stands out
about Ito’s model is how it transforms the space around it,
how the action and qualities inside the building envelope recast
the site as similarly being underwater, as if all of Sendai were
subsumed in some kind of liquid medium. One need only look

110 Log 50
Reiser + Umemoto RUR Architecture,
AEON, Dubai, 2006. Study models.
Materials vary (microcrystalline wax,
chipboard, wire, wood, 3D-printed
plastic). Photo courtesy the author.

at the rendering of the trees (no leaves, of course), the cars,

and the groundscaping in the model to see how they are all
are participants in the building’s aesthetic vision. Although an
ambition of all architecture, I’ll digress to say that this is spe-
cific to techniques of model making and particular to the scale
of the model. When painting aircraft models, for instance,
using the actual colors from the original, full-scale source
material makes the models appear like toys. Interestingly, one
has to paint “atmosphere” onto the plane models to make them
appear real, since aircraft in flight are often viewed from the
ground, thousands of feet away; thus the colors appear faded
and with a slightly sky-bluish hue. The results speak for them-
selves, and when hanging from the ceiling, these models truly
appear as though they are in flight. It’s a skill that errs on the
side of propaganda, and I assume such techniques were not lost
on Ito or the model makers he employed.
We use these techniques as well. Toward the front of
our office is displayed a series of clean, painstakingly pro-
duced presentation models for competitions, which are often
expensive and primarily meant to represent the project to
juries and clients in hyperrealistic detail. I would refer to this
genus of models as representational. On the other hand, scat-
tered throughout the office is a plethora of cheap wax and
paper models, often hastily yet attentively produced during
the course of design. Although lacking in relative detail, these
species of models have their own more direct and immediate
beauty, like precious yet alien creations. I would refer to this
genus of models as presentational, since their worth primarily
resides in the particular circumstances of their actualization
and the physical immediacy of their effects. To expand upon

111 Log 50
Reiser + Umemoto RUR Architecture, the actor metaphor, our models tend to function either like
O-14, Dubai, 2010. The design involved tightrope walkers engaged in real situational danger (presen-
an intensive back-and-forth with our
engineer, Ysrael Seinuk, who gave
tational) or like actors posturing as tightrope walkers (rep-
us a set of “rules” for the perfora- resentational). As one type of model clarifies and exaggerates
tions of the exterior concrete shell. experience, the other presents experience in all its particulari-
So long as we obeyed his general
ties and felt specificity.
rules of size and drift, laying out the
perforations could remain an entirely As with any healthy ecosystem, each type of model has
aesthetic concern from our point of its own place and purpose, and though we have our favor-
view. Conversely, we argued for a
ites, we would never claim that one type of model making or
thicker wall depth than structurally
necessary, solely because we thought functioning, representational or presentational, takes prece-
it looked better. The project passed dence over another. Furthermore, many, if not most, of our
through varying stages from the ideal- models rest uncomfortably in their categories. As material
ity of the rendered model to the inten-
sive material expression, cosmetic
actualizations, all representational models present themselves
ideality, and patination and perceptible and behave as contingent things in the world, and presenta-
sand streaking of the building. Images tional models always tend to represent in some capacity. The
courtesy the author.
intermixing and interbreeding of these types and functions of
models in our studio embody our architectural attitude, one
that mediates the representational and presentational roles of
all architectural artifacts.
Like all disciplines and ecosystems, ebbs and flows of one
attitude or species over another are but momentary fluctua-
tions in an otherwise general, longer-lasting stability. It seems

112 Log 50
Toyo Ito & Associates, Sendai obvious to me that our discipline currently finds itself in such
Mediatheque competition model, 1995. a moment. The presentational function of architecture has
Not only revolutionary phenomenally,
the project was a major challenge
ebbed, and representational concerns are flowing with an
to the preeminence of the structural alarming righteousness. From more than a few of the students
grid and engineering conservatism I encounter in design reviews to my colleagues and peers, any
traditionally encountered in Japan.
proposed attempt to actualize an architecture and take seri-
Normative structural grids were often
enforced as the only acceptable struc- ously the problem of its inevitable formal and material being
tural solution, more out of conven- is met with accusations of heroism, vanity, tragedy, or irrel-
tion than rationality. Photo © Tomio
evance. Perhaps taking cues from trends in the art world,
Ohashi. Courtesy the architect. Right:
Sendai Mediatheque, Sendai, Japan, proponents argue that what architecture can and should do in
2000. Photo © Nacasa & Partners. this present moment is represent the world in all its beauty,
Courtesy the architect. ugliness, banality, or indifference.
There is nothing wrong with such concerns entering our
discourse, so long as they are productive and enrich our work.
A healthy and diverse discourse can absorb an influx of ideas
both new and old. What is concerning is less the foreground-
ing of issues of representation, content, or meaning in archi-
tecture and more the concurrent dismissal of architecture’s
presentational function and formal, material reality. This
approach places architects in the position of mere passive rep-
resenters or commentators, subject to and making their work
a subject of the (often negative) forces in the world. In other
words, it forsakes our discipline’s unique power and agency:
to reconfigure and reject what already exists, to manipulate
the world by creating something foreign to it. What is lost is
the skill set to manipulate form and affect through the mate-
rials of our field, hard-earned techniques that require years
(or decades) of development.
Call me a romantic utopian, but the imminence of utopia
has always been for me a calling card of our field, something
that elevates it and separates it from other creative disciplines.
This is distinct from utopia’s actual delivery, a subtle but

113 Log 50
 often-overlooked distinction that can nullify architecture
when ignored. Accordingly, we are seeing the return of a
naive functionalism in architecture, an attempt to reify and
represent political programs and functions and to fix them
forever to things: a vain effort to forestall future events and
uses. A concurrent tendency is the return to politicization
through traditional representational forms, classical forms
of delivering meaning from narrative, autonomous pedestal-
bound sculpture, and even authorial sovereignty. I will coun-
ter with the words of my partner, Nanako Umemoto: “Form
and function do not correlate one to one, and that is the sav-
ing grace of architecture.” I would only add that the same
goes for meaning, and it is in ambiguity that architecture
becomes most meaningful.
This ambiguity necessarily involves awakening the pre-
sentational substrate that affords all representational endeav-
ors, and which allows images to surpass their surface-level
functioning as illusions. As artists explicitly contended with
during the 20th century and before, the material substrate of
the image does not run counter to the image but rather pro-
pels it into our world. Likewise, I do not blame myself or my
Spitfire model for its crashing in my room but rather hold
dear that it presented the possibility of flight and made an
eight-year-old boy believe that, if cared for and crafted atten-
tively, painted plastic parts could fly.
Quentin Meillassoux points to a performance of architec-
ture that perhaps more or less defines the field: architecture is
built impossibility. The mutability of perceptions and materi-
als, be they concrete or wax, is a precondition for architecture,
just as it is for all creative work. Things could always be oth-
erwise, and an architect must be an optimist. How wonderful.

Jesse Reiser (born 1958) is an archi-

tect and a professor of architectural
design at Princeton University. Prior
to cofounding Reiser + Umemoto, he
worked in the offices of John Hejduk
and Aldo Rossi.

114
Dora Epstein Jones
Models In and Out

Architectural history surveys tend to be toddlers among

surfaces, touching everything for a brief moment before
bouncing to the next thing. Depth is not expected, but some
explication of basic conventions is. For example, when
Marvin Trachtenberg and Isabelle Hyman present an image
of a wooden model by Michelangelo on page 310 of their
Architecture, from Prehistory to Post-Modernism, no explanation
1.  See Marvin Trachtenberg and Isabelle of the model’s provenance is given.1 The model is mentioned
Hyman, Architecture, from Prehistory to
Post-Modernism (New York: H.N. Abrams,
only as the preservation of a design, and after quite a few
1986), 310. pages on the representational techniques of the Renaissance
2.  See Robin Evans, Translations from
Drawing to Building and Other Essays
(plans, perspectives, lineaments), a discussion on models, or
(London: Architectural Association, 1997). even that model, is a curious omission.
3.  Robin Evans, The Projective Cast:
Architecture and Its Three Geometries
More curious is the relative silence on models in the oeuvre
(Cambridge: MIT Press, 2000), 302. of Robin Evans, the go-to scholar on all matters of architec-
4.  See Matthew Mindrup, The Architectural
Model: Histories of the Miniature and
tural representation. A discussion of “model dwellings” is with-
the Prototype, the Exemplar and the Muse out a model; a review of the model-heavy exhibition of Peter
(Cambridge: MIT Press, 2019).
5.  Jacques Derrida, Of Grammatology, trans.
Eisenman’s Fin d’Ou T Hou S concentrates intentionally only
Gayatri Chakravorty Spivak (Baltimore: on the subject of writing; and “Translations from Drawing
Johns Hopkins University Press, 1976), 158.
to Building” is without a model stop in between.2 The one
good mention of a model is the wire model of Le Corbusier’s
Ronchamp, but even there, Evans refers to it solely as “a rudi-
mentary but effective liaison among sculpture, drawing, and
building construction.”3
The most thorough text on the subject of architectural
models, by a good measure, is the recently published The
Architectural Model: Histories of the Miniature and the Prototype,
the Exemplar and the Muse by Matthew Mindrup.4 Chronolo-
gically scattered, dense, and overly comprehensive, Mindrup’s
argument is exemplified by the overwhelming feel of the text:
the model defies any one definition, any one use, any one his-
torical trajectory. Often reverting to language as a form of
analysis, Mindrup constantly reminds us that any definitive
fixity will be a temporary fiction. The word model alone serves
as a signal of its semiotic promiscuity: mode, as in modern,
and style, example, object, and ideal. Like Jacques Derrida’s
“There is nothing outside of the text,” a model can then be
everything – perhaps, to extend Mindrup’s line of reasoning,
a signified without a fixed signifier, a “trace.”5

115
 Without a clear definition or a precise canon on the
model, we, in architecture, are left with the model as an
enigmatic necessity: an undeniable part of the discipline
that automatically raises ontological questions. What kind
of model are we discussing? Is this a process model, a con-
cept model, a material model, a digital model, a presentation
model, a model of an idea, a model of a building, a model
of a model, or what? Subsequently, we have engaged in a dis-
cipline-wide task of critically discerning less what the model
is and more where the model “fits.” From Models to Drawings:
Imagination and Representation in Architecture implicitly prom-
ises a searching discussion of what editor Marco Frascari calls
a “metaphysics of presence,” architecture’s representation
6.  Marco Frascari, “Introduction: Models as “hidden,” and “not in some place.”6 This search is further
and drawings – the invisible nature
of architecture,” in From Models to
described in Bradley Starkey’s essay in the volume, “Post-
Drawings: Imagination and Representation in secular architecture”: “Architectural models exist somewhere
Architecture, ed. Marco Frascari, Jonathan
Hale, and Bradley Starkey (London: between the realm of ideas and the physical materiality of
Routledge, 2007), 6. buildings.”7
7.  Bradley Starkey, “Post-secular
architecture: material, intellectual, Of course, “somewhere” is made all the more confound-
spiritual models,” in From Models to ing by the use of the term model in the digital sense. While
Drawings, 234. Starkey sees the theorization
of architectural models to be historically probably not exactly a drawing nor precisely a physical object,
lacking because of the “post-secular” use the digital model is also both pervasive and seemingly place-
of the model in the modern discipline to
describe matter. He describes a resistance less.8 John May’s discussion of the digital model as an image
to engaging models because drawings are and all images as signals, or the model as a result of data inputs,
thought to represent ideas while models are
imagined to represent material realities. an effect of software, and the term model, bodes a kind of
8.  See Michael Meredith, “One Thing semiotic misdirection. We can call it a model or we can call
Leads to Another,” in Under the Influence,
ed. Ana Miljački (Cambridge: SA+P Press, it a movie or we can call it a cantaloupe, it’s all just data and
2014), 67–72; and Galo Canizares, Digital
Fabrications: Designer Stories for a Software-
rays of light carrying signals from eye to brain.9 Somewhere,
Based Planet (Novato, California: Applied maybe everywhere.
Research and Design Publishing, 2019).
9.  See John May, Signal. Image. Architecture.
So given what could be an utter promiscuity of the
(New York: Columbia Books on model, both in terms of what and where, I will temporarily
Architecture and the City, 2019).
pin into place an idea about architectural models. The term
that Evans used, liaison, is the chief clue. The architectural
model is primarily a communicative device of formal and
spatial understanding that cannot otherwise be derived from
a drawing alone. Models bridge gaps in understanding
concepts, volumes, spaces, tectonics, or any interaction of
architectural elements, or models tease out architectural
information. A model can be used to prognosticate or to con-
ceptualize based on the information that it provides to the
viewer, reader, or receiver, including the architect herself.
Certainly, there are now as many models as there are forms
of research in architecture. The point here is that, as a liai-
son, the model carries meaning, or to repeat the concept that
floats through Mindrup, the model is a kind of mechanical

116 Log 50
 Derridean “trace” that acquires meaning through an end-
less chain of signifiers, itself present, but its meaning gained
10.  I strongly urge a return to Derrida’s through sources outside of itself.10 As with all representational
Of Grammatology for more elucidation (or
confusion) on this topic.
devices in architecture, we are generally comfortable with
this arrangement, and certainly the model-as-trace can help
us account for the many modes of the model – analog, digital,
material, conceptual, and so on.

Somewhere, Out There

Following on the questions churned up by a discussion on
models, the question of the where is still open. One poten-
tial whereness, and the one of greatest concern to Frascari,
Starkey, et al., is the where in the design process. The “place”
of a model in a design process is one of daunting fluidity,
and on this “place” or even “space” of the process model, I
will remain silent.
My concern here is with the presentation model, the archi-
tectural model in its most traditional sense. These architectural
models, which are also the ur-models of our consciousness,
are not vaguely somewhere. They are extra-architectural.
They exist at the boundary point of the discipline between
those who can read a drawing and those who cannot, or in
blunter terms, “disciplinary insiders” and “disciplinary out-
siders.” We use models in the discipline to teach architec-
ture, and we (rightfully) admonish our students when their
models do not match their plans and sections, or when their
models are not accurately crafted. The model is a device for
bringing students into architectural thinking, as much as it
is a device for bringing clients in: it renders visible the spaces
of the proposed project, reveals relationships to the site, and
teases out the possibility of material and economic reality.
Maybe the omission of models in Evans and elsewhere,
or the frustration of placing the model in the larger field,
is because presentation models are, in the end, subject to
their place on the disciplinary boundary line. Architectural
models of any kind are expensive to produce; a presenta-
tion model often costs around $10,000, including labor, a cost
most firms will shoulder if the model can help tip the proj-
ect into economic reality. Architectural models are therefore
also implicated along the edge of a profit-driven exchange,
a way of creating a dialogue about a near-future building
with the motive forces of the developer, the client, and often
the builder. And, as charming as a model might be in terms
of its craft, as wondrous as Le Corbusier’s Villa Savoye as
it was presented in 1930, or as romantic as it might seem to

117 Log 50
reconstruct the fragments of a classical pile, it dwells pri-
marily in an exchange that has allowed architecture, and
Architecture, to exist in the world for centuries. Architecture
is not only a discipline, it is also the second-oldest profession.
A shift of attention to the model in this issue of Log and
elsewhere could signal a crisis or confusion over the term
model in a digital medium when once it referred to the physi-
cal. In these first decades of the 21st century, plenty of ago-
nizing over words is to be expected. Then again, a shift of
attention to the model might be signaling a turn to the modes
by which we communicate extra-architecturally – that is, a
shift in architectural intention. With this latter turn, we may
find ourselves staring into the face of a difficult matter: that
the profession and the discipline are not exactly aligned. And
nowhere is their misalignment more evident than in the pre-
sentation model.
The TV series Better Call Saul, often about the tensions
between conventional and unconventional lawyers, enacts
a dilemma similar to what we feel in architecture. Do the
briefcase, the suit, and the maple conference table make law-
yers seem more trustworthy or a safer bet? And are clients the
clearest path to legitimacy and value, both monetarily and
culturally? Kim Wexler (played smartly by Rhea Seehorn) is
a lawyer torn between her dual desires to do public service
and have corporate “success.” Over the course of many epi-
sodes, she lands a lucrative client, the local Mesa Verde Bank,
who needs her to represent them as the bank expands across
the Southwest. Wexler is the picture of legal-practice com-
mon sense, dressed neatly in tailored JCPenney separates,
ponytailed, fit, the human equivalent of the one-and-a-half-
inch heel. She fits well with the conservative but ambitious
Mesa Verde, but when her client proudly presents the bank’s
“room of models,” new branches of varying corporate styles
(modern, postmodern, high modern, faux modern) all dis-
played on pedestals, Kim reacts with… horror. For her, and
(presumably) the audience, the model display is the out-
ward sign of a world dominated by the transactional econ-
omy: little Mesa Verde banks everywhere. There is no visible
architect, only the legitimating force of the bank represented
by a lawyer who finesses zoning rules and building permits.
The bank meiotically phases into the architect of the future,
and in so doing, expresses through the presence of the many
models that the bank’s power over the world is both anony-
mous and absolute. Kim slowly backs out of the room and
then speed walks to her car.

118 Log 50
Mesa Verde Bank model room, Better The viewer’s take on this scene is “Kim questions her
Call Saul, “Something Beautiful,” sea- allegiance to the bank.” But the “architectural audience” sees
son 4, episode 3, aired August 20, 2018.
a somewhat different scene – a flicker of knowing recognition
and a momentary shiver of discomfort. The models are lit in
color, but they are all white, all midsized buildings, all differ-
ent styles, or as far as the cheap frame construction will take
it. Even close-up, the details speak to an architectural entropy
that we know too well: the more the model expresses its pop-
ular normativity, the less intentional it becomes to architec-
ture. And the less it belongs to architecture, the more it gains
meaning from somewhere, out there.

Or In or Both
Imagine the discipline of architecture as a very floppy bub-
ble. With apologies to Thomas Kuhn for the gross misin-
terpretation of his The Structure of Scientific Revolutions, the
edge of the bubble hosts both experimentation (“pushing the
boundaries”) and all possible interactions with other fields of
knowledge. The interactions, while on the edge, are capa-
ble of distorting the whole, of shifting the center or making
waves. Now, imagine the presentation model as a communi-
cative device that can talk in two directions: toward the cen-
ter or toward the edge. The model can speak inward to the
discipline or the model can speak outward to a consumer, cli-
ent, bank, television viewer, the faceless public, or anyone.
In the case of Better Call Saul, the presentation models
speak anonymously, “out there.” They are an extreme exam-
ple of a danger that exists in all presentation models: that

119 Log 50
Above and opposite page: Henry N. their existence as “trace” continuously and constantly leaves
Cobb, “Hypostyle,” SCI-Arc Gallery, them open to the full anonymity of corporate normativiza-
2015. Photos: Joshua White. Courtesy
SCI-Arc.
tion. This is not the same as an “authorship problem.” Rather,
it is more along the lines of what Joan Ockman observes as a
“normative architecture” in postwar America – a sliding
away of the utopian and socialist aims of the European avant-
garde in favor of the security of position and salary in a cor-
11.  See Joan Ockman, “Toward a Theory of porate world.11 It is no wonder that the presentation model
Normative Architecture,” in Architecture
of the Everyday, ed. Steven Harris and
seems to enjoy its starring role in the black-and-white photos
Deborah Berke (New York: Princeton of the era, as developers and architects shake hands in front of
Architectural Press, 1997), 122–52.
12.  Ockman’s argument revolves around
a large presentation model or Robert Moses leans over a table
the theoretical conception of Gilles with a model of Battery Park or Mies van der Rohe places his
Deleuze’s “major and minor” literatures,
with the corporate normative as major and hand on a diminutive Seagram.
the experimental act as minor. If the presentation model is capable of directionality in
13.  Keller Easterling, “The Dispositions
of Theory,” in 2000+: The Urgencies of its address, then it also makes sense that some presentation
Architectural Theory, ed. James Graham models can speak inward (which is more of a challenge),
(New York: GSAPP Books, 2015), 93.
and thus, the where of the presentation model could be a ful-
crum, both in and out. As a fulcrum, the presentation model
is anything but benign. The second that the intention of a
presentation model speaks outward, it carries implications
that genuinely inflect architecture’s meaning and impact in
the world – the real costs of labor and materials, the local
codes, urban and infrastructural planning, the client’s taste,
the developer’s pockets. In Ockman’s formulation, the “major

120 Log 50
tendencies” of the corporate profession speak to the immi-
grant’s desire to appear more “conservative,” especially Mies,
whom she casts as a bureaucratic poet.12 But what if the major
tendencies of a corporate profession are difficult for any paid
architect to resist?
Keller Easterling describes a tendency to imagine archi-
tectural theory as “isomorphic,” as a quasi-religious doctrine
“either for or against,” only engaged through argument and
dialectic. She then asserts that architectural theory needs to
change its “operating system” because the world simply does
not work in binary forms anymore: “One needs object forms
like those we are usually trained to make, but one also needs to
theorize active forms that are dispositional and time-released,
like bits of code to hack the operating system.”13
This canniness can be extended to the presentation
model. If the presentation model can speak both out and in,
and if it is simultaneously implicated as a commodity, then
why couldn’t we expect it to become a leveraging instrument
with a more curious, more intelligent, more contemporary
form of engagement?
Consider two different practices of presentation models:
Henry N. Cobb’s “Hypostyle,” an exhibition at SCI-Arc in
2015, and “Still Life: A Harvard GSD Exhibition 2015–2016,”
an exhibition of student projects in 2017, curated by Jennifer

121 Log 50
Bonner. “Hypostyle” was an experiment in spatial affect.
Near the heart of the gallery was a presentation model of a
typologically manifest columnar hypostyle arising from the
serial reproduction of Loos’s columnar Chicago Tribune
Tower competition entry. Surrounding the model and consti-
tuting a 1:1 model of a hypostyle were planar plywood struc-
tures, flange-like, with planar roofs that created a similarly
compressed but very different experience of space than in
the traditional columnar hall. Alongside the planar elements
were large crisp photos of the exteriors of eight building
projects by Cobb depicted as urban objects. Bonner’s “Still
Life” collected student presentation models from all areas
of the GSD, architecture, landscape architecture, and urban
design. Starting with chipboard models, she first grouped
the projects into odd associations such as “Las Vegas,”
“Inverted,” and “Pink Foam,” and then invited photographer
Adam DeTour to shoot the groupings under cartoonlike
lighting. The photos were then mounted on the wall behind
various other groups of models, which were also lit in pinks,
blues, and greens, and everything was “contained” behind
raw plywood-and-lumber stage flats. Viewing the exhibit
in the school’s hallway was like window-shopping for eye
candy, while at the same time trying to discern the many sets
of relationships and affinities among the projects.

122 Log 50
Right and opposite page: “Still Life: A
Harvard GSD Exhibition 2015–2016,”
March 20–May 12, 2017. Curated
by Jennifer Bonner. Photos: Justin
Knight Photography. Courtesy Harvard
Graduate School of Design.

The two exhibitions had common materials: scaled

project models on pedestals, plywood as a spatial separator,
effect-based lighting, and photography. However, they also
represented a wide spectrum of model “inwardness” and
“outwardness.” One inwardness was in the pedagogy – the
history of architecture that includes the hypostyle, the under-
standing of typology as a driver for design, the casting of the
model as a key mode of reading spatial relationships. Another
inwardness was the emphasis on experience over image mak-
ing, the critical commentary aimed at the discipline to value
the knowledges that experiences produce, whether the expe-
rience of the planar hall or the experience of being separated
from the student projects by the mechanisms of display.
The outwardnesses of these two exhibitions were can-
nier, more layered, exemplifying ways in which the presenta-
tion model can keep from slipping into the anonymous void

123 Log 50
 of Kim Wexler’s horrifying model room. One was the use of
photography to glamorize the project. Like a Snapchat filter
that softens facial lines or a pink light bath to “warm” the
scene, or even the use of Photoshop to enhance a rendering,
to crisp the edges or deepen the blue of the sky, the world of
capital and exchange exploits mood as desire. The dramatic
lighting of the Mesa Verde branch bank models is a tactic to
propel desire. The presentation model is expected to pull such
triggers, or to be accompanied by übersexy images or render-
ings of buildings in sunny, moody color palettes and atmo-
spheres. In both exhibitions, the tropes of atmosphere, color,
mood, and desire became wronged, slighted. Cobb’s impres-
sive buildings were relegated to walls that could not be seen
from a totalizing vantage point. At the GSD, the students’
models were held at bay behind the plywood flats. Neither
promoted a subsequent purchase of the projects, and that
resistance to commodification kept them from spilling over
to a place where the profession, or, to be more precise, cor-
porate or profit-driven normativization, could misalign or
malign the discipline. The outwardness of “Hypostyle” and
“Still Life” was also implicated by the idea of each exhibition
itself. Models were displayed, but whether through the many
misinterpretations of the categorized student projects or the
willful refusal to use the column, the exhibitions also resisted
the normative. In fact, neither exhibition felt like an exhibi-
tion – “Hypostyle” was labyrinthine, the model at the heart
like a stowed-away secret; “Still Life” felt like being back-
stage. As such, both exhibitions indicated what presentation
models can become for the discipline.
Always lurking in the presentation model is the devil’s
bargain of “selling out” to gain value – the fear that the cli-
ent, developer, builder will attempt to redesign the project
through the clarity and understanding that the model pres-
ents, and the fear that little can be done if that should happen.
Michelangelo could not prevent the pope from demanding
alterations to the crafted wooden model of St. Peter’s (the
very one in the Trachtenberg and Hyman book), and it’s
unlikely that architects today could do better. What we can do,
however, is see the open terrain of the model – the trace that
makes it difficult, if not impossible, to define; the same some-
where, everywhere of the design process, the same imper-
fect dialectic of in and out – as a distinct advantage. Make no
straightforward models. Serve architecture first.
Dora Epstein Jones, PhD, is chair
of architecture in the College of
Architecture at Texas Tech University.

124
Todd Gannon
Mind the Gaps!
Toward a Pedagogy
Of Models and a
Model Pedagogy
They appear just after Thanksgiving, around the time when
everything in Columbus begins to fade to wintry gray. There
are usually about 25 of them – squat, vaguely cubic, roughly
milk crate sized – arranged in an alien phalanx on the Big
Steps in Knowlton Hall, the Mack Scogin Merrill Elam
Architects–designed home of the Knowlton School at The
Ohio State University, where I lead the architecture pro-
gram. Easter-egg colors predominate and starkly contrast
with the coolness of Mack and Merrill’s interior concrete and
the clouds gathering outside. The objects exhibit a curious
density; their masses are less solid than stuffed with squishy
sleeves of space. Moving in for a closer look, notice that they
appear to have been sliced from larger wholes to reveal hints
of building-like orthogonality – a stair here, a balcony there,
the occasional stack of boxy rooms – nestled into otherwise
plump, invaginated volumes.
They are the final models of Knowlton’s first semester,
second-year graduate architecture studio (G2 for short). As
it turns out, they represent hotel designs for an urban site
in Cincinnati, but you would be hard pressed to figure that
out on your own. The drawings pinned up nearby uniformly
eschew labels, and the syllabus, assembled by instructors
Erik Herrmann and Sandhya Kochar and clotted with allu-
sions to the work of philosophers (Graham Harman, Martin
Heidegger, Bruno Latour), art critics (Michael Fried, Dave
Hickey), and mostly young architects (Andrew Atwood,
Jennifer Bonner, Jimenez Lai, Michael Meredith, Anna
Neimark, Hilary Sample), speaks not of hotels but rather, and
quite cryptically, of “a thing inside a thing inside a thing.”
Each year, the models trigger a wave of excitement that
ripples through the school. On final review day, undergradu-
ates who normally hang out on the Steps gawk at the curious

125
 invaders that have commandeered their lunch spot. G2 stu-
dents lean gregariously on one another with the air of post-
competition athletes. G1s chatter nervously, wondering if
they will be able to pull it off next year. Flinty-eyed G3s look
on in silence, like war veterans.
Studio reviews are uniformly lively and generate their
share of criticism, some of which – the kind that whines like
tinnitus at Ohio State, SCI-Arc, Penn, Pratt, UCLA, and other
institutions where advanced formal speculation remains the
norm – I’ll use as a jumping-off point here. The criticism takes
aim at the purportedly problematic relationship of speculative
work to the realities of conventional building tectonics and
functional performance. Critics (often the same students who
produced the models!) are quick to point out that, while inter-
esting, these models tend to be long on form and short on func-
tion, underdeveloped in terms of structure and program, and,
in some cases, improbable as buildings, raising thorny ques-
tions as to their ultimate value to the education of an architect.
True enough. In this studio, faculty suspend some techni-
cal questions and encourage students to pursue compositional
and spatial effects unperturbed by the often-stifling realities of
1.  These technical questions are pursued structure, program, and budget.1 If architecture is understood
more pointedly elsewhere in the curricu-
lum, including in the comprehensive studio
merely in the traditional (and dictionary) sense as the art and
that immediately follows this one. Beth science of building, a case can be made that these models fall
Blostein and I taught that studio last spring.
Recent student work from both studios can short. But if architecture is understood more broadly as “an
be seen at https://knowlton.work/G2. unfolding investigation into the possibilities of occupying
2.  Aphorism 14 in Jeffrey Kipnis, In the
Manor of Nietzsche: Aphorisms around and material form,” as Jeffrey Kipnis memorably puts it, these
about Architecture (New York: Calluna models are right on the money.2
Farms Press, 1990).
3.  See Todd Gannon, “Figments of the To achieve this broader understanding of the discipline,
Architectural Imagination,” Log 37: The Knowlton faculty often encourage students to proceed via
Architectural Imagination (Spring/
Summer 2016): 65–68. For a more counterintuitive experimentation and calculated risk-tak-
expansive treatment of the relationship ing. In other words, we sometimes ask them to do things the
of the profession and the discipline,
see Andrew Zago and Todd Gannon, wrong way on purpose. We do so not out of recklessness but
“Tabloid Transparency, or, Looking out of a desire to produce new forms of rightness. If all we
through Types, Legibility, Abstraction,
and the Discipline of Architecture,” in The asked was for our students to conform to best practices, all
Routledge Companion for Architecture Design
and Practice: Established and Emerging
they would ever do is repeat known successes. If we want
Trends, ed. Mitra Kanaani and Dak Kopec new successes, we need new practices. That, in a milk crate, is
(London: Routledge, 2015), 21–34.
what these models are about. If it ain’t broke, break it!
A key difference between this position and that of our
critics is that the latter generally want to close gaps they iden-
tify in the work – gaps between the models and the buildings
they represent, between education and the professional prac-
tice it predicts, and even those pesky gaps between materials
that signal poor craft in models and allow the rain into and
the heat out of poorly made buildings. Though we agree that

126 Log 50
Instructors Sandhya Kochar and Erik craft-related gaps should be eliminated, at Knowlton we usu-
Herrmann, A Thing Inside a Thing ally aim to open gaps – between models and buildings, schools
Inside a Thing Inside a Thing, final
models, Knowlton School, The Ohio
and offices, theory and practice, discipline and profession – in
State University, Fall 2019. All photos order to demonstrate that the most interesting work in any field
courtesy the Knowlton School. tends to emerge from the dissonance between the polarities by
which that field is defined. As I have argued elsewhere, a defin-
ing polarity in architecture is that between a profession under-
stood as a technical pursuit and a discipline understood as an art
form.3 By denying the assumption that an architect’s job is to
resolve the inherent misalignment of these aspects of architec-
ture, we aim to teach our students how to exploit the inevitable
gaps between them to generate new possibilities for the field.
All architecture is a function of a gap between concepts
and actuality, as even the dullest project architect will con-
firm. You know who I am talking about, that grouchy guy
(it’s always a guy) in every office who can’t stop telling the
interns to forget about that theory stuff they learned in school
and face up to the real world. Next time you visit a building
with that guy, point out a flaw. A misaligned light fixture, a
badly finished wall surface, any punch list item will do. As
if on cue, he will protest: “Well, it’s not supposed to be like
that.” In this, our rock-ribbed realist draws attention away
from the supposedly real world of the building to an appar-
ently realer one of its architecture. The conceptual doppel-
gänger that haunts all material form, what Rosalind Krauss

127 Log 50
 once labeled the “hermeneutic phantom,” is not an irrelevant
example of obscure theorizing, as our hapless PA might have
it, but a precise description of his favorite excuse for slipshod
4.  See Rosalind Krauss, “Death of a construction.4 And though he might not believe it, his protes-
Hermeneutic Phantom: Materialization of
the Sign in the Work of Peter Eisenman,” in
tations make it clear that he at least intuitively understands
Peter Eisenman, Houses of Cards (New York: that architecture cannot be reduced to buildings. Rather,
Oxford University Press, 1987), 166–84.
5. For an elaboration, see N. Katherine
architecture must be understood in terms of the persistent
Hayles and Todd Gannon, “Virtual gap between material form and immaterial concepts.5
Architecture, Actual Media,” in The SAGE
Handbook of Architectural Theory, ed. C. Greig In my drawing attention to gaps, Log readers may notice an
Crysler, Stephen Cairns, and Hilde Heynen echo of Harman, whose book on H.P. Lovecraft centers on the
(London: SAGE, 2012), 484–500.
6.  “One of the most important decisions philosophical implications of the production and destruction of
made by philosophers concerns the produc- gaps.6 Less obvious, but perhaps more important to my argu-
tion or destruction of gaps in the cosmos.
That is to say, the philosopher can either ment, is the debt I owe to the late sociologist Murray S. Davis.
declare that what appears to be one is actually In an influential 1971 essay, Davis points to the importance of
two, or that what seems to be two is actually
one.” Graham Harman, Weird Realism: denying the expectations of one’s audience in order to produce
Lovecraft and Philosophy (Winchester: Zero
Books, 2012), 2. Emphasis original.
interesting social theories.7 Here, theories have to do primar-
7.  See Murray S. Davis, “That’s Interesting! ily with the relationship between what seems to be and what is.
Towards a Phenomenology of Sociology
and a Sociology of Phenomenology,”
That is, in its most basic form, a theory states whether phe-
Philosophy of the Social Sciences 1, no. 2 (June nomenology aligns with or diverges from ontology. Whether
1971): 309–44.
8.  In the interest of brevity, I will refrain
a theory is found to be interesting, by contrast, has more to do
from swerving through the literature on with its audience, and occurs when a prevailing assumption
interesting as a concept and a category.
Those who wish to do so ought to begin
about the relationship between seeming and being is denied.8
with Sianne Ngai’s Our Aesthetic Categories: Most theories, like the common sense into which suc-
Zany, Cute, Interesting (Cambridge: Harvard
University Press, 2012), which is required cessful ones degrade, work to eliminate gaps to make the
reading for anyone wishing to use the term world more manageable, understandable, and predictable: if
intelligently; Mark Dorrian’s “What’s
Interesting? On the Ascendency of an it seems too good to be true, it probably is; humans evolved
Evaluative Term,” Architecture and Culture 4, from apes; E = mc2. Interesting theories, on the other hand,
no. 2 (2016): 173–84, which provides an
incisive analysis of the term’s recent use “constitute an attack on the taken-for-granted world,” and
in architecture; and Andrew Atwood’s always upset prevailing assumptions.9 Humans evolved from
Not Interesting: On the Limits of Criticism
in Architecture (San Francisco: Applied apes?!? Mass is equivalent to energy?!? Of course, interest is
Research and Design Publishing, 2018),
which, in spite of its arguments, is interest-
always context dependent. However shocking they once were,
ing in exactly Davis’s sense of the term. Darwin’s theory of evolution and Einstein’s theory of special
9.  Davis, 311.
relativity have, for most of us, long since slouched into the
10.  Ibid., 313. Cf. Harman’s take on
philosophical gaps, quoted in n. 6. comforting familiarity of common sense.
11.  Ibid., 309. Emphasis original. Davis’s formula for interesting social theory is simple and is
12.  Ibid., 311.
easily transported to other fields: “What is accepted as X is actu-
ally non-X.”10 Importantly, Davis is careful to point out that
theories need not be true to be interesting. (His opening lines
ought to be tattooed on the inner forearm of every aspiring the-
orist and PhD student: “It has long been thought that a theorist
is considered great because his theories are true, but this is false.
A theorist is considered great, not because his theories are true,
but because they are interesting.”11) In fact, theories that merely
verify prevailing assumptions are generally dismissed as their
truth is affirmed. “That’s obvious,” we huff and move on.12

128 Log 50
Mack Scogin Merrill Elam Architects,
Knowlton Hall, Columbus, Ohio, 2004.

As responsible professionals, architects often, and rightly,

work to eliminate gaps by aligning phenomenology with
ontology. This column seems sufficiently strong to carry the
load placed on it and, in fact, is sufficiently strong. This win-
dow seems to be the right size to provide light and air to this
room and, in fact, is the right size. Useful buildings often
result from the application of such commonsense thinking.
Interesting architecture requires more.
Interesting architecture (which is to say, architecture)
most often results when architects deny prevailing assump-
tions. Henri Labrouste understood as much when he speci-
13.  A Questioner, “Questions for Colin fied columns that appear insufficiently strong to carry the
Rowe,” ANY 7/8: Form Work: Colin Rowe
(1994): 34; reprinted as “Interview: 1989,”
vaults they support in the reading room at the Bibliothèque
in Colin Rowe, As I Was Saying: Recollections Nationale in Paris. Colin Rowe knew it too when he quipped,
and Miscellaneous Essays, ed. Alexander
Caragonne, vol. 2, Cornelliana (Cambridge: “If you see a building with windows of a size to admit an
MIT Press, 1996), 356. Here, Rowe guesses appropriate amount of light, it may or may not be a work of
that the anonymous questioner was
Richard Ingersoll. architecture; but, if the windows are definitely too big or def-
14. If architecture exists as an affront to initely too small, then you can be almost certain that you are
building, it also sometimes finds itself at odds
with theory, which too often is expected in the presence of an architectural endeavour.”13 The implica-
to serve as a model for practice. Normally, tion is startling. Adherence to convention only sometimes leads
architectural theories become interesting
when individual buildings conform to a to architecture; affronts to convention nearly always do.14
theory’s general principles. Individual The G2 models should be understood in the same spirit.
buildings, on the other hand, become
interesting when they confound a theory’s If you see a model with interior spaces appropriate to their
necessary generalizations. For proof,
compare the built work of any architect who
intended uses, it may or may not be a work of architecture;
has written a theory to the general principles but, if you see a model with interior spaces that wildly exceed
of that theory. You will find that the work
is most interesting when the gap between
what is appropriate for their intended use, you can be almost
the architecture and the theory is most certain that you are in the presence of an architectural
pronounced. Given this, we should return to
understanding (and teaching!) building as endeavor. I allow that some (but not all) of these models are
a cause for, not an effect of, theory, at least implausible as buildings. I also submit that most (if not all) of
until that approach ceases to be interesting.
them point to exciting new possibilities for architecture.

129 Log 50
 In this, our students take up the challenge Mack and
Merrill posed to them with Knowlton Hall, which my col-
leagues and I often hold up as a model for speculative prac-
tice. The building, with its undulating curves, odd shingles,
and crystalline voids chiseled into its marble mass, flies in the
face of its staid, brick-clad context to open a yawning chasm
between what it is and what campus buildings are expected
to be. Where most buildings at Ohio State offer bland stacks
of single-height classrooms clipped to double-loaded corri-
dors, Mack and Merrill conjure an architecture of dizzying
volumetric complexity. Triple- and even quadruple-height
spaces soar up past faculty offices, through design studios, and
around glass-clad laboratories that cling to the ceilings like
wayward balloons. Outlandishly long ramps guide visitors up
through a section torn open by split-level floor slabs pinned in
place by concrete columns. Resisting the regimentation of an
orthogonal grid, these columns are distributed loosely across
the plan and scaled like oaks. On entry, first-time visitors typ-
ically raise their eyes skyward and fall reverently silent. Only
after they have collected themselves do they begin to scratch
their heads and wonder how the hell they are supposed to
find the lecture hall. (Don’t worry, just head up the ramp!)
Once occupants have spent a little time figuring the place out,
new possibilities begin to emerge. Curious nooks offer cozy
spots to cram for a quiz or sneak in a nap. Gaps between floor
levels provide more than just spatial excitement; they foster
casual communication between normally isolated floors, with
students exchanging tips or gossip and occasionally lobbing
snacks across the void.
Where most buildings segregate occupants into clearly
defined rooms, this one joins them in loosely layered, over-
lapping zones. The G2 models attempt something similar.
Though they may not yet exhibit professional control to the
degree that is everywhere apparent in Knowlton Hall, they
signal a commitment to speculative possibility that is utterly
lacking in far too many buildings today. Reliance on con-
ventional methods and best practices alone, which manage
expectations by foreclosing uncertainty, quite simply cannot
produce possibilities like these. Indeed, the more we adhere to
them, the less likely we are to produce architecture at all.

Todd Gannon is Robert S. Livesey

Professor and head of the architecture
section at The Ohio State University’s
Knowlton School.

130
Jimenez Lai
Ephemeral Models’
Permanent Ghosts
In December 2019, the final review of students’ work in the
Market Forces studio took place at Cornell’s College of
Architecture, Art, and Planning. A photograph shows a stu-
dent standing next to an oversized final model and facing a
group of critics. Literally onstage in the auditorium space of
Milstein Hall – thus truly a theatrical event – that occasion
may have been one of the last in-person final reviews of the
fall semester. December 2019, a time when breathing the same
air together did not raise fear of contamination.
The student soliloquy that encapsulates his or her efforts
that semester is a commonplace in schools of architecture.
The evaluation process in the design studio, unlike examina-
tions in most other university disciplines, is highly perfor-
mative. Students orate a narrative about a fictional building,
guiding attention by using scaled physical models – almost
like puppets – to animate that fiction, while standing in front
of a backdrop of plotted drawings, diagrams, renderings, or
photographs. In this ephemeral performance – lasting 15 to
30 minutes, sometimes up to an hour – the critics have the
front-row seats. Supportive classmates, friends, or family
constitute the remainder of the audience, usually at the
fringes or on the sides. This theatrical space transforms the
review from a proscenium stage for a solo act into a theater-
in-the-round, where the quarrels or politics between the crit-
ics become the central focus, often eclipsing the very model
they are gathered around.
In 2020, the reality of this review changed, as the first
wave of Zoom-based final studio reviews began in response
to the need to contain the global pandemic. How, in a world
of screen-based live broadcasts, with students and critics flat-
tened in a grid of faces, could we deliver the same experi-
ence of the interactive puppet theater we are so accustomed
to? The pandemic forced us to engage a distant audience, but
more significantly, it also drew new awareness to the medi-
ums we rely on, such as the exhibition, documentation, and
dissemination of models.
After final reviews, for many architecture schools, a year-
end exhibition plays a vital role in maintaining the pulse of

131
Final review of the studio Market
Forces, taught by Kevin Carmody,
Andy Groarke, and Rodolfo R. Dias, in
Milstein Auditorium, Fall 2019. Photo:
Bill Staffeld. Courtesy College of AAP,
Cornell University.

the school and its intellectual and social properties. For exam-
ple, the “End of Year Show” at The Irwin S. Chanin School
of Architecture at The Cooper Union has been a consistent
and reliable space of congregation for decades, both as a dis-
cursive space to present academic content and as an immer-
sive space for social interaction. However, when the schools
had to discontinue in-person teaching, adjustments had to
be made quickly. To host their traditional show, the Cooper
Union community (led by Dean Nader Tehrani, Steven
Hillyer, and Farzin Lotfi-Jam) endeavored to construct a dig-
ital version of the school’s Foundation Building. The result,
launched online on June 17, is a virtual version of the building
and contains digital presentations of student work. As if in an
emergency retreat, Cooper Union hauled its treasured student
work into a web space beyond the real.
Jorge Luis Borges, in his short story “On Exactitude in
Science,” describes a 1:1 mapping effort in which the scale
of the map would match the scale of the territory it docu-
mented. Similarly, the “End of Year Show” is an interactive
3D model that exports the ghost of Cooper Union as a precise
nonphysical counterpart to the physical reality we usually
inhabit. The exhibition design captures the spirit of how the
individual studio projects normally occupy their setting, now
in a digital realm and aided by technological features impos-
sible to achieve in the physical realm, such as the defiance of
gravity, the ability to zoom in to models, pop-up annotations,
and digitally enhanced navigation systems. Users point and
click to advance to various adjacent rooms, which are seen
in first-person perspectival views. Alternatively, viewers can
navigate with a global map of the building to identify loca-
tions of studio exhibits. In each room, a studio setup includes

132 Log 50
“End of Year Show,” The Cooper Union, pinned-up texts, drawings, and sometimes models, like an in-
launched June 17, 2020. Image cour- person installation. At times, the digital setup enhances the 1:1
tesy Cooper Union.
map with embedded digital footnotes, such as floating speech
bubbles that, clicked, link to prerecorded YouTube videos. The
videos present instructors explaining studio briefs or anima-
tions and short films of studio projects. The effect renders both
the world and the map at whatever scale viewers zoom in to.
Suddenly, the scale model no longer retains the resolutions of
study, nor does it transform into a 1:1 puppet when students
present work. In the “End of Year Show,” scale becomes a
necromantic instrument for resuscitating the nearly lifeless
Foundation Building as a space of magical realism.
This exhibition format has parallels with video games
like Grand Theft Auto, an open-world game that elevates
the exploratory potential of game worlds, and also expands
representational possibilities for architects. In open-world
games, players can freely explore the map and write their
own stories in the process. Similarly, the “End of Year Show”
presents elements of free will and exploration. The repre-
sentational model of open-world games takes the subjec-
tivity implied by first-person perspective one step further
by allowing for the element of choice. To enable viewers to
choose how they navigate within the perspectival drawing is
to introduce free will into the model.
Visual and textual media persist in a way that physical
buildings cannot. For example, the photograph of the demoli-
tion of the Pruitt-Igoe housing complex represents a nanosec-
ond in the life of the buildings, but through this time freeze,
Pruitt-Igoe, as media, became permanent. We can never phys-
ically visit this moment, much in the same way we cannot
visit the fictionalized Trellick Tower in J.G. Ballard’s novel

133 Log 50
 High-Rise. The permanence of the media representation,
when compared to the fleetingness of the moment, trans-
forms the media that document the event into a reality. We
use these media to extract the ghost of a reality that is more
meaningful than the experience of the space itself.
Final reviews of architecture studios used to disappear
into thin air, but the new online review and exhibition for-
mats now demand that architectural representation trans-
form into more permanent media for repeat broadcasts.
Whereas the in-person review offers a space of an outside-
in, the online digital exhibition format generates a separate
space of the inside-out. The 1:100 physical model, once play-
ing the vital role of a 1:1 puppet in the ephemeral theater of
academic evaluation, is now dissolved into the realm of the
1:1 digital map, eternally documented as a discrete and closed
reality. Where Robin Evans once declared, “Architects do not
make buildings; they make drawings of buildings,” in a future
post-COVID world, architects may encounter a new pres-
sure to move beyond drawings, models, and performances,
1.  Robin Evans, “Architectural Projection,” and toward a multimedia set of interactive documentations.1
in Architecture and Its Image: Four Centuries
of Architectural Representation: Works from
In practice, architects are likely to renegotiate the function of
the Collection of the Canadian Centre for the 3D model as a separate 1:1 reality for an audience waiting
Architecture, ed. Eve Blau and Edward
Kaufman (Montreal: Canadian Centre for to experience the 1:1 physical realm. In education, the online
Architecture, 1989), 21. delivery and exhibition will become a new medium, in addi-
tion to the representational tool set we already cultivate.
Two branches of the same discourse, the immersive and the
discursive spaces of the architectural review present two
separate models: one live and seemingly ephemeral, and one
mediated and seemingly permanent.

Jimenez Lai works at Bureau

Spectacular, an office based in
Los Angeles.

134
Marshall Ford
Blue Foam

The story of Styrofoam begins with US Patent 2,450,436, filed

July 26, 1947, by Dow Chemical Company engineer Otis Ray
McIntire, and issued October 5, 1948. By its description, the
patent “concerns an improved method for the manufacture of
cellular thermoplastic products”; more specifically, it outlines
a “method for the preparation of cellular polystyrene which
involves heating solid polystyrene and a gas such as methyl
chloride in a closed vessel under a pressure. . . . During flow
from the vessel the polymer is swollen by expansion of the gas
and is caused to assume the form of a somewhat elastic, non-
brittle, cellular body composed for the most part of individual
1.  McIntire, Otis Ray. Manufacture of closed cells.”1 The resulting product is the extruded polysty-
cellular thermoplastic products. US Patent
2,450,436, filed July 26, 1947, and issued
rene (XPS) brand Styrofoam – not to be confused with the
October 5, 1948. colloquial use of “styrofoam” for the expanded white poly-
2.  Tim Lacey, Dow business director,
quoted in Lydia DePillis, “You have never
styrene used for disposable cups and plates – better known in
actually used a Styrofoam cup, plate or architecture as blue foam. Originally produced in white logs,
takeout box,” Washington Post, December
18, 2013, https://www.washingtonpost. the now recognizable blue color is a branding strategy for
com/news/wonk/wp/2013/12/18/you- Dow: “When people see the blue dye, and they see Styrofoam,
have-never-actually-used-a-styrofoam-
cup-plate-or-takeout-box/. to our customers, that’s a promise that they’re going to get
3.  Ronald Rietveld, “Hardcore Heritage” the [Dow] people, the knowledge and the relationship. We
(lecture, Carnegie Mellon University,
Pittsburgh, October 24, 2016), available actually do make it blue for that reason.”2
online as “Fall 2016 Lecture Series – Ronald Today, blue foam is commonplace in architectural prac-
Rietveld,” Vimeo video, 1:13:39, posted by
“CMU School of Architecture,” August 2, tice and academia as a liberating medium for designers to
2018, https://vimeo.com/282913339. roughly and quickly explore concepts. At the 2010 Venice
Architecture Biennale, “Vacant NL,” the Dutch Pavilion
curated by Rietveld Architecture-Art-Affordances (RAAAF),
used blue foam to model the nation’s vacant government
buildings. RAAAF worked with a multidisciplinary team of
Dutch designers and artists to build scale models of thousands
of buildings, compiled into a massive field suspended on wires
in the otherwise vacant pavilion. In an embrace of blue foam’s
universally interpretive qualities, Ronald Rietveld said of his
team’s project, “Once you came upstairs you suddenly saw
the whole sea of vacancy, and then it was not about blue foam
buildings, but about real buildings with a real location and
real qualities.”3 In the 2019 MVRDV exhibition “Architecture
Speaks: The Language of MVRDV,” the blue foam model
was specifically defined as an element of their architectural

135
Rietveld Architecture-Art-Affordances,
“Vacant NL,” Dutch Pavilion, Venice
Architecture Biennale, 2010. Blue
foam, steel cables, and wood. Photo:
Rob ’t Hart. Courtesy the architect.

vocabulary, and visitors could “create their own stacked

4.  For more on the exhibition, see designs from blue foam.”4
“Architecture Speaks: The Language of
MVRDV,” MVRDV, https://www.mvrdv.
The recent adoption of blue foam to coherently yet
nl/projects/394/architecture-speaks-the- abstractly communicate organization, form, and concept has
language-of-mvrdv.
5.  See Rob Krier, Urban Space, trans.
even led to firms producing computer-generated rendered
Christine Czechowski and George Black images with blue foam as the design material. This treatment
(London: Academy Editions, 1979).
of blue foam as an objective medium with endless opportu-
nities for architectural manipulation and interpretation has
established it as a universal expressive material of design.
Perhaps the first architect to embrace the potential of
conceptualization in blue foam was Rob Krier, whose influen-
tial book Urban Space (1979) represents several design schemes
with dramatically choreographed black-and-white photo-
graphs of three-dimensional cityscapes built in Styrofoam
painted gray.5 The clarity of intent found in the gracefully
carved insulation material brings the urban diagram to life,
introducing a new dimension, scale, and voice through a
chemically complex and infinitely versatile medium. Both
trivial and critical, blue foam was simply easy to work with.
Parallel to the crude but promising digital methods emerg-
ing in the ’80s, ’90s, and into the 2000s, blue foam became an
instantly recognizable method for exploration in architecture.
Used for the rough expression of programmatic elements or
formal moves, the material-focused process often manifested
the building concept itself after designers embraced the elemen-
tal moves permitted by the simplicity of the material. This
relationship between architecture and blue foam became partic-
ularly evident in the work of Dutch architecture and urbanism
offices specializing in data, polemic, irony, and iterative physical
model making. OMA’s S,M,L,XL and MVRDV’s KM3, released
in 1995 and 2005 respectively, exhibit a range of projects whose
initial and ultimate expressions are through blue foam mod-
els, including OMA’s City Hall The Hague competition proposal

136 Log 50
Chris Norton Riley, Readymades,
Preston Scott Cohen’s design studio,
Harvard Graduate School of Design,
2016. Photo courtesy the architect.

and MVRDV’s Book Mountain in Spijkenisse, Netherlands. The

images for both NL Architects’ and UNStudio’s early projects
tell a similar story. “In a blue foam model, you cannot incorpo-
rate all the details you can include in a BIM model,” says former
OMA architect and HQ Architects founder Erez Ella. “You have
to edit it down to the essential. . . . You immediately recognize
OMA buildings because they are clear. You understand the big
moves and the small moves, because you have to edit down
6.  Erez Ella, “Interview,” CLOG: Rem the ideas to the bare essentials.”6 Former OMA partner Ole
(2014): 73.
7.  See “One Month Without Models!,”
Scheeren says the process would start with between 30 and 50
in Albena Yaneva, Made by the Office for conceptual models.7 MVRDV cofounder Winy Maas explains the
Metropolitan Architecture: An Ethnography of
Design (Rotterdam: 010, 2009), 45.
purpose of “using an enormous amount of blue foam to make
8.  Winy Maas, “Future-utopian,” interview an endless number of models for comparison. Afterwards, you
by inter•punct, March 2014, https://inter-
punct.pub/inter-view-Winy-Maas.
see distinctions, the offspring, and all of the interesting differ-
ences as well.”8
As the use of blue foam grew, from the established Herzog
& de Meuron to then up-and-comer Bjarke Ingels Group, firms
explored formal iteration with carved blocks of blue foam, con-
necting the process of their professional studios with their aca-
demic studios. Interns at OMA often found themselves working
on the notorious floor of the Rotterdam office dedicated to blue
foam models. In his 2014 studio Indebted Architecture at the
Harvard Graduate School of Design, Preston Scott Cohen man-
dated the use of blue foam as the sole method of construction for
final models. Former students, including Preliminary Research
Office principal Chloe Brunner, describe Cohen’s choice as an
attempt to apply a unifying quality to all projects, while simulta-
neously using blue foam’s thickness in unorthodox ways to gen-
erate productive conflicts and architectural tensions.
Before its importance to and ubiquity in the design pro-
cess today, blue foam was originally a rediscovery of Swedish
inventor Carl Georg Munters’s material for insulating refrig-
erators. Munters successfully developed an expanded foam

137 Log 50
 plastic by introducing a soluble gas (typically methyl chlo-
ride) under pressure to a melted polymeric compound like
polystyrene. When the pressure was rapidly reduced, the
substance bubbled and hardened into an extremely light,
9.  Munters, Carl Georg, and John highly porous structure with excellent insulating properties.9
Gudbrand Tandberg. Heat insulation. US
Patent 2,023,204, filed August 20, 1932,
After acquiring the right to use Munters’s patents, McIntire
and issued December 3, 1935. Also see improved this method in efficiency and scale. The foam could
“Carl Munters posthumously honored as
ASHRAE Pioneer of Industry,” Plumbing be produced in logs three feet in diameter and 60 feet long,
& Mechanical, March 1, 2017, https:// then cut into sheets and used as extraordinarily light and flex-
www.pmmag.com/articles/100425-carl-
munters-posthumously-honored-as- ible insulation at an architectural scale. With a better R-value
ashrae-pioneer-of-industry. per inch than almost all other insulation, and given its supe-
10.  See Alex Wilson, “Rigid Foam
Insulation and the Environment,” rior performance in controlling air gaps and its inherent water
BuildingGreen, July 1, 1992, https:// resistance, XPS rigid insulation is often the preferred choice in
www.buildinggreen.com/feature/
rigid-foam-insulation-and-environment. construction, with these benefits typically outweighing rela-
11. “DuPont™ Styrofoam™ Brand Extruded tively higher costs. Architects specify the use of rigid insula-
Polystyrene (XPS) Foam Insulation reduced
GWP update, DuPont,” https://beyondblue. tion in many projects to allow for thinner building envelopes
dupont.com/compliance. and simplified exterior waterproofing details.
12.  See Make it With Styrofoam™, KAP
Kraft Books PB-47 (Chicago: Graff Unfortunately, blue foam has never been environmentally
Publications, 1969). friendly. Methyl chloride’s extreme toxicity and combustibil-
ity led to most manufacturers using chlorofluorocarbons in the
’60s. In the ’90s, producers veered away from CFCs and toward
less toxic but still environmentally damaging and ozone-
depleting gases, as well as carcinogens.10 Blue foam scraps in
design studios may be repurposed for successive rounds of con-
ceptual studies, but the large-scale use of blue foam as building
insulation leads to waste that is much more difficult to recycle
and can take thousands of years to biodegrade. At the same
time, blue foam also plays an outsize role in creating energy-
efficient buildings, and starting in January 2021, according
to DuPont, which merged with Dow in 2017, Styrofoam will
transition to a reduced global warming potential (GWP) for-
mulation and be produced in a “design friendly grey.”11
But it could be argued that the use of blue foam for mod-
eling was intended, or at least foreseen, from the material’s
inception. In the ’50s and ’60s, Styrofoam was used for practical
things like life jackets and building insulation, and encouraged
as a versatile crafting material for children. A 1969 pamphlet
from KAP Kraft Books, in collaboration with Dow, offers
instructions for complex Styrofoam models of tepees, dolls, a
birdcage, and a rickshaw.12 The material’s vulnerability to heat
made it easy for people to cut through sheets of it with a hot
Marshall Ford is an architectural knife, needle, or wire – techniques that would later be used in
designer and writer. He is cofounder
architecture schools. Perhaps one of the most resonant symbols
and principal of Small Office, an
architecture studio based in San Luis of architecture today, blue foam now embodies a discipline that
Obispo and Tokyo. spans from concept to construction, from play to work.

138
David Erdman
Mottle Behavior

Given that the discipline, profession, and academies of archi-

tecture are struggling to find a voice amid multiple pandem-
ics – virological, climatological, and social – the notion of
“model behavior” in this issue of Log can be seen as both on
and off topic. The idea that a scientifically driven and math-
ematically supported predictability model shapes our daily
rituals could not be more pronounced than in this moment.
Our very sense of culture and the urbane is presently being
dictated by epidemiological models that are forcing us to rein-
scribe personal, urban, and social modalities in response to the
behavior of a “novel” virus. This particular scientific model
of predictability underscores many of the aspirations of digi-
tal modeling and relates to the premise of architectural mod-
eling as we know it today, where the model is a “theorem” of
1.  While indeterminacy and unpredict- a building to be “proven” through full-scale construction.1
ability were central themes in architectural
discourse in the late 20th century, one
Whether model behavior describes people and their moral
could understand the role of the model or ethical behavior or the behavior of architectural models,
and its requisite behavior as going in the
inverse direction. I would cite Frei Otto’s the phrase supports the idea of the model as a tool for divining
computationally based large-scale physical a conceptual framework for success and perfection: something
models for the Munich Olympic Stadium
(1968–72) as the opposite of the discourse to emulate. But what if we make a slight linguistic adjustment
on indeterminacy emerging around the – or slip in pronunciation – from model to mottle? To mottle is
same time in Vienna, New York, and
(later) Los Angeles. In Otto’s work, model to stain or mark inconsistently, actions not generally associ-
behavior was all about testing and predict- ated with exemplary design technique. Thus mottle behavior
ing structure. This rationalization of
structure via computation modeling set the is contrary to the established modalities of modeling in archi-
stage for 30-plus years of software develop-
ment and prototyping, propelling the
tectural education and practice. It suggests an embrace of the
model to be largely understood as a scaled imperfections, defects, stains, and patinas of materiality in
replica of what will be proven through
full-scale construction. This concept of the
the behavior of a model. It can be understood as “noise” in
model and the citation of the very same the system, deviations from the behavioral mean, distracting
Otto model plays a central role in an essay
by Rob Whitehead titled “Model Behavior: from the data and introducing unpredictable outcomes.
The Evolving Use of Physical Prototypes The different cultural connotations and inferences of
in Structural Shell Design, 1959–1974,”
(presented at the Association of Collegiate model and mottle are important in light of architects’ rising
Schools of Architecture Annual Meeting interest in social and environmental justice, where “model
104, Shaping New Knowledges, Seattle,
March 18, 2016). See also Rob Whitehead, behavior” is being defined by governments, corporations, insti-
Structures by Design: Thinking, Making, tutions, and individuals alike. Architectural models, whether
Breaking (New York: Routledge, 2019).
digital or physical, have the unspoken mission to be clean,
precise tools for conveying instructions at a detailed, tectonic
scale. They draw into focus overall intent and advocate for
the full and visible transparency and predictability of data.

139
 Redefining a term like mottle as an attribute and behav-
ioral asset could introduce the possibility of unpredictable
outcomes and ulterior agendas into the milieu of architec-
tural models. There is a shrewdness to mottle behavior that
may be useful to examine in our current pandemic situation.
Two particular examples come to mind: an exhibition of
architectural models from ancient American civilizations and
a recently completed memorial to enslaved laborers.

Pre-Columbian Models
“Design for Eternity: Architectural Models from the Ancient
Americas,” an exhibition at the Metropolitan Museum of
Art in New York (October 2015–September 2016), assem-
bled some of the earliest examples of architectural models
from an array of pre-Columbian civilizations. The exhibi-
tion suggested alternative ways one might understand the
scaled architectural model by drawing upon the scholarship of
archaeologists and the design ingenuity of these historic and
creative cultures. According to curator Joanne Pillsbury, the 56
models, which constituted a small exhibition by the museum’s
standards, had a significant public impact. Borrowed from
various collections around the world, the models seemed to
have a kind of cult status among museumgoers, for both the
number of visitors and the duration of their visits were unan-
ticipated, if not record-breaking.
Mottling was manifest in three ways in the 56 mod-
els: geographically, temporally, and materially. The precise
geographic location depicted and the representational accu-
racy of each model’s structure or internal composition were
mottled – that is, unclear. The makers seemed to combine
elements from various places or depict actual buildings in
new ways by misscaling various elements. In a stirrup spout
vessel, for example, stairs from a known structure are exag-
gerated. The models often have a somewhat distant relation-
ship with the buildings they represent: they are not replicas
per se. As Pillsbury notes, these objects collectively expanded
the term model to include representations that are “smaller-
scale embodiments” of actual structures, an idea reinforced in
Juliet Wiersema’s catalogue essay, “The Art of Ancient Andean
2.  Joanne Pillsbury, “Building for the Architectural Representations.”2 More than an encrypted scale
Beyond: Architectural Models from the
Ancient Americas,” in Joanne Pillsbury et
representation of something “other,” this notion of embodied
al., Design for Eternity: Architectural Models representation is akin to mottling in that it suggests the model
from the Ancient Americas (New York: The
Metropolitan Museum of Art, 2015), 5. has interactive dimensions inextricable from the artifact itself.
A number of the models were funerary, buildings that
the deceased had been fond of, which represented their

140 Log 50
House Model, 100 BCE–200 CE. careers or depicted certain events in their lives or even their
Nayarit culture, Mexico. Ceramic, 12 afterlives. In this sense, the models were representations of
by 10 by 6 3/4 inches. Right: Temple
Model, 200 BCE–200 CE. Colima cul-
simultaneous pasts and futures, temporally mottled, unfixed
ture, Mexico. Ceramic, 7 by 7 by 4 to a singular place or time. In Pillsbury’s catalogue entry,
1/2 inches. Shown in the exhibition “Building for the Beyond,” these models are seen to introduce
“Design for Eternity: Architectural
ulterior interpretations and content as much as they depict or
Models from the Ancient Americas,”
Metropolitan Museum of Art, New York memorialize elements of an actual building.
City, October 2015–September 2016. An important aspect of these models is their materiality,
Photos courtesy the Metropolitan
mottled with the patina of aging, which enhances the sur-
Museum of Art.
face qualities of ceramic, stone, bronze, and wood. In many
cases, including a scepter from Sipán, mottling was designed
into the artifact by using multicolored pigments. Rather than
objects of perfection, these are objects of interaction and nar-
rative with a multimedia quality, as Patricia Joan Sarro and
James Doyle write in their catalogue essay “Monumental
Imaginings in Mesoamerican Architectural Models.” Given
the number of typologies embodied in the models, rather than
depicting a single outcome, multiple media and scales operate
in the models simultaneously, mottling a spectrum of possible
messages. One could argue, for example, that various “figu-
rines” encrusted or embedded in the “scenes” of the models,
as in several from Peru and Mexico, are evidence of an inter-
est in multimedia methods. In other examples, models are
converted into vessels or even whistles – the latter something
that seems to be specific to Andean designers.
The concept of mottling can be expanded beyond the
material finish or quality to a broader framework in which
the model also serves as a multimedia cultural artifact that
engages multiple dimensions and times. The alignment of
these objects with ancient ritual practices, philosophies, and
religions is profound, and particularly acute in indigenous

141 Log 50
University of Virginia Campus,
Charlottesville, with the Jefferson
Rotunda in the background and the
Memorial to Enslaved Laborers in the
foreground. Photo: Sanjay Suchak.
Courtesy UVA Communications.

cultures of the Americas, where the temporality, circular-

ity, and multivalence of inanimate objects have a distinctly
different subjectivity and politics than their inert, inorganic
counterparts in modern Western civilization. In these ancient
cultures, the model was a form of writing, of myth, narrative,
and active engagement with multiple futures and pasts, and
these qualities are still evident 2,000 years later. The nota-
ble attentiveness of nonarchitectural audiences to “Design
for Eternity” was important, Pillsbury said, as it suggested
a degree of inclusivity, the capacity of the models to spark
the imagination – perhaps something less tenable for today’s
architectural models, which have intentionally limited capac-
ities, are without material qualities, and generally focus solely
on their relation to a future building.

Robotic Handmadeness
Jump ahead several hundred years and one might find similar
behavioral attributes in the recently completed Memorial to
Enslaved Laborers at the University of Virginia, designed by
the interdisciplinary team of Höweler + Yoon Architecture,
Dr. Mabel O. Wilson, Gregg Bleam, Dr. Frank Dukes, and Eto
Otitigbe – an architecture firm, a historian and designer, a
landscape architect, a community facilitator, and a polymedia
artist. The project entailed a process of prototyping at full
scale that is explicitly postdigital while associating methods of
storytelling that oscillate between the abstract and the literal,
between the handcrafted and the computational, with the
politics of social justice and the labor practices of architecture
and construction.
Thomas Jefferson’s master plan and design for UVA for-
mally, geometrically, and materially embodied postcolonial

142 Log 50
Höweler + Yoon Architecture in col-
laboration with Dr. Mabel O. Wilson,
Gregg Bleam, Dr. Frank Dukes, and
Eto Otitigbe, Memorial to Enslaved
Laborers, University of Virginia,
Charlottesville, 2020. Photo courtesy
the architects.

early democratic ideals of the US and served as the template

for many aspirational university campuses across the country.
As is well-known, UVA is also a site of hardship, inequity, and
enslavement, the latter being instrumental in the labor and
building of the institution and its architecture, a narrative
now made visible and commemorated through the memorial.
As a design artifact, the memorial embodies some of the
same qualities as the models in “Design for Eternity.” Its pres-
ence is delicate, weathered, almost as if a fossil encrusted with
soil has been partially excavated in an archaeological dig. A
partially submerged, circular pathway is lined with a retaining
wall that is etched with the names of the enslaved laborers who
built UVA. The exterior facing wall includes a partial image of
Isabella Gibbons, a former enslaved laborer who, upon eman-
cipation, became an educator, and reinforces a sense of history
by registering subtle asymmetries between the implied interior
and exterior of the memorial – between its differing optical
scales, between the temporal durations in these zones, and in
the relation of the memorial to its surrounding environment.
The memorial vibrates between the visible and the invis-
ible in an almost haunting, apparitional manner. It envelops
and entombs the visitor, swelling up like a cresting wave as
one ambulates the circumference of the circular plan. As in
the ancient models, the entrancing and temporal affects of
the design can be attributed to the material sensibility, which
is mottled. It is loud and quiet at the same time, present and
barely there, object and landscape, floating and simultane-
ously bound to the topography of the site.
The multimedia aspects of the memorial augment its
temporal qualities. The stone elements, the primary material
of the memorial, were prefabricated – carved, curved, and

143 Log 50
 etched – using computationally based robotic techniques. The
stone is a single monolithic material, but its deployment is
mottled. The V-etched line matrix “transferring” and “image
mapping” of Otitigbe’s depiction of Gibbons’s eyes is in direct
contrast to the parabolic scallops, which are in contrast to the
roughly diagonal veining of the marble.
The design sensibility introduces “noise,” mottling
materials, patterns, and design moves. It is an architectural
murmur, not quite static, but also not wholly legible. Each
textural system is “rendered” in varying shades of gray, slid-
ing and reverberating off one another, imbuing the overall
assembly with an artificially induced, handmade quality.

Mottled Behaviors
Mottled architecture and mottled models are not entirely new
ideas in architectural discourse. Threads of both were visible in
Eto Otitigbe, the eyes of Isabella the mid-1980s and early ’90s in the infamous pigment-Bondo-
Gibbons V-etched on the memorial’s spackle-covered models of the Los Angeles firm Morphosis
exterior wall. Photo: Alan Karchmer.
Courtesy Höweler + Yoon Architecture.
and in the use of stucco, patinated copper, or COR-TEN steel
among many notable Los Angeles projects from the same
period. But those effects were generally representational and/
or passive, the result of the actions of other forces. “Design for
Eternity” and the Memorial to Enslaved Laborers expand the
possibilities for mottling to engage the multiple temporalities
and mediums of contemporary design and the vicissitudes of
its unpredictability.
The introduction of multiple narratives, of multiple effects
imbricated into the materiality of a scale model or prototype, is
ultimately an act of nonbinary behavior. This inclusivity brings
diverse oppositions into one space, allowing for the noise, the
messiness, and the stain. Mottle behavior is a vague proposal
of behaviors that may allow us to draw upon the history of the
discipline and its origins to understand the temporal, mate-
rial agencies we have at our disposal in the development of
scale models and built work. In this way, unforeseen behaviors
might emerge in the things we make and move against the ste-
reotype of the model as an inert replica, ultimately allowing us
to see the value of mottle behavior within the frame of our cur-
rent crises and beyond.

David Erdman is the chair of gradu-

ate architecture and urban design at
Pratt Institute, author of Introducing
(2020), and cofounding principal of
davidclovers and servo.

144
John McMorrough
Not a Role, Model

To anyone following the recent play-by-play of collective

imaginaries, architecture has dropped the ball. The outbreak
of the COVID-19 pandemic shows that architecture puts us
too close together. The Black Lives Matter movement reveals
that it keeps us too far apart. Economically exorbitant, politi-
cally ambivalent, environmentally destructive, and socially
exploitative, architecture appears to be a solution that only
produces more problems. Despite its utopian vocation, or per-
haps because of it, architecture has perennially failed in its
presumptive totalities large and small (society and its build-
ings). With a long-suffering fan’s weariness, the avid follower
of the field has to ask how long it has been since team archi-
tecture has had a win.
Given the permeating sense of failure, the consensus is
that architecture needs to be rethought, old relations severed,
and new connections made. But with so much wrong, where
is the right place to begin? In the vast array of possible avenues
of reconstitution (its definition, its process, its history) and
the variety of constituencies (who pays, who builds, who
dwells), it seems appropriate to take a moment to examine
the conception of the “architect” as the problematic singular
embodiment of volition. It’s not the most critical issue fac-
ing the world, or even architecture, but in the era of social
media, where every whisper is into a megaphone, the personal
is political, and the political is taken personally, the architect
appears as the public face of a broken promise. It is the identity
of the architect as the embodiment of hierarchy that seems
most out of step with contemporary codes of behavior. It is
not just what architects do, or how, but rather who they are,
or at least who they are perceived to be. What follows is a brief
sketch of possible ways of seeing the architect. As a model for
such modeling, a detour into another field’s anecdotal history
offers a different public debate about the relation of character-
ization and performativity from a relatively more straightfor-
ward period of (social) media signification.
In 1992, the Cold War was over, Los Angeles was on fire,
and what everyone wanted was to “be like Mike.” In a com-
mercial for Gatorade, Michael Jordan appeared in a series

145
of clips that intercut game footage of his famously athletic
leaps, suspended midair as he dunked a basketball, with video
of him in a casual playground game with a group of chil-
dren and then merrily practicing and drinking Gatorade with
his “teammates,” showing him simultaneously as a man of
accomplishment, a man of the people, and a man among men.
All the while, the commercial’s jingle described how one could
move and groove, achieving a state of grace only “if I could be
like Mike.” At that time, Jordan was described, perhaps with-
out hyperbole, as the greatest basketball player in the world.
With numerous national championships and corporate spon-
sorships, including his namesake Air Jordan global sneaker
brand for Nike, Jordan was the personification of excellence
in athletics and image. A player on both court and screen,
the superstar could only allow for the most abstract perfec-
tion of persona. But while his athletic record remains stellar,
with time, perceptions of Jordan’s totality have become more
modulated (as seen in ESPN’s 2020 The Last Dance documen-
tary miniseries). He could be unsupportive of his teammates
(as in Scottie Pippen’s contract negotiations), he was driven
by animus (winning games to enact personal grudges), he was
a gambler of some notoriety, and he was not a good baseball
player (relatively). Perhaps unsurprisingly, Jordan was more
and less than his public image.
The architectural analogy for Michael Jordan is the
heroic figure of the architect as a singular genius. The spe-
cific source of genius is still a matter of some debate: divine
intervention talent, intense diligence (or practice, at least
10,000 hours), or, more likely, some combination, with a lib-
eral application of inheritance (from trust funds or trusted
sources of institutional, political, and/or media support). But
the descriptions of heroic genius, regardless of their origins,
are troubling for what they leave out. From Giorgio Vasari’s
The Lives of the Most Excellent Painters, Sculptors, and Architects
to Dezeen, such glosses have never depicted reality’s fullness.
As anyone who has stepped into an architecture office (or
onto a basketball court) realizes, “teamwork makes the dream
work.” The pinnacle of public greatness rests on a founda-
tion of others’ toil. It is not so much the lack of shared credit
that is at issue (at least not entirely); rather, it is the cult of
genius itself, which implies hierarchy (designs are deter-
mined from the top down), enforces rank (design choices by
a select few are implemented by others), and is not egalitarian
(firm members, clients, users, citizens, and other stakehold-
ers do not all have an equal opportunity to inform designs).

146 Log 50
Frank Lloyd Wright to Le Corbusier, I.M. Pei to Zaha Hadid,
Jeanne Gang to David Adjaye; the names change, but a firm’s
figurehead is an idol to be idolized, a cult of personality of
which the “starchitect” is only the most recent manifestation.
Following Jordan’s example, it is not just the figure’s singu-
larity that is problematic but also the false projection based on
a two-dimensional representation. The development of this
Potemkin personification is an issue in both architecture and
basketball because it reduces the totality (of the project, of the
person) to a marketing claim.
Charles Barkley was both a contrast and countermodel to
Michael Jordan. Though also a remarkable basketball player,
his style of play and persona could not have been more differ-
ent. Where Jordan was aerial, Barkley was grounded, known
for his aggressive play and rebounding. Where Jordan was
often smiling but silent (when asked why he didn’t speak out
more about the issues of the day, he opined, “Republicans buy
sneakers too”), Barkley could be scathing in his observations,
always ready with an outrageous line: “These are my new
shoes. They’re good shoes. They won’t make you rich like me,
they won’t make you rebound like me, and they definitely
won’t make you handsome like me. They’ll only make you
have shoes like me. That’s it.” Following, and seemingly in
direct response to, Jordan’s Gatorade commercial, Barkley’s
commercial for Nike played as a rebuttal in the court of pub-
lic opinion as much as in basketball. Shot in stark black and
white, Barkley, a visage of sweaty effort, is seen practicing
on a basketball court alone; the commercial starts with him
bluntly stating, “I am not a role model,” followed by, “Just
because I dunk a basketball doesn’t mean I should raise your
kids.” The explicit point was that parents should be the role
models; more implicit was that the socially performative
expectations of the player are not part of the deal (the player’s
contract does not extend to the social contract). However,
while Barkley’s wholesale rejection of social performance as
the cost of admission to the public sphere is a form of rebel-
lion, this too is a marketing strategy.
In his avowed rejection of being a “role model,” Barkley
suggested that excellence in one domain does not necessarily
translate to others. In architecture, this would compare to the
bracketing of discrete arenas of relative clarity (geometry,
precedent, construction) from the messy difficulty of the het-
eronomous (such as the social and political). In a line that runs
from l’art pour l’art to autonomy, from aestheticism to deter-
minism, such a limitation of claim (and duty) acts as a fire wall

147 Log 50
between the flames of expectation and the safety of a precisely
delimited zone of competency. It’s a claim that benefits from its
categorical clarity. The gain in one arena (ability) does not nec-
essarily translate to another (morality), but the converse can-
not be fully dismissed, as it must be admitted that the singular
agent operates in an extended field of activities and actors that
underwrite each performance, whether aesthetic or athletic.
The implicit debate between Jordan and Barkley and the
conundrum of acquiescence to, or disregard for, the expecta-
tions of image were addressed subsequently by a third basket-
ball player, Karl Malone, in his 1993 opinion piece “One Role
Model to Another” in Sports Illustrated. Rather than a generic
signification of Jordan or a wholesale repudiation of Barkley,
in Malone’s rendition, performance and performativity are
bound, because the exceptional basketball player, whether out
of privilege (acquired wealth) or position (fame), has a moral
responsibility to act knowing that “kids and even some adults
are watching.” It is a reasonable argument and expectation,
and although, as Malone qualifies, “people shouldn’t expect
perfection,” the idealism of exemplarity remains as a latent
expectation (as Malone would find out in late career scandals
stemming from earlier lapses).
If Jordan represents the architect as a fountainhead of
singular exemplarity (be like ____) and Barkley repre-
sents the architect’s limited liability as a technician of spe-
cific expertise (I am not a role model. I am not paid to be a
role model), then Malone represents an architect’s moral
duty as one as a responsibility of privilege (noblesse oblige).
Overcoming the limits of the previous two, this third posi-
tion aspires to be beyond reproach, which is not to say the
previous two positions are not still present, even operative, in
the final formulation. Following Malone, one is called to be a
role model by the de facto extrapolation of “the glory and the
money that comes with being a famous” …architect. Here the
analogy fails, and the specifics of practices must be reintro-
duced. An architect and a basketball player may share many
things, but outrageous financial opportunity and popular
visibility of the position are not among them. The problem
to address, the problematic aspect of the current definition
of roles, is to understand the relation between the subject of
architecture and the subject position of the architect.
In the envisioning of a new game, let us first recall the
rules and roles of the previous one. Good intentions don’t
make for better play, on or off the court. Depending on the
game, some players are better than others. Regardless of how

148 Log 50
 deep the roster of the team, at some point someone with the
ball has to take the shot. To question the game is to understand
each moment and movement, not as just one role but many:
of performance, of representation, and of their mediation.
Whatever the freedom of games is, it occurs within a specified
field. Speculation, whether in the market or the gallery, comes
with associated limits and costs. A correction of the problem
of the “architect,” which proposes the substitution of one
player with another, misses the more fundamental crisis vis-
à-vis architecture’s relation to authority. As one is replaced
with another, what are the conditional limitations of any
occupation? Architecture is subject matter (a topic, a scale, a
practice, and a discipline), but not a subject (a person); it is
more like a subject position (a vector of images, responsibili-
ties, necessities, and freedoms). What the fans and philoso-
phers get so wrong is the understanding of the architect as the
embodiment of absolute control, and their subsequent rejec-
tion of architecture stems from never getting over the dream
of realizing perfect intentions (geometric and otherwise).
The long-standing crisis of architecture is the inevitability of
its failure on those unrealistic terms.
The contemporary (and historical) problems for archi-
tecture are existential issues – namely, how or what should
be done at a given juncture amid the manifest availabilities
and scarcities. It is not a matter of turning on architecture’s
relevance because it was never turned off. Architecture works
through its own relation to power (of societies and econo-
mies) and its connections to authority (institutional, edu-
cational, and legal) in both affirmation and negation. The
questions that architecture has faced and still faces are pre-
cisely the issues that society now faces: What is the relation
between performance and representation? What is the connec-
tion between form and formation? How is something made,
how is it contained, and how can it supersede these limits? To
engage architecture’s problematic disciplinary history is to
understand the availability and limitations of agency unfold-
ing – historically, economically, and conceptually. The archi-
tectural discipline, through the means at its disposal, works to
understand authority’s reflective limits and to come to grips
with complicity, which occurs in every plan conceived, in
John McMorrough writes about the
relationship between culture and every building built. In building (or rebuilding) the world,
design, with a focus on architecture’s the architect is not a role model, but in navigating multiplici-
extended field. He is associate profes-
ties (of agendas) and constraints (of material), architecture
sor of architecture at the University of
Michigan and a partner of studioAPT is, was, and continues to be the model.
(Architecture Project Theory).

149
Observations on Natural Models

Architect-engineer Frei Otto’s practices reflect concerns that are “Models, Media, and Methods: Frei
“pursuit of material and central in today’s climate crisis. Otto’s Architectural Research,” Yale
School of Architecture Gallery, New
resource efficiency . . . prefigured Otto found inspiration in Haven, February 20–May 2, 2020.
many of the tenets that now fall natural configurations and used Photo: Richard House Photography.
under the umbrella of sustain- combinations of natural and Courtesy Yale School of Architecture.

ability,” writes Deborah Berke artificial materials to make

in the preface to the catalogue of ephemeral and dynamic models. flexible column, reveal the
“Models, Media, and Methods: On one table, a photograph of influence of these models on his
Frei Otto’s Architectural a spiderweb accompanied a design work. More than find-
Research,” an exhibition pavilion model made of fabric ing form based on material and
that opened at the Yale School and wire. On others, images of structural properties, through
of Architecture Gallery in sandpile and soap bubble studies these models, Otto developed sys-
February 2020. Curated by theo- were displayed alongside the tems based on natural processes;
rist of architecture and technol- architectural models they beyond producing particular
ogy Georg Vrachliotis, the show informed. As illustrated in the solutions, these experiments
featured several physical models publications of the research engendered a greater under-
and many black-and-white institutes he led, Otto conducted standing of nature. As archi-
photographs of representational, a wide variety of physical tecture continues to rethink
structural, and form-finding experiments to study natural approaches to the environment
models. While some were famil- elements, attempting to simulate today, fostering new adapt-
iar, such as the suspended chain the growth or structure of trees able systemic strategies, there
models used to shape projects like or bones with water, plaster, and is clearly more to learn from
the 1975 Mannheim Multihalle, nets, for example. Unrealized Otto’s natural models.
Otto’s lesser-known modeling projects, such as a spinelike  – Anna Renken

150
Christian Hubert
Model Behavior?

Several grammatical functions are embedded in the title

“model behavior.” In one, the term model functions as a verb –
to model, to predict, or to enable behavior. Another function
is adjectival, whereby model behavior means exemplary behav-
ior, but the ambiguity of the phrase raises the question of how
to enable exemplary behavior. In a third function, model is a
simple noun and its behavior a topic of inquiry.
I have previously explored the expressive registers of
architectural models, especially their dual functions as rep-
resentations and actual objects, along with the desires they
1.  See Christian Hubert, “The Ruins of embody, and some of their underlying ideological functions.1
Representation,” in Idea as Model, ed.
Kenneth Frampton and Silvia Kolbowski
One consequence of the digitization of the practice of archi-
(New York: Rizzoli International, 1981), tecture that has occurred in the interim has been to change
17–27. See also Christian Hubert, “‘The Ruins
of Representation’ Revisited,” OASE 84: the defining features and ontological status of the model.
Models (2011): 11–25. Although the model’s physicality previously enabled it to func-
tion as both object and representation, that physicality has
waned in importance today. Whether it is milled, rendered,
or fabricated through aggregation, a 3D model produced from
a digital file remains primarily an expression of its digital
instructions. It can effectively persuade, or its verisimilitude
can be questioned, but one source of its aesthetic criticality
has been canceled. It doesn’t really matter if the model is a
physical object. Today, it is the information and the scenarios
expressed through the model that have become primary, and
the goal of modeling a process has taken the place of modeling
a product. As a result, the model has potentially become an
agent in the process itself. Modeling has become performative.
One example of materially performative modeling is the
Urbach Tower, designed by a team from the University of
Stuttgart and erected in 2019 in the Rems Valley in Germany.
The project mobilizes properties of the specific material it
is made from to create shapes that form themselves. It relies
on the deformation of 12 cross-laminated timbers as a direct
effect of industrial drying processes on their initial flat-cut
shapes. Those deformations can be visualized beforehand in
3D computer models that integrate material properties into
the resultant forms. Here, the model is predictive (of the end
form), performative (as a set of instructions for milling), and

151
 testable (through the results of the actual fabrication process).
But in this case, a scale physical model would have been
irrelevant, as it could not have embodied the actual form-
giving process.

A Model for the Anthropocene

A defining feature of our time is the growing consensus among
scientists to call the current geological epoch the Anthropocene,
by virtue of the significant human impact on Earth’s geology
and ecosystems. Its distinguishing features include anthro-
pogenic climate change and loss of biodiversity through a
University of Stuttgart Institute “biological annihilation” – a sixth mass extinction event –
for Computational Design and that is currently well underway. In this sense, what is model
Construction (ICD), Urbach Tower,
Rems Valley, Germany, 2019. Photo
behavior? The concept of the Anthropocene gives humans
© ICD/ITKE University of Stuttgart. global roles and responsibilities. On the one hand, it recog-
nizes the defining role of humans as forces of nature. On
the other, it implicates them in processes that they do not
fully understand and seem unable control, in which natural
and human-made environmental calamities converge. The
future geological record of the Anthropocene will presum-
ably include fossilized seabirds with stomachs full of plastic,
as well as plastiglomerates – stones fused with plastic found on
beaches. As global “apex predators,” humans themselves are
at risk from the environmental stresses they have caused or
exacerbated, such as rising sea levels, the fragility of human
food chains, water shortages, and the possibilities of global
pandemics. In this sense, “model behavior” is not a metaphor.
It is an obligation.

Modeling Tomorrow Today

Most contemporary modeling is embedded in scenarios for
the future, and the new hope for the model is for it to be
instrumental in realizing them, to promote a reflexive real-
ity informed by hopes and fears that motivate its dynamic
force. In this sense, the function of performative models is
to provide “proof of concept.” They not only point to the
future; their purpose is to build new worlds, with the full
awareness that those will not necessarily be better worlds
unless humans become stewards of life. The evolution of
increasingly accurate computer models to predict and plan
for climate change cannot be undervalued. While some of
the fundamental forces at work in climate change are simple
and well understood, integrating a host of new data, feed-
back processes, and complex behaviors at many scales is a
demanding and ongoing process – one that many humans

152 Log 50
 are unwilling to trust and more often find reason to deny.
Modeling takes place today in an expanded cultural and sci-
entific context, with new biotechnical possibilities for visual-
ization, prediction, and implementation. In the global context
of life on Earth, the dreams of new worlds seem to coin-
cide with the imminent collapse of life on the old one. In the
Anthropocene, it is essential to consider model behavior as a
form of repair – of repairs to Earth and its forms of life.
Model behavior requires an explicit ecological and social
agenda in which life in every form is understood as process
and humans, their technologies, and the planet are all stake-
Kelly Jazvac, Plastiglomerates, 2013. holders. This agenda requires human inquiry and design to
Plastic debris and beach sediments, work symbiotically with other agents – other processes or
7 1/2 by 6 1/2 by 3 inches. Photo cour-
tesy the artist.
other species – to learn from symbiotic organisms like lichens,
slime molds, etc. – complex life-forms that move beyond con-
cepts of the individual or a population. The current task of
models is to incorporate those parallel strains, to embody their
potentials, and to function as “models” in the sense of exem-
plars. The role of humans in this sort of work is more akin to
gardening than manufacturing, which poses a question: Can
architectural models be performative in this sense?
A new appreciation for the creative capacities of other
life-forms is evident in symbiotic projects that integrate
other species into design and building processes, such as Neri
Oxman and the Mediated Matter group’s work with silkworms,
as well as other projects that help us move away from think-
ing of the world as made just for humans and toward an
interrelated web of life.
Examples of those interrelated webs are the topic of
a recent book by Merlin Sheldrake, titled Entangled Life:
How Fungi Make Our Worlds, Change Our Minds & Shape Our
2.  See Merlin Sheldrake, Entangled Life: Futures.2 The growth and transformation of the World Wide
How Fungi Make Our Worlds, Change Our
Minds & Shape Our Futures (New York:
Web is the metaphorical model for Sheldrake’s evocations of
Random House, 2020). the “Wood Wide Web,” the ubiquitous underground networks
between fungi and trees that support their exchanges of
nutrients and information. If the development of the World
Wide Web was the result of technical and mathematical devel-
opments that ultimately produced a transformative human
artifact, the very concept of the Wood Wide Web illustrates
how a natural “model behavior” provides a productive way
of grasping ideas. These conceptual transformations enable us
to see mycelia as neural networks that occupy the intersection
of biology, technology, and critical thought. Gilles Deleuze
and Félix Guattari developed tools for exploring some of
this philosophical territory in their writings on the rhizome

153 Log 50
Merlin Sheldrake’s book, Entangled some 45 years ago. Today, the life of this concept is much
Life, seeded with spores, September more explicitly tied to biological research in the growth and
4, 2020. Right: Studio Klarenbeek
& Dros, Mycelium Chair (Mycelium
working of mycelial networks and to psychedelic experi-
Project), 2013. Photo courtesy the ences, including the sense of being one with nature and the
designers. feeling of cocreatureliness. But cocreatureliness needs to be
more than a feeling or projection. In the Anthropocene, it
needs to provide a model for living.
Sheldrake’s particular metaphorical transfer brings the
idea of mycelial networks to life. As he lightheartedly points
out, some objectionable human thought patterns may also
carry over into this view of the world. Seeing fungi as con-
nectors between plants may be a form of “plant-centrism.”
To illustrate the ways that fungi “make worlds” and “also
unmake them,” Sheldrake announced his intent to seed a copy
of his book with Pleurotus mycelium, let it eat its way through
the words on paper, and then to eat the oyster mushrooms
3.  Ibid., 160, 225. himself, thus “eating his words.”3

Architecture as Symbiosis
How do the ideas expressed in Sheldrake’s book function as
a model for architecture? Frei Otto’s Ökohaus (Eco-house)
in Berlin, built as part of the 1987 International Building
Exhibition, embodies a number of qualities that qualify it as
a model in our sense. It was designed and built by and for an
ad hoc, self-selected, informal population, with particular
attention to incorporating existing and projected plant life.
Otto designed two reinforced concrete structures that allowed
separate infill construction a measure of independence and
freedom of material. In fact, the upper stories of the concrete
armature were infilled first, so that the lower levels could be

154 Log 50
Ophiocordyceps unilateralis, also
known as zombie ant. From Maj-
Britt Pontoppidan et al., “Graveyards
on the Move: The Spatio-Temporal
Distribution of Dead Ophiocordyceps-
Infected Ants,” PLoS ONE 4, no. 3
(March 2009).

built out in response. The design process and its outcomes,

like the growth of mycelium, integrated the most local deci-
sions into a loosely defined whole, in symbiosis with other
life-forms – especially plant life. Here the architect indeed
played the role of gardener.
In Entangled Life, Sheldrake points to ways that we might
partner with fungi to help us to adapt to life on a damaged
planet. For designers such as Eric Klarenbeek, for example,
mycelium is both a figure and a potential building material,
as seen in his design for a chair with fungi continuing to grow
out from it. It is in these senses that mycelium can function as
a model.
In the late 1990s, a disquieting image was exchanged
through e-mails. It was an early internet meme: an idea that
takes hold of unsuspecting minds. The Vacanti mouse, also
known as the “earmouse,” circa 1996, had what looked like a
human ear on its back. The “ear” was actually an ear-shaped
cartilage structure, grown by seeding cow cartilage cells in a
biodegradable, ear-shaped mold, implanted under the skin
of the mouse. Over time, the mouse grew extra blood vessels
that supplied nourishment to the cow cartilage cells. The
mouse itself, known as a “nude mouse,” was a laboratory
strain used in cancer research, with a compromised immune
system. The picture prompted a wave of protests against
genetic engineering – although in this specific experiment
no genetic manipulation was performed. An antigenetics
group placed a full-page ad in the New York Times, with the
caption, “This is an actual photo of a genetically engineered
mouse with a human ear on its back.” This was totally untrue.
The cartilaginous structure that looked like a human ear

155 Log 50
 could never have been transplanted onto a human because it
was full of cow cells and would have been rejected by a per-
son’s immune system. Nonetheless, the earmouse constituted
a step forward in the development of tissue technology. The
goal of this line of research is to be able to repair or replace
human tissue directly. The earmouse functioned as a contro-
versial model for design and fabrication along the way.
Similarly, Sheldrake describes the workings of the so-
called zombie ant fungus in the same pages as psilocybin’s “fre-
Vacanti mouse, also known as earmouse. quently inexplicable and indeed uncanny” effects on humans.4
Ophiocordyceps unilateralis is known for its parasitism on car-
penter ants that normally live in the high tree canopy, where
they have an extensive network of aerial trails. Sometimes the
canopy gaps are too wide to cross, so the trails descend to the
forest floor where the ants can be exposed to the spores. The
process, which ultimately leads to the death of the ant and the
reproduction of the fungus, includes a stage in which fruit-
ing bodies grow from the ant’s head, eventually rupturing to
release the fungus’s spores. The fungus’s skill at colonizing the
ants is surpassed only by its skill at colonizing popular culture.
4.  Richard Evans Schultes and Albert The Ophiocordyceps is often referred to as the zombie ant fun-
Hofmann, Plants of the Gods: Origins of
Hallucinogenic Use (New York: Alfred van
gus, and its sinister form of mind control resonates with the
der Marck Editions, 1987), 9. fear of the “living dead.”
While the two biological examples may function more
as metaphors for designers than models, all three examples
model different worlds that have the power to shape the
human imagination. The earmouse is an affirmation of
human technique and the manipulation of life. It seems to
introduce a world in which human technologies use other
species for “spare parts.” The zombie fungus, itself a com-
pound organism, hijacks another species for its own ends.
These three models actively stand in for worlds in which tra-
ditional boundaries between the human and nonhuman, or
between machines and organisms, have become porous. They
do not simply represent these worlds. These models are steps
along the way. If the Anthropocene period starts with recog-
nizing the power of human interventions on the Earth, the
concept of the Anthropocene also needs to recognize the vul-
nerabilities of all living species, including humans, to those
effects. The models illustrated here are meant to serve as
guideposts, both positive and negative, along the way.

Christian Hubert is an architect and

writer who lives and works in New York.

156
Kiel Moe
Architectural
Agnotology & Broken
World Models
Our collective model of architectural education and practice is
a shared form of climate change denial. For some, that denial
manifests in positions and pedagogical models that main-
tain architecture as independent, autonomous from changing
climates and other directly related environmental and social
dynamics. In such cases, the denial is active; the realities of
changing climates and social inequity are known but do not
substantively influence models of architectural education
or practice. Further, protagonists of this denial may actively
work to exclude or suppress positions and models of archi-
tecture that can more directly address our changing climates.
Another example of active denial, perhaps somewhat rarer
but still present among architects, includes the position that
debates the veracity of climate change science. Obdurate dis-
belief about the causes of climate change serves as a tactical
bypass to increasingly pervasive climate realities. This denial
is active through its engagement in misinformation and its
subsequent obstruction of practices and policies that might
better address our changing climates. In both cases, the active
denial is one of studied and cultivated avoidance.
A third genre of denial is curiously passive. Here, the
realities of climate change are accepted en tout, yet the protag-
onists act in ways that deny what climate science actually indi-
cates. This includes designers who aim to slow or dampen the
effects of climate change. While perhaps well intended, such a
mindset nonetheless denies the science and political science of
changing climates because the window for mitigation closed
before any of us were born. Decisions, policies, and infrastruc-
tures regarding how hydrocarbon capital and its contradic-
tions structure nearly every aspect of contemporary life were
determined several decades ago in the developmental phases of
modernity. We live, and design, in the aftermath of those accu-
mulative decisions that have covered the world over with per-
vasive hydrocarbon emissions and saturated our ocean gyres
and our landscapes with petroleum-derived microplastics.

157
 “Sustainability” and “resilience” praxes are prime examples of
this passive genre of denial in architecture. More recently, pro-
ponents of the Green New Deal have considerably expanded
the scope of this passive denial by ignoring the immense car-
bon emissions and social inequity that would be required for
the scale of action implied by H.R. 109. The paradoxical mix-
ture of good intentions and de facto acceptance of climate
change – but not acceptance of its actual impacts or implica-
tions – results in a diabolically passive form of denial.
I borrow the above categories of active and passive denial
from a framework that history of science scholar Robert N.
1.  Robert N. Proctor, “Agnotology: A Proctor developed in his essay “Agnotology: A Missing Term.”1
Missing Term to Describe the Cultural
Production of Ignorance (and Its Study),”
Agnotology is the study of ignorance; more specifically, it is
in  Agnotology: The Making & Unmaking of the study of the cultural production of ignorance and how
Ignorance, ed. Robert N. Proctor and Londa
Schiebinger (Stanford: Stanford University our models preclude us from knowing what we could other-
Press, 2008), 1–33. wise know. A canonical example of an agnotological query is
2.  Daniel R. DeNicola, Understanding
Ignorance: The Surprising Impact of What how the tobacco industry actively cultivated ignorance about
We Don’t Know (Cambridge: MIT Press, the carcinogenic effects of smoking in the 20th century. In
2017), 84.
3.  Ibid., 82.
turn, we might ask, how did architecture actively cultivate
4.  Ibid., 81–82. ignorance about the climatic effects of building and urbaniza-
tion in the 20th century? How did a profession native to the
processes of hydrocarbon urbanization – putatively justi-
fied through its ethics of health, safety, and welfare – miss
its climatic capabilities and culpabilities as a discipline? How
did our models of architecture come to sanction its climate
change–inducing pedagogies and practices?
Other work on agnotology helps to refine and address
these questions. The active forms of denial identified above
would, in the work of philosopher Daniel R. DeNicola, fall
into several nuanced agnotological subcategories. Someone
engaged in willful ignorance, in DeNicola’s view, “persistently
ignores the topic despite its likely salience and even resists
learning about it or assimilating facts that bear on it.”2 This
would include architects who maintain that architecture is
autonomous from environmental and social phenomena such
as changing climates. Strategic ignorance, on the other hand,
applies to those who actively deny the veracity of climate
change science. This category, DeNicola notes, is “calculated,
but in this case the intent is to use ignorance as an advantage.”3
Rational ignorance applies to the passive denials of “sustain-
ability” and “decarbonization” advocates in both descriptive
and normative ways. “The descriptive sense,” DeNicola
observes, “includes the possibility that we may mistakenly or
imprudently choose ignorance; the normative implies that we
have a plausible justification.”4 Accordingly, individuals in this

158 Log 50
 genre are selective about what forms of facts and knowledge
they include and exclude from discourse and practice.
Architectural ignorance about changing climates is not
outright problematic. An “ignorance-based worldview,” as
agriculturist Wes Jackson describes, is not only inevitable
but perhaps epistemically prudent for a topic as complex as
5.  Wes Jackson, “Towards an Ignorance- changing climates.5 That is, admitting one’s necessarily lim-
Based Worldview,” in The Virtues of
Ignorance: Complexity, Sustainability, and the
ited grasp of an immensely complex topic is not an alibi for
Limits of Knowledge, ed. Bill Vitek and Wes ignoring the topic but rather its best point of epistemic entry.
Jackson (Lexington: University Press of
Kentucky, 2008), 21–36. Operationalizing one’s ignorance is an important methodolog-
6.  For issues of environmental load dis- ical step forward. Only then can a more intricate agnotologi-
placement, see Alf Hornborg, “Towards an
Ecological Theory of Unequal Exchange: cal inquiry about climate change in architecture fully reveal
Articulating World System Theory the limitations of our current models of architectural peda-
and Ecological Economics,” Ecological
Economics 25, no. 1 (April 1998): 127–36; gogy and practice. Such agnotological inquiry helps reveal,
for economic and ecological exchange, for instance, that our present models of pedagogy and prac-
see Arghiri Emmanuel, Unequal Exchange:
A Study of the Imperialism of Trade, trans. tice have long relied on systemic forms of environmental load
Brian Pearce (New York: Monthly Review displacement, unequal economic and ecological exchange, and
Press, 1972); for processes of underdevelop-
ment, see Stephen G. Bunker, “Energy processes of underdevelopment.6 The political ecology of these
Values in Unequal Exchange and Uneven
Development,” in Underdeveloping the
processes are ardently externalized from the discipline, but
Amazon: Extraction, Unequal Exchange, and they nevertheless underlie our present model. In this agnotol-
the Failure of the Modern State (Chicago:
University of Chicago Press, 1988), 20–57.
ogy, it becomes increasingly clear that our technical and for-
mal endeavors have long fetishized architectural objects while
actively suppressing and externalizing their systemic terrestrial
processes and feedbacks. When considered alongside architec-
ture’s other immense externalizations – not the least of which
is systemic racism – the abject conundrums and privations
of architecture’s object-oriented model become increasingly
impossible to maintain or justify. As such, this agnotological
inquiry into architecture’s relationship with changing climates
is not merely retrospective but also reflexive and projective,
revealing how the current model might evolve and adapt.
A persistent and pernicious feature of our collective
models of architecture in the past several decades has been
an assumption of stability. This stability engendered a formal
model of autonomy on the one hand and a technomanage-
rial model of putatively sustainable autarky on the other. In
the first case, an Anglo-American reading of an exclusively
European canon stabilized the terms of a discourse on the
autonomy of architecture. The narrowness of this canon
granted its relative fantastical stability. Variants and progeny
of that autonomy agenda persist today in and outside of the
academy. In all cases, the associated protagonists either ignore
or outright reject social and environmental dynamics, which
are deemed a dilution, if not a destabilizing threat, to archi-
tecture as an autonomous discipline. The accumulated social

159 Log 50
 and environmental externalities of this restricted formal
praxis left other architects ample room to develop social and
environmental positions most often at odds with the central
formal propositions of design. As such, most environmental
concerns regarding building were too often treated as a dis-
mally technocratic issues, involving little more than the fuel
management of buildings: first “energy conservation,” then
“energy efficiency,” and more recently “net-zero energy.”
In this autarkic model, an implausibly narrow praxis then
became the basis of claims about a larger, more elaborate fan-
tasy of stability: the putatively “sustainable” building.
Both the autonomy and autarky models of architecture in
this period relied on forms of stability to frame their respec-
tive purview and architects’ behaviors. Fantasies of stability
enabled a common ground upon which tribes of architects
learned to agree to disagree in both the academy and practice.
However, as the social, ecological, and, most recently, viro-
logical stability of the world erodes, these inherited models
are similarly falling apart. Global climate change, pandemics,
and the persistence of ever-vertiginous economic and sys-
temic social disparities are challenging the relevance and via-
bility of any model so tenuously premised on stability. It turns
out that architecture, like climate change, is far more inter-
sectional than its traditional tribes, topically divided depart-
ments, and power geometries intimated.
One of the telltale traits of a stability-enabled model is that
it must suppress its own inherent instabilities. To paraphrase
philosopher Paul Virilio, to invent the ship is also to invent the
7.  “When you invent the ship, you also shipwreck.7 Further, to invent the car along with myriad other
invent the shipwreck; when you invent the
plane, you also invent the plane crash; and
hydrocarbon infrastructures is not only to invent the car crash
when you invent electricity, you invent but also to trigger slower processes of environmental violence.
electrocution... Every technology carries
its own negativity, which is invented at To look past the veneer of stability in our inherited models of
the same time as technical progress.” Paul architecture is to peer through the traditional asymmetry of
Virilio, Politics of the Very Worst, ed. Sylvère
Lotringer, trans. Michael Cavaliere (New its capabilities and culpabilities and to understand architec-
York: Semiotext(e), 1999), 89. ture’s wrecks as much as its ambitious narratives.
8.  Steven J. Jackson, “Rethinking
Repair,” in Media Technologies: Essays on In response to models premised on elusive stability, an
Communication, Materiality, and Society, ed. alternative and more projective model might begin with what
Tarleton Gillespie, Pablo J. Boczkowski,
and Kirsten A. Foot (Cambridge: MIT computer scientist Steven J. Jackson calls broken world think-
Press, 2014), 221.
ing. He asks, “what happens when we take erosion, break-
down, and decay, rather than novelty, growth, and progress,
as our starting points in thinking.”8 For Jackson, this model
begins with “an appreciation of the real limits and fragility
of the worlds we inhabit – natural, social, and technologi-
cal – and a recognition that many of the stories and orders
of modernity (or whatever else we choose to call the past

160 Log 50
 two-hundred-odd years of euro-centered human history) are
in process of coming apart, perhaps to be replaced by new and
9.  Ibid. better stories and orders, but perhaps not.”9 This is a model
10.  Ibid., 222.
not based on the pseudo stabilities that enabled the modern-
11.  See David Harvey, “The Fetish of
Technology: Causes and Consequences,” ist narratives of form and technics in architecture but that
Macalester International 13: Prometheus’s
Bequest: Technology and Change (Summer
addresses the instabilities and failures increasingly seen in
2003): 3–30. public, in the news, and in nature.
Design would follow – and yield – a different model of
order, through a different set of practices, in broken world
thinking. To return to Jackson’s perspective, an alternative
model of models emerges, in part from practices of care and
repair. As Jackson writes, “Broken world thinking asserts that
breakdown, dissolution, and change, rather than innovation,
development, or design as conventionally practiced and thought
about” are constitutive of another model and thus another way
forward.10 To be clear, the aim of broken world thinking is not
to learn to “fix” that which is broken in a modern technologi-
cal object or system. As David Harvey demonstrates, a techno-
logical fix has more to do with the reentrenchment of extant
disabled systems, knowledge, and practices, much like how he
describes “spatial fixes” in geography.11 In other words, “fix-
ing” tends to only further ossify the broken system, a dynamic
that operates as an antipode to broken world thinking. Nor is
the aim of broken world thinking to learn tricks and develop
designs for how life can persist as it is now (as with sustain-
ability discourse) or for how life can easily snap back to where
it was (as with resilience discourse). Rather than persisting or
returning to the same comfortable state, a broken world model
is pragmatically attentive to next states of existence. In archi-
tecture, this invokes an anticipatory conception of architec-
tural objects that will inevitably adapt through design. Broken
world thinking is optimistically curious about the inevitability
of next states of life on this planet, which helps characterize its
genre of creative activity.
Broken world thinking inverts agnotology from a his-
torical, reflexive look at the discipline and its blind spots,
transforming similar queries into a more projective and gen-
erative activity for designers. It offers lessons on how sys-
tems and designs continuously reconfigure, reassemble, and
reconstitute themselves. Accordingly, designers develop an
appreciation for the enabling social and practical systems that
emerge in the aftermath, once things break down. The design
of maintenance, next uses, and repurposing reflects a differ-
ent account of material and social progress. This redesign of
progress – guided by what design can convivially enable, how

161 Log 50
 it can emancipate, engender solidarity, and help us evolve – is
a salient question for architecture’s model of pedagogy and
practice today. That is, the object of architecture transcends
architectural objects in broken world thinking.
Rather than clinging to inherited models of stability,
forays into contemporary building and urbanization on the
basis of a broken world model are one way to productively
interrogate, and act on, the agnotology of extant models of
architectural pedagogy and practice. This implies a much
more literal study and practice of architecture as a terrestrial
phenomenon. Against a traditional celebration of abstrac-
tion, a more literal model of architecture would inherently
afford a more inclusive description of architecture’s purview,
one that more directly reflects the actual terrestrial exchanges
and displacements that occur through the design and speci-
fication of building. As a blunt reframing of architecture’s
politics, a broken world model would systematically include
all peoples implicated in building – from miners who dig for
cobalt to janitors and other building care workers to archi-
tects themselves – all to better understand the matrix of lives
and myriad living environments that inhere through archi-
tecture. A more literal model of architecture would imag-
ine its energetics far beyond the hydrocarbon fuel fetters that
overdetermine and understate architecture’s relations to ther-
modynamics and ecology. At each disciplinary turn, a more
literal model would overturn the habits of abstraction that are
characteristic of architecture’s agnotology.
The present moment is not just about addressing changing
climates or finally changing the racist basis of the prevailing
model of architecture, although both must occur. It is about
changing the model of pedagogy and practice in ways that
overtly interrogate these, and other, unjustifiable agnotologies
of architecture. An evolved model would exercise the virtues
of ignorance studies by identifying and then internalizing cur-
rent environmental and social externalities, accordingly alter-
ing the daily habits and routines of its pedagogies and practices,
and awakening its leaders and protagonists from administra-
tive docility. Acknowledging that fundamental aspects of our
current models of practice are broken is not only a generous
and generative place to begin but central to the very methods
and practices that can raise architecture out of the habits and
routines of its failed model. Broken world thinking is not just a
model for contending with the aftermath of broken buildings
Kiel Moe is a practicing architect
and the Gerald Sheff Professor of and territories evidenced in our changing climates but also a
Architecture at McGill University. means of redefining the entire model of architecture itself.

162
Neyran Turan
Unnatural Models

For the concept of nature has a history, and the figuring of nature
as the blank and lifeless page, as that which is, as it were, always
already dead, is decidedly modern, linked perhaps to the emergence
of technological means of domination.
– Judith Butler, Bodies That Matter, 1993

Nature’s Antiquity
It was 1774. An engraving of Fingal’s Cave on the uninhab-
ited Scottish island of Staffa was published as one of the first
1.  This text reflects on some of the topics recorded descriptions of the geologic structure.1 The engrav-
discussed in my recently published book
Architecture as Measure (New York: Actar,
ing illustrates the account of naturalist Joseph Banks, who
2020). visited the island in 1772.2 With its prismatic basalt pillars
2.  The engraving is by Thomas Major,
based on an original drawing by James
covered with a rooflike lava flow, Fingal’s Cave was regarded
Miller. Published in Thomas Pennant, A as a natural cathedral – one of the most important geologi-
Tour in Scotland, and Voyage to the Hebrides,
1772, vol. 1 (Chester: John Monk, 1774), cal wonders and tourist attractions of the 19th century. What
300. Joseph Banks became the president of was particularly striking about both this engraving and Banks’s
the Royal Society of London a couple of
years later and was also the first director of account was their truly architectural character, depicting the
the Royal Botanic Gardens at Kew. cave as a classical architectural model. This antiquarian view,
3. For a compelling and thorough elaboration
on this interaction, see Allison Ksiazkiewicz, instead of a purely geologic interpretation, had its reasons. At
“Geological Landscape as Antiquarian Ruin: certain angles, the vertical basalt pillars and the lava flow look
Banks, Pennant and the Isle of Staffa,” in
Enlightenment Travel and British Identities: like a Greek colonnade supporting a ruined triangular pedi-
Thomas Pennant’s Tours in Scotland and ment. But this was more than a matter of visual analogy. On
Wales, ed. Mary-Ann Constantine and
Nigel Leask (London: Anthem Press, 2017), a broader level, the cave showcases the intersecting interests
183–202. Related discussions can be found
in Noah Heringman, Sciences of Antiquity:
of naturalists and antiquarians – between natural history and
Romantic Antiquarianism, Natural History, aesthetics.3 In his account, Banks wrote, “Compared to this
and Knowledge Work (Oxford: Oxford
University Press, 2013); and Barry Bergdoll,
what are the cathedrals or the palaces built by men!,” and went
“Of Crystals, Cells, and Strata: Natural on to suggest that nature – instead of the Grecian school of
History and Debates on the Form of a New
Architecture in the Nineteenth Century,” thought – should be the real basis for architecture.4 Uno von
Architectural History 50 (2007): 1–29. See Troil, one of Banks’s companions on the excursion, expressed
also Geoffrey Grigson, “Fingal’s Cave,”
Architectural Review 104 (August 1948): 51–54. similar thoughts on Fingal’s Cave: “How splendid do the porti-
4.  Banks continued: “Where is now the cos of the ancients appear in our eyes. . . . But when we behold
boast of the architect! Regularity the only
part in which he fancied himself to exceed the cave of Fingal . . . this piece of architecture, executed by
his mistress, Nature, is here found in her nature, far surpasses that of the colonnade of the Louvre, that
possession, and here it has been for ages
undescribed. Is not this the school where of St Peter at Rome, and even what remains to us of Palmyra
the art was originally studied, and what and Paestum, and all that . . . the Greeks could invent.”5
has been added to this by the whole Grecian
school? . . . How amply does nature repay More than a Romantic aesthetic appreciation of ruins,
those who study her wonderful works!”
Pennant, A Tour in Scotland, vol. 1, 301.
during the first half of the 19th century, ruins were under-
Emphasis original. stood as geological remnants, while rocky mounds, glacial

163
5.  Uno von Troil, Lettres sur l’Islande deposits, and quarries were appreciated because of their par-
(Paris: Chez P. Fr. Didot le jeune, 1781),
376, quoted in Barthélemy Faujas de
allels with human constructions.6 The correlation between
Saint-Fond, A Journey Through England geologic fossils and antiquarian ruins pointed to the emer-
and Scotland to the Hebrides in 1784, ed.
and trans. Sir Archibald Geikie, vol. 2 gence of the field of geology as a distinct discipline in late
(Glasgow: Hugh Hopkins, 1907), 43. 18th-century Europe, a consequential moment for the inter-
6.  See Michel Makarius, Ruins, trans. David
Radzinowicz (Paris: Flammarion, 2004). action between natural and human history. During this time,
7. Martin J.S. Rudwick, Bursting the Limits as geologist and historian Martin J.S. Rudwick elaborates,
of Time: The Reconstruction of Geohistory in
the Age of Revolution (Chicago: University “the sciences of the earth became historical by borrowing
of Chicago Press, 2005), 181. For a related ideas, concepts, and methods from human historiography.”7
discussion on fossils and antiquity, see
Rudwick, Earth’s Deep History: How It Was Rudwick suggests that late 18th-century naturalists were
Discovered and Why It Matters (Chicago: informed by the theories and methods of antiquarians and
University of Chicago Press, 2014), 31–77;
and David Sepkoski, “The Earth as Archive: their reconstructions of human history from ruins and mon-
Contingency, Narrative, and the History of uments. Fossils and rocks were seen as nature’s own antiqui-
Life,” in Science in the Archives: Pasts, Presents,
Futures, ed. Lorraine Daston (Chicago: ties and as evidence for deciphering the history of the Earth.
University of Chicago Press, 2017), 56–60. See
also Alain Schnapp, The Discovery of the Past,
In the case of Fingal’s Cave, as historian Allison Ksiazkiewicz
trans. Ian Kinnes and Gillian Varndell (New argues, naturalists were both observers of natural history and
York: Harry N. Abrams, 1996), 266–316.
8.  Ksiazkiewicz, 195–96.
“agents of aesthetic sensibility, who justified and perpetuated
9.  In other images from their collaboration, a framework for making knowledge and knowing Nature,” all
an archive of architectural models and plas-
ter-cast fragments of ancient ruins were
of which efficiently mitigated an “empirical sensibility within
depicted in modern architectural spaces. On an aesthetic discourse that sought to quantify art and taste.”8
John Soane’s collections, see John Elsner, “A
Collector’s Model of Desire: The House and
It is no coincidence, then, to see Fingal’s Cave in one of
Museum of Sir John Soane,” in The Cultures the most important images of architectural history. Titled
of Collecting, ed. John Elsner and Roger
Cardinal (Carlton: Melbourne University Architecture: Its Natural Model (1838), this watercolor was
Press, 1994), 155–76. Also see Amy Catania produced by English architect and theorist Joseph Michael
Kulper, “Architecture’s Lapidarium: On
the Lives of Geological Specimens,” in Gandy during his collaboration with John Soane.9 Collecting
Architecture in the Anthropocene: Encounters and recontextualizing several natural artifacts and fragments
Among Design, Deep Time, Science and
Philosophy, ed. Etienne Turpin (Ann Arbor: of the Earth in the same scene, the painting depicts various
Open Humanities Press, 2013), 87–110. geologic ruins as prehistoric prototypes for architecture. As
10.  Brian Lukacher, “Joseph Gandy and the
Mythography of Architecture,” Journal of historian Brian Lukacher puts it, Gandy’s image conveys the
the Society of Architectural Historians 53, no. 3 message that “the future history of architecture was already
(September 1994): 293.
11.  Anthony Vidler, “The Hut and the written in the landscape, merely waiting for human civiliza-
Body: The ‘Nature’ of Architecture from
Laugier to Quatremère de Quincy,”
tion to catch up.”10
Lotus International 33 (1981): 102. On Gandy’s emphasis on mimesis and origin is striking
architectural theories of the primitive hut,
see Joseph Rykwert, On Adam’s House in
compared to the two prevailing antiquarian doctrines almost
Paradise: The Idea of the Primitive Hut in a century earlier. It is not like the naturalism of the primitive
Architectural History (New York: Museum
of Modern Art, 1972). See also Jo Odgers, hut in Marc-Antoine Laugier’s 1755 An Essay on Architecture,
Flora Samuel, and Adam Sharr, eds., nor like Johann J. Winckelmann’s Neoplatonic idealism, in
Primitive: Original Matters in Architecture
(London: Routledge, 2006). which ideal form is seen as “more natural than nature her-
self.”11 Instead, Gandy’s painting depicts the Earth as a col-
lection of various geologic parts, in which the rock is the
primordial object, suggesting another kind of antiquity for
architecture. Collapsing the fossil’s geohistory and antiq-
uity – and thus natural and architectural history – Gandy’s
aesthetic techniques of reconstruction and reassembly posi-
tioned architecture as a way of knowing about the planetary,

164 Log 50
Joseph Michael Gandy, Architecture: while producing new modes of knowledge for architecture’s
Its Natural Model, 1838. © Sir John troubled quest for an origin.
Soane’s Museum, London.

Chamber of Wonders
How would Gandy compose this image today to refer to a
“natural model” in the Anthropocene? If the emergence
of natural history as a “modern episteme” in the late 18th
century went hand in hand with antiquarian and natural-
ist origin myths and redefinitions of aesthetic categories for
architecture, what kinds of architectural stories might future
12.  Vidler alludes to the emergent necessity archaeologies tell?12 In this age of technofossil accumulation,
during this era to revisit aesthetic catego-
ries and the origin narrative in “The Hut
the material residues of electronic and plastic waste create a
and the Body.” On the idea of the “modern new mineralogy, sedimenting human and natural histories
episteme,” see Michel Foucault, The Order
of Things: An Archaeology of the Human together.13 How can architecture help imagine ways of know-
Sciences (London: Tavistock, 1970). ing that pertain to our future archaeologies?
13.  On technofossils, see Sy Taffel,
“Technofossils of the Anthropocene: Media, A good reference point here might be the Wunderkammer,
Geology, and Plastics,” Cultural Politics 12, or cabinet of curiosities, of the 17th and 18th centuries, which
no. 3 (2016): 355–75. On new minerals of the
Anthropocene, see Robert M. Hazen et al., is regarded as an early natural history museum. Since the
“On the Mineralogy of the ‘Anthropocene Wunderkammer made no distinction between natural speci-
Epoch,’” American Mineralogist 102, no. 3
(March 2017): 595–611. mens and human artifacts, the collections ranged from miner-
14.  On this point, see Lorraine Daston and als, fossils, plants, and stuffed animals to sculptures, paintings,
Katharine Park, Wonders and the Order of
Nature, 1150–1750 (New York: Zone Books, and scientific instruments. This flattening of categories
1998), especially 255–301. showcased neither the universalist and objective classifica-
tion of encyclopedism nor a chaotic, muddled miscellany.14
Similar to Gandy’s painting, the Wunderkammer conveys an

165 Log 50
Thomas Major (based on a drawing by
James Miller), Fingal’s Cave, Staffa,
1772. Drawing from Joseph Banks’s
voyage. © British Library Board.

evidence-based and thus encyclopedic character, while also

portraying subjective elements of myth and enchantment
in its selection of artifacts and specimens. The story that the
Wunderkammer tells, then, is a self-aware way of seeing and
15.  For more on this idea in the context knowing the world.15 This is observable in the actual con-
of museums, see Svetlana Alpers, “The
Museum as a Way of Seeing,” in Exhibiting
tents of the collections and the way in which their rarity was
Cultures: The Poetics and Politics of Museum equated with allure.
Display, ed. Ivan Karp and Steven D. Lavine
(Washington, DC: Smithsonian Institution, Historian Robert Felfe demonstrates that a considerable
1991), 26–32. change occurred in drawings of the Wunderkammer between
16.  Robert Felfe, “Collections and the
Surface of the Image: Pictorial Strategies
the early 17th and late 18th centuries. Initial depictions show
in Early-Modern Wunderkammern,” in interior views of the collections; however, later on, the inte-
Collection, Laboratory, Theater: Scenes of
Knowledge in the 17th Century, ed. Helmar rior is depicted as a stage behind a proscenium, communicat-
Schramm, Ludger Schwarte, and Jan ing the growing self-awareness of the distinction between the
Lazardzig (Berlin: De Gruyter, 2005), 247.
Emphasis original. collection and the real world. Felfe argues that this condition
17. For a history of the development of the can be understood as a propensity “towards a medialization
diorama, see Erkki Huhtamo, Illusions in
Motion: Media Archaeology of the Moving of the museum collection.” In these images, the human as “a
Panorama and Related Spectacles (Cambridge: subject of the gaze and of knowledge – the object and order
MIT Press, 2013), 139–69; and Stephan
Oettermann, The Panorama: History of a Mass of knowledge – is renegotiated.”16 This multidimensional stag-
Medium, trans. Deborah Lucas Schneider ing of layered architectural space and historical time on the
(New York: Zone Books, 1997), 69–83.
surface of an etching or drawing also reaffirms the capacity of
a model representation to act as a space of assembly, instead of
as an origin, just like the Wunderkammer itself.

Miniature Nature
Dioramas, another microcosm of the natural, were initially
installed in specially designed theaters, combining aspects
of 18th-century panoramas, illusionistic painting, and stage
design.17 By the end of the 19th century, the term was most

166 Log 50
Frontispiece. From Johann Ernst
Hebenstreit, Museum Richterianum:
Continens Fossilia Animalia, Vegetabilia
Mar., 1743.

frequently applied to natural history museum habitat displays.

Reconstructing a three-dimensional scene frozen in time and
space as life-size sceneries, this now familiar type of diorama
18. See Karen Wonders, Habitat Dioramas: depicts an idealized nature.18 These dioramas contain zoologi-
Illusions of Wilderness in Museums of
Natural History (Uppsala: Acta Universitatis
cal specimens and precisely modeled environmental artifacts
Upsaliensis, 1993); and Stephen Christopher in a three-dimensional foreground that replicates their sur-
Quinn, Windows on Nature: The Great Habitat
Dioramas of the American Museum of Natural roundings and merges with a two-dimensional painted back-
History (New York: Abrams; American drop. They reveal a complicated array of tensions as well as
Museum of Natural History, 2006).
19. Donna Haraway, “Teddy Bear Patriarchy: various political entanglements between the imaginaries of
Taxidermy in the Garden of Eden, New the natural and the cultural.
York City, 1908–1936,” Social Text 11 (Winter
1984–85): 24, 52. On the pedagogical role Donna Haraway alludes to the social construction of
of the museum in 19th-century North dioramas in her essay “Teddy Bear Patriarchy,” which focuses
America, see Sally Gregory Kohlstedt,
“‘Thoughts in Things’: Modernity, History, on the Akeley Hall of African Mammals at the American
and North American Museums,” Isis 96, Museum of Natural History. Haraway demonstrates how
no. 4 (December 2005): 586–601. See
also Karen A. Rader and Victoria E.M. taxidermist Carl Akeley’s dioramas reflect inherent cultural
Cain, Life on Display: Revolutionizing US
Museums of Science and Natural History in the
hierarchies of race, gender, and class in the early 20th-century
Twentieth Century (Chicago: University of American cultural imagination. “A diorama is eminently a
Chicago Press, 2014); and Michael Osman,
“Representing Regulation in Nature’s
story, a part of natural history,” she writes. “The story is told
Economy,” in Modernism’s Visible Hand: in the pages of nature, read by the naked eye. The animals
Architecture and Regulation in America
(Minneapolis: University of Minnesota in the habitat groups are . . . actors in a morality play on the
Press, 2018), 81–106. stage of nature. . . . Nature is, in ‘fact,’ constructed as a tech-
nology through social praxis. And dioramas are meaning-
machines.”19 Habitat dioramas were invented around the time
when the exploitation of natural resources in the American
wilderness was accelerating. Accordingly, while reflecting the
discursive shifts in the various fields of natural sciences in
the 19th century, dioramas displaying nature were frequently
linked to imaginations of natural heritage, national identity,

167 Log 50
 and the disappearing wilderness, as well as questions of race,
eugenics, and conservation. These displays exemplify another
instance of nature gaining a “moral authority,” in the words of
historians Lorraine Daston and Fernando Vidal, presenting
20.  Lorraine Daston and Fernando itself “as a source of values and a value in itself.”20 In its
Vidal, eds., The Moral Authority of Nature
(Chicago: University of Chicago Press,
promise to demystify nature through a legitimate yet fictional
2010), 5. For an interesting reading of natu- verisimilitude, the diorama constructs new myths of scientific
ral history dioramas in the context of scien-
tific models, see Lynn K. Nyhart, “Science, validity and realism.
Art, and Authenticity in Natural History The salient aspect of dioramas, then, is their collapsing
Displays,” in Models: The Third Dimension
of Science, ed. Soraya de Chadarevian of two kinds of artifice. The first exists in their politics of
and Nick Hopwood (Stanford: Stanford nature, in that they prompt a natural idealism as a way of
University Press, 2004), 307–35. For a
discussion of late 19th-century taxidermy seeing and understanding the world. The second resides in
and museum displays of animal forms as their politics of appearance – in their actual construction of a
a potent mode for negotiating human and
nonhuman relationships as well as the picture that includes all of the techniques involved to achieve
wider politics of nature, see John Herron,
“Stuffed: Nature and Science on Display,”
their illusionary effect. It is the nature of this collapse that
in Rendering Nature: Animals, Bodies, allows the diorama to act like a staged tableau.
Places, Politics, ed. Marguerite S. Shaffer
and Phoebe S.K. Young (Philadelphia:
Here, instead of deliberately invoking doubt, artifice aims
University of Pennsylvania Press, 2015), to suspend disbelief temporarily and to replace it with magic
48–69. See also Rachel Poliquin, The
Breathless Zoo: Taxidermy and the Cultures and wonder. Theorist Susan Stewart argues that miniatures,
of Longing (University Park: Pennsylvania in their attempt to “bring historical events ‘to life,’ to imme-
State University Press, 2012).
21.  Susan Stewart, On Longing: Narratives of diacy . . . lose us within their presentness” and remove any
the Miniature, the Gigantic, the Souvenir, the sense of “not only labor but causality and effect.”21 Similarly,
Collection (Durham: Duke University Press,
1993), 60. the collapse of these two kinds of artifice in the diorama can-
cels any awareness of causality in both the backstage of con-
struction and the cultural biases inherent in an idealized
nature. In other words, dioramas are constructs of fake
natures made into a microcosmic theater.

Nature Under Construction

In his Animal Logic series, photographer Richard Barnes
depicts natural history dioramas as they go through con-
struction, renovation, or maintenance. In Man with Buffalo
(2007), a maintenance worker is inside a diorama vacuum-
ing artificial snow while facing a taxidermied buffalo. In the
foreground, we see the metal studs and plywood frames of
the diorama in construction. Giraffe (2005) depicts a taxi-
dermied giraffe with its head wrapped in plastic and its neck
held up by a crane in front of a natural scenery backdrop. The
remnants of the quotidian labor of restoration and mainte-
nance are visible: a broom is leaning against a table full of
small pieces of the diorama and other animals, models, and
equipment, such as light fixtures. If in dioramas causality is
replaced by magic and wonder, Barnes’s photographs restage
causality as quotidian labor. By destabilizing the collapse
between the artifice of appearance (illusionism) and nature

168 Log 50
Richard Barnes, Man with Buffalo, 2007. (idealism), Barnes’s photographs open up a space of inquiry.
From Animal Logic. © Richard Barnes. Comprehending nature and its staging both as construc-
tion and in construction opens up an alternative imagination of
the diorama, and the idea of the model, in the Anthropocene.
In her book Possessing Nature, historian Paula Findlen shows
how the creation of science museums produced new attitudes
toward nature and influenced the development of the disci-
22.  See Paula Findlen, Possessing Nature: pline of natural history during the 16th and 17th centuries.22
Museums, Collecting, and Scientific Culture
in Early Modern Italy (Berkeley: University
If the rise of natural history museums goes hand in hand
of California Press, 1994). It should also with the cultural imagination of nature, what kind of stories
be added that in the context of colonial
expansion, natural history museums were will the dioramas, as new models of the Anthropocene, tell?
architectures of the imperial archive where In parallel with the climate crisis, how does rethinking the
botanical, animal, and mineral samples
were collected. unnatural model help us produce contemporary architectural
imaginaries of the environment?
As the Anthropocene destabilizes distinctions between the
natural and the artificial, the first step is to let go of the very
idea of the “natural” model as it relates to the environment. If
we are intervening in everything from the upper atmosphere
to the deep seas, it is difficult to argue for the idea of a natural
state. The second step is to forgo the very idea of the “natural”

169 Log 50
NEMESTUDIO, Our Junk, Their Ruin,
2018. Elevation perspective drawing of
a building from the Wunderkammer.
Opposite page: Middle Earth:
Dioramas for the Planet, 2017. View of
Plastic Pacific Hall. Drawings courtesy
the architect.

model as it relates to architecture’s quest for its own origin.

Once a cushion for myths about architecture’s disciplinary
stability, the idea of the origin waits to be debunked, desta-
bilized, decolonized, and stripped of its relevance as a divine
source. When liberated from these ideas, we might soon real-
ize that there is actually nothing special about the new models
of the Anthropocene themselves, or any architectural refer-
ent, for that matter. The potential, as we may come to under-
stand, might instead lie in a model’s capacity to instigate ways
of knowing – about architecture and about the world at large.

Postscript
In an attempt to provoke alternative environmental imagi-
nations of the Anthropocene, in some of our recent work at
NEMESTUDIO, the very staging of the unnatural model – as
construction and in construction – is taken as an architec-
tural problem. On the one hand, through the restaging of
familiar models and replicas such as the diorama, the theater
set, and the Wunderkammer with the messy stories and con-
troversies of resource extraction, waste, and environmental
injustice in the Anthropocene, the notion of nature as an ideal
and universal model is denaturalized. On the other hand, a
slowed-down attention to construction and assembly is pro-
posed by revealing the mundane details of these models’ mak-
ing and unmaking.

170 Log 50
 Instead of conceptualizing the environment as preserv-
ing nature, as technological management, or simply as visu-
alization, it is important for architecture, at this moment, to
consider the potential of a planetary imagination from within
its inner workings and techniques. To rethink the potential of
the unnatural model is to appreciate the power of these inner
workings and techniques to reveal, to unhinge, and to recon-
stitute larger questions of environmental imagination for the
worlds we design, the institutions we build, and the systems
of knowledge and value through which we imagine architec-
ture and the world. Other architectural imaginations of the
environment are only possible with this kind of recognition
and self-awareness.

Neyran Turan is a founding partner of

NEMESTUDIO and associate profes-
sor at the University of California,
Berkeley. She is editor-in-chief of New
Geographies 1: After Zero and author
of Architecture as Measure.

171
Observations on a Border Model

Blane De St. Croix, Broken Landscape Broken Landscape IV is a 72-foot-long, seven-foot-high, and two-
IV, 2009. Installed in “Blane De St. and-a-half-foot-wide model of the US-Mexico border built by artist
Croix: How to Move a Landscape,”
MASS MoCA, North Adams,
Blane De St. Croix in 2009. It is constructed out of natural materials
Massachusetts, July 2020–September such as branches and dirt as well as foam and plywood. Years before
2021. Photo: Kaelan Burkett. Courtesy the recent focus on the border, De St. Croix conducted research along
MASS MoCA.
over 3,000 miles of the fence that separates the US and Mexico, vis-
iting 15 border crossings and speaking with people living and working
on both sides. Through this on-site research, De St. Croix approached
his work – the model – in the context of landscape painting.
But from an architect’s perspective, what is perhaps most
intriguing about this model is the way De St. Croix cut a section deep
into the ground, slicing it into a long, narrow strip and exposing its
geological layering. The fragile, manufactured, and temporary qual-
ity of the fence is an obvious contrast to the layers of earth, which
are rendered with the same precision as the fence itself, diminishing
the importance of the barrier. Thus the border is disempowered.
The model most recently went on view at MASS MoCA in 2020,
as part of De St. Croix’s exhibition “How to Move a Landscape.”
It was placed diagonally as an obstruction in the gallery space,
becoming a border itself that directly confronts the viewer. Due to
its length, the model cannot be seen in its entirety all at once. The
viewer must move around it, walk along it, catch details, get closer,
and walk back to gain perspective. Then, having had an embodied
experience of the model as an obstacle, the viewer can no longer
remain indifferent to it and its subject. – Alican Taylan

172
Joe Day
One to One

Whether the megaliths enter our contemporary consciousness or

we move closer to the Stone Age is not all that important. What is
important is to feel a slight release from the present, to feel at ease
and at home here and then.  – Anna Neimark

In the early, less complicated months of 2020, two remarkable

and complementary installations opened in Los Angeles, one
hosted by the MAK Center for Art and Architecture and the
other by SCI-Arc. Michelle JaJa Chang’s “Scoring, Building”
was built in stages in the courtyard of the MAK-operated
Mackey Apartments, and it overlapped for precisely the leap-
year extent of February with “Rude forms among us,” by
Anna Neimark, exhibited in the SCI-Arc Gallery. Both offered
fresh templates for full-scale environmental conjecture, as
well as some hope for the increasingly fraught genre of instal-
lation architecture. Eight years after Sylvia Lavin’s withering
and much-circulated disavowal of pavilions as exhausted, and
exhausting, endgames for emerging architects, such as both
Chang and Neimark, these two works breathe unexpected life
1.  See Sylvia Lavin, “Vanishing Point: The into that recently moribund form.1
Contemporary Pavilion,” Artforum 51, no. 2
(October 2012): 212–19.
Chang’s piece followed a competition-winning (but at
the time unsited) proposal for a structure based on a series
of musical-cum-spatial notations or “scores.” Materials &
Applications, the public design nonprofit that organized the
competition, secured the interstitial MAK site, and Chang’s
proposition was tailored accordingly. A standard litany of
building materials – a fixed number of two-by-four steel
studs and four-by-eight sheetrock panels, and allotments of
screws, spackle and paint – was gradually deployed in three
iterations in the open space between the apartment block and
garage/gallery building of the Mackey Apartments. In the first
iteration, Chang bounded the void with a long rectangular
frame of metal studs, left open above. In the second, the fram-
ing was clad in gypsum board to form an open-air “room”
between the two main volumes of the Mackey complex. In the
final iteration, new wall elements were tightly layered inside
of one another, revealing considerable nautilus-like internal
complexity when viewed from above.

173
Michelle JaJa Chang, iterations two It’s tempting to read the three stages of Chang’s installa-
and three of “Scoring, Building,” tion as a neat reverse archaeology of the techniques of insti-
Mackey Apartments, The MAK Center
for Art and Architecture, Los Angeles,
tutional demystification pioneered by the late Michael Asher.
January 18–February 29, 2020. Photos Chang’s first stud wall iteration closely resembled Asher’s
courtesy the architect. 2008 “retrospective” at the Santa Monica Museum of Art,
in which he reframed every display wall of every exhibition
ever hosted by that museum with open metal studs; her sec-
ond room echoed the ambivalence of his vacated art spaces,
and her third stage interpolated the second repeatedly, like a
series of “Russian doll” art spaces nested each inside the pre-
vious. If Asher removed gallery walls to reveal private spaces
of judgment and commerce, Chang repeated her white
walls until they foreshortened all views within – a gallery
occluded. Both Asher and Chang repeatedly abrogated the
pristine white cube, as well as its assumptions about where
architecture must stop in order for the art to start.
But to focus on the fixed manifestations is perhaps to
mistake Chang’s broader intentions. The procedures and
durée of building and rebuilding that she addressed through
the logistical score, a timetable involving many volunteers,
were central components of the work. As such, the project
and its title also recall Matthew Barney, both in his parametric

174 Log 50
Eugène Trutat, Dolmen du Mas d’Azil,
Ariège, France, later 19th century.
Digital image from a negative glass
plate with gelatin and silver bromide
intended for projection, stereograph,
plate format 3 1/2 by 6 1/4 inches.
Photo courtesy Bibliothèque munici-
pale de Toulouse.

Drawing Restraints and in the many space-defining rituals in

The Cremaster Cycle. As Barney once enlisted Richard Serra
to sling Vaseline in the Guggenheim the way Serra had long
before hurled lead into the corners of his studio, Chang invited
Neil Denari to play guitar at her opening, underscoring the
role of music, and specifically the techniques of reverberation
and remixing, in both of their practices. The sound behind
Chang’s scores was William Basinski’s The Disintegration Loops,
a haunting hour-long cycle of found, relooped sonic “facture”
in which the magnetic information on cassette tape is heard
gradually flaking away to silence.
Neimark’s installation relied less on a process of for-
mation than on the obdurate power of objects to inspire
and propagate kindred objects. Neimark’s long fixation on
dolmens, the Neolithic stacked-stone monuments scattered
across Eurasia, led to a collaboration with the Natural History
Museum of Toulouse, France, which provided both a pair of
19th-century images of the Dolmen de Vaour by the photog-
rapher Eugène Trutat and curatorial support for the exhibi-
tion – very much a show within a show, a notional Russian
doll in contrast to Chang’s compositional one.
While the reverse prints are stunning, shrouding the
megalithic dolmen in hazy, noir mystery, in Neimark’s
revelry they became something more, something closer to
Rosalind Krauss’s remarkable cross-reading of two prints,
photographic and lithographic, of Tufa Domes, Pyramid Lake,
2.  See Rosalind Krauss, “Photography’s Nevada, by Timothy O’Sullivan.2 Just as Krauss plumbs
Discursive Spaces: Landscape/View,”
Art Journal 42, no. 4: The Crisis in the
two “identical” images for their very different exposures,
Discipline (Winter 1982): 311–19. and famously favors the less nuanced version for its stark
contrasts, Neimark reads her dolmen source images with a
strategically blind eye, inverting high and low resolution for
her own ends.

175 Log 50
Anna Neimark, exhibition model for
“Rude forms among us.” Photo cour-
tesy the architect.

The crux of Neimark’s exhibition is an imposing, crisply

executed model home of airbrushed structural insulated
panels composed, like the trabeated dolmen, of a looming
3.  This pavilion, a compressed prototype roof plane held aloft by heavyset piers.3 Monolithic C-shaped
for an accessory dwelling unit, was,
according to Neimark, based on a slightly
extrusions, these six 10-foot-high piers would encapsulate
larger commission for a 2,000-square-foot all functional requirements of the accessory dwelling unit
custom home.
4.  Adolf Loos or Marcel Duchamp likely
(ADU), thus transforming, as Gordon Matta-Clark might
inspired Gordon Matta-Clark: “One of have it, art into architecture.4
my favourite definitions of the difference
between architecture and sculpture is This loose amplification of the dolmen form, under-
whether there is plumbing.” James Attlee, taken with her collaborator Andrew Atwood at First Office,
“Towards Anarchitecture: Gordon Matta-
Clark and Le Corbusier,” Tate Papers 7 surprised even more as one deduced the scale of original
(Spring 2007), https://www.tate.org.uk/ lithic structure. At first it appeared to be roughly the size of
research/publications/tate-papers/07/
towards-anarchitecture-gordon-matta- a hut, with habitable space below, but on closer inspection it
clark-and-le-corbusier. was the size of a large altar. Neimark’s leviathan structure,
shoehorned into the SCI-Arc Gallery, while at the same time
rigorously disengaged from it, allowed just two comprehen-
sive views of the “Rude form”: the view from the gallery
entry and a view across its roof from a second-floor catwalk
that runs along the gallery’s eastern edge. The snug fit of the
“Rude form” within the gallery echoed that of Tony Smith’s
Smoke, lodged in LACMA’s entry atrium until its recent
demolition and often the basis for first-year graduate studio
exercises headed by Neimark at SCI-Arc.
To argue that these two projects by Chang and Neimark
offer new provocations, one might need to look past the
recent quarter-century wave of architecture-in-the-gallery,
and even beyond postwar amalgamations of art and

176 Log 50
Anna Neimark, “Rude forms among
us,” SCI-Arc Gallery, Los Angeles,
January 31–March 27, 2020. Photo:
Marten Elder. Courtesy the architect.

architecture such as Peggy Guggenheim and Frederick

Kiesler’s Art of This Century gallery. A better touchstone for
Chang and Neimark might be the Polish sculptor Katarzyna
Kobro, who developed her Spatial Composition 3 in 1928, fol-
lowed by a series of architectural proposals that precisely
magnified that original planar construct up to a variety of
inhabitable scales and uses. All versions are rigorously similar
in the mathematical sense: in all manifestations, all internal
and surface proportions remaining constant, with no loss (and,
as crucially, no augmentation) of detail. A diagram, a sculp-
ture, a store, a house, such was Kobro’s faith in her Composition
that she dismissed the idea of “one size fits all” in favor of a
more radical proposition: that one shape suits all – simply scale
as needed.
Kobro’s work is also a swan song for synthesis among the
arts – Composition 3 is simultaneously and unapologetically
art, architecture, and design. However, Chang and Neimark
operate in a world both less and more siloed. Art and archi-
tecture have shared and stolen much from each other in the
last half century, but their practitioners and critics – in ways
that would likely have perplexed Kobro – have also grown
far more insistent about the divides between their discourses.
Lavin elegantly laid out in Artforum the complicities and con-
tretemps between overlapping environmental practices, as
has Hal Foster in his many volumes. Neither architects nor
artists can be innocents in the field of spatial production any-
more, least of all the postpavilion generation of Chang and

177 Log 50
 Neimark. To repurpose Philip Johnson’s dictum on history
for the 21st century, one cannot not know art.
My resort to considering these projects exclusively in
terms of artists rather than architects began in the name of
precision, but it became a reflection of the current extent of
disciplinary displacement. Critics often mention an artist early
in an architectural review as a flattering aside, but here it
makes more sense simply to stay in that milieu. These projects
clearly extend possibilities in fine art, if obliquely or modestly,
while refusing, and thus perhaps refuting, many of the issues
central to the millennial installation craze in architecture.
And yet, neither Chang nor Neimark purports to be an
artist. In both installations, the notion of the model is dis-
lodged from its standard transitional role in architecture to
a more terminal one. In neither show is there a scaled rep-
resentation of, or toward, a final design – the “Rude form”
itself is simultaneously model, prototype, and end result. In
the case of “Scoring, Building,” the cyclical and, in certain
respects, casual production resembles crowdsourced model
making more than conventional construction (as it also recalls
Sol LeWitt’s recipes for works of art to be enacted by others).
For that matter, neither is a projection of a drawing, not in
the archaic sense of a 3D extrapolation from a flat plan nor in
the timelier one of a file 3D printed. Instead, Chang amplifies
a musical composition and Neimark interprets a century-old
pair of photographs, each eschewing as many conventions of
architectural production as possible.
If these projects are themselves models, just what are
they modeling? A new brokerage between art and architec-
ture, perhaps, one that favors the ways of the former while
still claiming allegiance to the latter. Many of their specific
gestures – the shifting, pinched shape of Chang’s enclosures
and the portals into and out of those enclosures; the manifold
attunements of Neimark’s wall-piers to ADU programming
– play to a narrow home audience, but their major moves
reflect a broad canon of postminimalist art. Both Chang and
Neimark choreograph spaces and expectations in novel ways,
redefining a primary challenge in their field by raiding the
arsenal of another.

Joe Day is a designer and critic

in Los Angeles.

178
David Eskenazi
Tired . . . and
Behaving Poorly
In today’s digital work space, objects easily swivel from size to
size. They can be enlarged or reduced without consequence,
enabling architects to playfully assemble forms from a palette
of references, big and small, that are then deployable as build-
ings, furniture, paperweights, and even cities. Their flexibility
permits architects to momentarily ignore the particularities
of materials, structure, and social organization. For some,
this has led to the use of enlarged or miniaturized references,
while others have imagined architectural compositions that
1. A recent example is Tom Wiscombe’s idea operate at sizes between building massing and material detail.1
of the supercomponent, an element that is a
massing-sized detail. See Tom Wiscombe, “A
As a creative vector, the digital object is a shape-shifter, end-
Specific Theory of Models: The Posthuman lessly eluding expectations of being pinned down to a specific
Beauty of Weird Scales, Snowglobes and
Supercomponents,” Architectural Design 89, size or material form.
no. 5: Beauty Matters: Human Judgement Galileo observes in Two New Sciences that if two ships
and the Pursuit of New Beauties in Post-
Digital Architecture (September/October were built identically in every way except that one was an
2019): 80–89. enlargement of the other, the larger ship would undoubtedly
2.  Galileo Galilei, Two New Sciences, trans.
Henry Crew and Alfonso de Salvio (New need more structure to prevent it from collapsing under its
York: Macmillan, 1914), 131. own weight.2 For this simple reason, Galileo explains, giant
animals cannot exist. In other words, forms in the world
effectively have something to do with size. Galileo’s observa-
tion suggests that a digital object that embeds the material,
structural, and mechanical limits of the physical world into
its data set can no longer be sized up or down. For architects,
the specificities of building inevitably move the digital object
toward an intended building size, at which point it is size-
specific and unenlargeable.
Models that behave well match the proposed building as
faithfully as possible. The size-specific digital model is well-
behaved: it accounts for building properties like wall thick-
nesses and structural systems. Meanwhile, a good physical
model’s geometries and materiality suggest those of the build-
ing. Following Galileo, the physical model is a smaller object
that resembles the larger building yet has its own material
properties – the two cannot be identical. An attempt to make
both model and building with the same material and structure
would inevitably fail, suggesting that a separate digital model
must be created to embed the materials and fabrication tech-
niques specific to constructing the physical model. Whether

179
Opposite page: David Eskenazi, Two a physical model behaves well depends on what it is intended
Scrolls, 2020. Left column (top to bot- to do: either it is a dutiful depiction of the building or it is a
tom): nine-inch-tall kraft paper model;
nine-inch-tall model at 1:12 scale with
faithful output of the digital object.
structural scaffold on interior, kraft In a recent proposal for two differently sized scroll-
paper and 3D prints; digital simulation like structures, I constructed a set of physical models whose
of nine-foot-tall structure; nine-inch-
attempts to faithfully reproduce digital models and depict
tall model at 1:12 scale with structural
scaffold on exterior, 3D prints. Right buildings render a vague resemblance to both without duti-
column (top to bottom): three-foot- fully portraying either. In effect, the physical models behave
tall kraft paper model; nine-inch-tall
poorly. Their depicted architecture is unsettled, and they are
model at 1:60 scale with structural
scaffold on interior, 3D prints and kraft disconnected from a digital model.
paper; digital simulation of 45-foot- Both physical models are made of paper, a material whose
tall structure; nine-inch-tall model at blankness gives the impression that it is simply a neutral sur-
1:60 scale with structural scaffold on
exterior, kraft paper. All images cour-
face ready to be shaped or drawn on – much like the digital
tesy the architect. work space. But paper is more than a container for content.
Before it is ready for use, it has been shaped and stiffened
into a complexly manufactured material. Despite predictions
that paper would become superfluous due to smartphones
and tablets, it is enjoying a renaissance as a newly engineered
shipping material for online shopping delivery.3 It has shifted
from a material for printing content to one whose very mate-
riality is significant. This is why the paper model can resist
the neutrality suggested by the digital work space.
The project, Two Scrolls, began by curling a small and
a large piece of kraft paper into two connected cylinders. In
the smaller model, the paper behaves well, with some slight
bowing. At the larger size, the model misbehaves as the paper
slumps until the cylinders press against one another in what
resembles a lamenting posture. The paper models produce
two size-specific forms that don’t comply with geometric cer-
tainty. They don’t resist gravity. They’re fragile – tired, even.
The same posture is then simulated digitally. Starting
with digital models of the scrolls, identical to the two paper
sizes, a computer graphics physics engine simulates the behav-
ior of the small and big paper models, mimicking the stiffness
and weight of the physical experiment. The real and simu-
lated papers slump as expected.
3.  “The cardboard box as we know it has It isn’t until the scrolls are proposed as a small and a large
changed more in the past 10 years than in
the previous 100.” With home delivery,
building that the paper slumps begin to behave poorly. As archi-
shipping companies are increasingly tecture models, we’d expect them to faithfully replicate the
conscious of the aesthetic of the cardboard
box and are using artificial intelligence to forms of the anticipated buildings. In the proposal, however, the
engineer thousands of specific boxes. Mark paper slumps are both pre- and postdigital in that the physical
Wilson, “The hot new product Amazon
and Target are obsessing over? Boxes,” models incorporate digitally simulated paper slumps and actual
Fast Company, May 6, 2019, https://www. paper. The effect of these models is to make it debatable whether
fastcompany.com/90342864/rethinking-
the-cardboard-box-has-never-been-more- these are geometries we can anticipate in the building or if the
important-just-ask-amazon-and-target. models simply denote the proposed material of the building.

180 Log 50
David Eskenazi, Two Scrolls, 2020. The proposed buildings are nine and 45 feet high respec-
Three-foot-tall kraft paper model and tively, with a smattering of six-foot-tall window cutouts. The
still from the digital simulation.
physical models are equal in size but their scales are different.
They consist of kraft paper, 3D prints of digitally simulated
paper surfaces, structural scaffolds, and tape. In the nine-foot-
tall proposal, the kraft paper is on the exterior, surrounding a
structural scaffold and 3D-printed simulations of nine-foot-
tall paper. Due to the size of the kraft paper, the outside of this
model appears as two stiff, leaning cylinders, while the inte-
rior forms appear to struggle to stay upright. For the 45-foot-
tall proposal, the 3D-printed simulation of the paper is on the
exterior while the stiff kraft paper is on the interior. Tired and
struggling to stand, the outside of this larger proposal appears
to be collapsing while its interior seems relatively stable. The
two physical models propose layered architectures whose
interiors are shielded from exterior view and whose relation-
ship to gravity shifts from resistance to acceptance. The inside
of one model is similar to the outside of the other, relating the
small and large buildings to one another.
The models are held together with two types of tape.
Brown kraft tape holds the kraft paper together while colorful
graphic tape stitches the rigid digital prints. The tape is either
a trace of the simulation software’s colorful notation system
or, in the case of the brown tape’s changing width, a sign of
the scale of the model. The windows also indicate scale; their
sizes change between the models to indicate the change in
height of the two buildings. As a simple cutout, each window
reads like a graphic notation that complements the tape.

182 Log 50
David Eskenazi, Two Scrolls, 2020. Top Together, the models are a pairing that complicates the
view of nine-inch-tall models at 1:12 reading of each element as representing something or not.
scale and at 1:60 scale.
The kraft paper models are specific to the size of the paper
when it is slumped, while the digital simulations of the paper
slumps are able to swivel to larger and smaller sizes as they
simulate the physical models. It is difficult to tell whether
the proposed building would resemble the paper model or
whether the building would produce an entirely different
slumped form. Do these models behave well? Some informa-
tion, like the kraft paper, suggests that they don’t, while other
information suggests that they do: the windows point toward
the buildings and the tape back toward the digital model.
These models construct an architecture that appears
tired, struggling to reconcile an upright geometry against
the weight of its material. Paper is allowed to behave freely;
digital simulations are not. Buildings are suggested but kept at
bay, and digital information only accounts for half the geom-
etries. These paper models may act poorly and their archi-
tecture may be downcast, but a new possibility arises when
models are freed from their digital predecessors and from a
faithful adherence to a building.

David Eskenazi leads d.esk in Los

Angeles and was recently awarded a
MacDowell Fellowship and the League
Prize from the Architectural League of
New York. He teaches at SCI-Arc.

183
Observations on a Model City

Michael Elmgreen and Ingar Dragset, Presenting a top-down bird’s-eye view from a bottom-up worm’s-eye
The Hive, 2020, Moynihan Train Hall, perspective, Elmgreen & Dragset’s The Hive (bees enter the mix
New York City. Stainless steel, alumi-
num, polycarbonate, LED lights, and
too) is sure to entrance travelers when the long-awaited Moynihan
lacquer, 45 feet long by 22 feet wide by Train Hall opens in Manhattan on January 1, 2021. A Berlin-based
12 feet deep. Photo: Nicholas Knight. artists duo known for their wit (see their Prada Marfa with Rael
Courtesy Empire State Development
San Fratello, for example), Michael Elmgreen and Ingar Dragset
and Public Art Fund, New York.
were commissioned in 2019 to produce one of the three major public
artworks installed in the expansion to Pennsylvania Station. Their
model city, which includes both replica and fictional towers built
at 1:100 scale and weighs more than 30,000 pounds, is also a “city
of light,” suspended from the ceiling just inside the West 31st Street
midblock entry. In yet another nature metaphor, the artists com-
pare the upside-down towers to stalactites, towers that worms and
humans typically experience as stalagmites.
The Hive appears to derive from the duo’s City in the Sky
(2019), also an inverted model of towers, but smaller and sus-
pended from a tall rectangular frame suitable for traveling exhibi-
tions. The popularity of City in the Sky suggests that the larger,
permanent Hive, its mirror base reflecting mingling 1:1 oglers,
will become an Instagram-worthy tourist attraction on a par with
Thomas Heatherwick’s “honeycomb” Vessel in nearby Hudson
Yards. But its 2020 installation will also be a fitting reminder of
the year urban life itself was turned on its head. – The Log Team

184
Patrick Templeton
A Model Painter:
Behind the Scenes
With Amy Bennett
Amy Bennett is a painter who makes models. For her, models func-
tion as still lifes that she poses for her narrative paintings. In her
studio in Cold Spring, New York, she makes models that range from
isolated domestic interiors to entire landscapes dotted with hun-
dreds of suburban houses. These scenes are not typical maquettes,
which are minimal in detail and primarily help an artist maintain
proportion. Rather, Bennett meticulously details them, incorporating
model train figures, foliage, and even functioning lights to produce
voyeuristic glimpses into the mundane and the everyday. As a result
of this attention to detail, her paintings become facsimiles of her
models – representations of representations – producing a dream-
like or perhaps a distant memory quality, which is heightened by the
paintings’ small size: often no bigger than a page in a diary that
the viewer must lean in to read. Our “conversation” took place via
e-mail in October and November 2020. – PT

PATRICK TEMPLETON: When did you start making mod-

els and how did the process of model making change the way
you paint? 

AMY BENNETT: I have always preferred working from life,

painting while observing light hitting a figure, a still life, or a
landscape. In searching for an approach to make narrative paint-
ings, I worked first from toys and then from dollhouse furniture,
and eventually I discovered the world of model railroad min-
iatures. I have been working at that scale, or even smaller, for
the past 15-plus years. For each painting I create a miniature
three-dimensional model to serve as a still life that captures a
certain narrative moment. I construct miniature houses and
interiors using cardboard, foam, wood, paint, and glue. The
model gives me complete control over lighting, composition,
poses, colors, etc., and allows me to view the scene from all
angles before choosing a vantage point from which to paint.
Painting something from the imagination in a way that
looks uncannily real results in a style that I think Eleanor

185
Amy Bennett, model for Rubber Gloves. Right: Rubber Gloves, 2018. Oil on panel, four by five inches. All images courtesy the artist.

Heartney characterized well as “an unset- find what I’m after. Establishing the lighting
tling kind of realism – simultaneously and contrast is essential to the mood, and is
artificial and naturalistic.” Because I’ve cre- therefore my top priority.
ated every element of the scene and know Another use of the model is in the “cast-
it inside and out, my paintings have more ing process” when I decide which figures
detail and information in them than they to use in which poses. Those figures have to
might have otherwise. I am also able to look be painted ahead of time also. The traits of
at and focus on different parts of the model the people I choose or don’t choose – race,
rather than being limited to a singular focal weight, gender – can have big implications
point, like with a camera. My biggest chal- for how the narrative is interpreted. I empa-
lenge is to keep the paint feeling loose and thize with the characters, and I hope that
fresh even at a tiny scale. It is difficult to be translates through the paint.
precise but not overwork the brushstrokes.
PT: Sometimes you exhibit your models
PT: Model houses made of cardboard, foam, alongside the paintings and sometimes you
and wood are ubiquitous in architecture, but only exhibit the paintings. How do you decide
they’re almost never painted, as yours are. when to exhibit the models? And why is it
In a sense, you’re painting twice – first you important to put so much time into produc-
paint the model and then you produce the ing these intricately detailed models when
painting. How does that doubling impact the they may not ever be shown?
decisions you make about the final painting?
AB: I think of a model as means to an end.
AB: You’ve hit upon one of the many ways that It is a tool for extracting images from my
model making is useful. It replaces the tradi- imagination and it is a still life from which to
tional sketching process, giving me an oppor- paint. I want to put only as much time (it’s
tunity to try out different color schemes and very time-consuming!) and craft into mak-
vantage points. I can experiment with colors, ing one as needed so that I can spend most of
lighting, materials, and poses until I finally my time painting. That is, I need to be con-
reach a moment when it comes together and I vinced enough by the illusion that I can paint

186 Log 50
Amy Bennett, model for In the Throes. Right: In the Throes, 2018. Oil on panel, four by four inches.

an almost real-looking scene from it. It needs AB: Sometimes I alter the model while I’m
to make me want to stare at it for the hours painting, but I usually try to work out the key
and hours that it will take for me to make a elements in three dimensions. That way, I’m
painting. I use and reuse models, repainting not wasting a lot of time painting tiny figures
and reconfiguring or reconstructing as needed, that I might ultimately want to paint out or
and they usually take a beating in the process. move. When I’m satisfied with the arrange-
If I consider having to exhibit the model and ment, it shifts from three-dimensional sketch
craft it as such, I lose a lot of freedom to play to still life and my job becomes translating it
around in hopes of arriving at something even to two dimensions. But the painting takes on
more interesting than I initially imagined. And a life of its own, through exaggerations here
tiny models made from recycled cereal boxes and there, unintentional alterations, expres-
and such don’t age well. Dust and cobwebs sive brushwork, and slight shifts of scale, and
have a proportionally greater impact at such a hopefully, the painting itself surpasses the
tiny scale. The few times I have exhibited the model at some point.
models were in shows specifically about mod-
els or artists’ processes. I feel a little squeamish PT: Your paintings are often quite small, and
sharing them because they’re so flawed, and they seem fairly close to 1:1 with the models.
an embarrassing admission of an adult at play. What role does scale, both of the model and
of the final painting, play in your work?
PT: It’s interesting that you reuse and recon-
textualize the model pieces. Is the modeling AB: Since I began working with 1:87 scale
process really the iterative phase, where models, which is HO scale for model trains,
you’re working loosely and trying things out I started using a consistently tiny scale in my
– the “adult at play,” as you say? Or is there paintings, with figures measuring about an
a particular moment when you switch to inch tall. The size of the painting is deter-
painting the final image? Does that make the mined by the space around the figures, so
painting phase a process of documentation, or interiors are only a couple of square inches
do you work back and forth, developing the while landscapes are necessarily larger. What
model and the painting in tandem? I find so useful in painting from miniatures

187 Log 50
 or something I’ve read. Things often come to
me while I’m lying around in bed or sitting in
an audience or reading fiction. Model making
helps me translate that into something new and
less documentary. The challenge then becomes
looking carefully at one’s environment and
assessing the proportions of things, and the
cues specific to certain places that describe
those spaces as institutional, private, medical,
suburban, contemporary, etc. Re-creating tiny
portions of the world in miniature forces me
to look with a fresh eye, with wonder both at
the relative size and shape of things but also
at the relationship between people, the gestures
of their bodies and the spaces in between in
order to tell stories without words.

PT: You’ve said the story behind one of your

paintings, Smoke Signals, came from a report
of a church in Florida burning a Koran. How
much of that scene, and your scenes in gen-
eral, is pure invention and how much is doc-
umentation of real spaces and real events?

AB: Everything I make is fiction. There is a

Amy Bennett, Smoke Signals, 2011. Oil on panel, 30 by 40 inches.
lot of freedom and anonymity in fiction. I
is how the small scale reduces the detail to take inspiration from my own life and obser-
the essential expressions of the poses and the vations and things I read, of course, but the
relationship between the figures. It weeds translation into a model and then the paint-
out extraneous information so I don’t waste ing is never direct and literal. For instance,
time on the things that ultimately don’t mean I have no idea what that church in Florida
much to me, like the print on someone’s shirt looked like, how many people took part in the
or even their exact facial expression. It also burning, or where that might have occurred.
establishes a more intimate connection with I painted the scene I imagined, and it’s one
the viewer, who must get close enough to that could pair more appropriately with a
become a fly on the wall. Shirley Jackson story than the news story
that inspired it. Usually I am trying to depict
PT: That distillation of detail probably allows a feeling, so everything is at the service of
you more control over how the narrative of the setting the stage for that feeling to emerge.
painting is interpreted. Where do these narra- That’s the only thing I am beholden to. With
tives originate and what kind of research goes paintings so small, things are reduced to
into preparing these scenes that tell stories? symbols and metaphors – a church, a dress,
a kitchen – in hopes they can be more uni-
AB: Scenes usually percolate in my imagina- versally relatable. Creating and lighting new,
tion, maybe sparked from my own experience imagined scenes is half the fun.

188 Log 50
 For example, I’ve just finished a painting PT: So the viewer is positioned somewhere
of a family in their living room covering their between being a fly on the wall and having a
windows in newspapers. This is not some- God’s-eye view. That simultaneous intimacy
thing I have ever witnessed directly, but I and distance seem to come directly from
have felt the need to protect my family from working with models – especially when the
external threats and know others are going to paintings are diorama-like, with walls evi-
great measures to protect their families from dently cut away. Is that what you mean by an
COVID, extreme weather, and racism. Rather unsettling realism?
than illustrating someone bleaching their
groceries on their porch, since we all know AB: That is definitely part of it, but I think
exactly what that is, I was searching for some- Heartney is also referring to the scenes and
thing more metaphorical that might translate landscapes being fakes. While they are painted
beyond one specific thing or moment. in a naturalistic way, from observation of real
light hitting real objects, something unset-
PT: Part of the feeling you elicit also seems to tling cues the viewer that what they’re seeing
come from how you position the viewer rela- isn’t painted from real life or even a doctored
tive to the scene. The vantage points of your photo of real life. Like when Manet dresses
paintings are usually set very distant from his model, Victorine Meurent, in a costume
the subjects of these scenes, so you see every- for Mademoiselle V. . . in the Costume of an
thing. Why is it important to see the entirety Espada, although he goes to the trouble of
of the spaces in which the events you’re painting a background, we don’t believe she is
depicting take place?  actually an espada in the bullfighting ring. We
understand the scene is a fake. It’s not even a
AB: The distant, godlike view is akin to an real portrait. It is fiction playing at reality. In
omniscient narrator, a dispassionate, arm’s- my paintings, the impossibility of the vantage
length viewer who has seen it all before. It can point, slight inconsistencies of scale, even
impart a sense of insignificance and contrast the nature of the scene itself reveal that the
with the drama being depicted. For instance, model is my tool for translating reality.
in my painting Delivery a child is born via
C-section and depicted from so impossibly
far across the room that I had to remove two
walls. It is the start of a life and one of the
most significant, if not the most significant,
moments in the parents’ lives, a moment that
was anticipated and that will be remembered
and celebrated annually, perhaps a traumatic
moment for the mother, and yet for the doc-
tors and nurses, it is a routine moment. There
are about 250 babies born around the world
each and every minute and about a third of
those are via C-section. A distant vantage
point can exploit the tension between the per-
sonal and dramatic and our compulsive search
for meaning, contrasted with the indifference Patrick Templeton is a quarantine gardener and
and vastness of time and human experience. the managing editor of Log.

189
Observations on Model Education

Samantha Ding, model of a suburban The first architecture model I ever made sits in my childhood bed-
house, 2011. Photo courtesy the author. room, covered in nine years’ worth of dust. It is a tiny replica of an
idyllic suburban home that I made in my junior high school archi-
tecture class, the nascent beginning of my interest in architecture.
It is a pastiche, inconsistent in scale, conflicted in its use of mate-
rial, and made with the arts-and-crafts techniques I had learned:
the spray-glued moss lawn; the driveway lined with tiny hot-glued
rocks that I perhaps gathered from the school property; the clum-
sily cut and comically thick balsa wood porch railings; the digni-
fied white columns made out of what appear to be plastic straws; the
scored shingles; and the Prismacolor-drawn stone and stucco facade.
My role in actually designing this house, however, was minimal – I
was only tasked with filling in the blanks of its coloring-book facade.
Its size and shape, its roof pitch, the locations of its apertures, and
its neocolonialness were all dictated by our teacher; the origins of the
design are unknown. Four years later, my first model in architecture
school speaks of a different sort of education. I was introduced to the
tools and methods with which architects create form. The model’s
stark white bristol board and foam core walls implied an eerie
absence of materiality, shunning ornament to direct attention to a
strong spatial argument. My university architectural education
was largely concerned with broad concepts, while the finer grain of
phenomenal textures often came as an afterthought – resulting in
models left smooth, white, and blank. While effective in achieving
its goal, perhaps that white model might have benefited from some
added moss, rocks, or plastic straws. – Samantha Ding

190
Shane Reiner-Roth
Internet Browser

For their 2014 multimedia art performance Art Project 2023

as part of a symposium at the Whitney Museum, artists João
Enxuto and Erica Love produced a speculative narrative cen-
tered on an economic crash in 2020 that swiftly destabilizes the
global art market. Seven years after the Metropolitan Museum
of Art’s 2015 acquisition of the Marcel Breuer–designed
Whitney Museum building on Manhattan’s Upper East Side
(changing its name to the Met Breuer), the Met board, accord-
ing to the artists, “elected not to renew its eight-year lease.”
In 2023, the multinational technology company Google pur-
chased the vacant structure to transform it into the Art Project
1.  João Enxuto and Erica Love, “Shared Museum, touted as the “first virtual museum.”1
Spaces: Social Media and Museum
Structures” (lecture, Whitney Museum,
Rather than undergo a yearslong renovation of the now
New York, January 14, 2014), available 57-year-old structure, and as a means of rebranding it as a Google
online as “Shared Spaces Symposium: João
Enxuto and Erica Love,” Vimeo video, 8:05,
product to showcase other Google products, the company chose
posted by “Whitney Museum of American to demolish and then replicate it on-site as a full-scale model with
Art,” January 31, 2014, https://vimeo.
com/85575655. For more on the piece, see the aid of a digital scanner, the original blueprints, and a subsid-
“ART PROJECT 2023,” http://theoriginal- iary 3D printer. “Google reasoned that if there was no break with
copy.net/art-project-2023.
2.  Ibid. the past, then no one would be alarmed that a new dawn was
3.  Ibid. upon us.”2 In a matter of weeks, the distinctly stacked facade and
window boxes, once clad in gray granite slabs, were re-created
as a hollow concrete sculpture (for their performance, the art-
ists printed a miniature, powder-based model of the museum).
The Art Project Museum contains no physical objects. Its
virtual artwork can only be accessed through an arsenal of
proprietary software and devices in a process that starts with
visitors logging into their Google+ accounts at the front desk
for personal data collection. Once inside the gallery space,
“with the aid of a Google Glass prototype, visitors can assem-
ble virtual walls, partitions, and pedestals that follow Breuer’s
original grid” and are “invited to curate the collection with
the help of a Google Art Scholar.”3 However, the virtual assets
spontaneously glitch, blur, and fracture until they all disappear
without a trace, eventually leaving the Art Project Museum
virtually bare – a vacant shell masquerading as the Breuer-
designed original that once stood in its place.
The artists’ parable of a rapidly reproduced museum,
emptied by a dysfunctional suite of elements required to

191
 activate its interior, satirizes Google’s treatment of the art
world through Google Arts & Culture (originally titled the
Google Art Project), an online platform launched in 2011 that
allows users to view artworks from partnering cultural orga-
nizations around the world. The exchange of users’ private
data for tailored creative content is as easily permissible to its
users as it is corrosive to the centuries-old museum typology.
“As Google Art Project continues to grow by consuming con-
tent from established institutions,” Enxuto and Love explained
in an interview, “the question remains whether host museums
will atrophy in the networked future.”4
Six years after the presentation of Art Project 2023, the
2020 economic crisis Enxuto and Love imagined came to frui-
tion (along with the Met’s dispossession of the Met Breuer)
as a result of the global spread of COVID-19, leaving cultural
institutions with the task of reevaluating their online pres-
ences to reach newly homebound audiences. Prior to the pan-
João Enxuto and Erica Love, 3D-printed demic, the Met, the Whitney, and over 2,000 other cultural
model of the Marcel Breuer–designed institutions had already established profiles on the Google
Whitney Museum of American Art from
Art Project 2023, 2013. Gypsum powder,
Arts & Culture website, supporting its mission “to preserve
6 3/4 by 6 1/4 by 7 1/2 inches. Photo and bring the world’s art and culture online so it’s accessible
courtesy the artists. to anyone, anywhere.”5
While a handful of these institutions can be walked
through virtually using Google Street View–adapted photo-
graphic panoramas, the experience is compromised by low
and inconsistent resolutions that make it a poor substitute for
physical engagement. The majority of institutions are rep-
resented, instead, as visual catalogues. The viewer can see an
image of an artwork from any given museum, with the option
to browse through a collection of artworks from the same
location or countless others in Google’s seemingly inexhaust-
ible database – searchable by time, place, medium, and color,
and always against a blank, spatially indeterminate expanse.
Like the superficially reproduced Breuer-designed museum
imagined in Art Project 2023, Google Arts & Culture reduces
the concept of the museum to a repository of images excised
from their contexts. To scrap the space in museum space, in
other words, is to replace the slow, contemplative stroll one
might take through a museum with the disembodied, poten-
tially endless scroll of a web page made free of charge by the
cultivation of user data.
As an alternative to Google’s database-as-space approach,
the contemporary and modern art gallery Hauser & Wirth
unveiled Hauser & Wirth Virtual Reality (HWVR), the first
virtual reality modeling tool of its kind to emerge from the

192 Log 50
“Beside Itself,” Hauser & Wirth, art world. Developed by ArtLab, the gallery’s recently estab-
launched April 30, 2020. Interior view lished research and innovation arm, HWVR converts Hauser
of a virtual reality exhibition sited in
a model of the future Hauser & Wirth
& Wirth’s collection into 3D assets and models its exhibition
Menorca gallery. Image courtesy spaces from the ground up without the aid of combined photos.
Hauser & Wirth. “Drawing from techniques applied in architecture, construc-
tion and video-game authorship,” Hauser & Wirth’s announce-
ment of the new research division reads, “the tool creates
true-to-life scale and accuracy, as well as the authentic look,
4.  João Enxuto and Erica Love, “See feel, and interactivity of our galleries.”6 The inaugural HWVR
Some Art While You Can – Google Will
Eventually Replace Museums,” interview
exhibition, “Beside Itself,” takes place in a digital model of
by Pete Brook, Wired, September 27, Hauser & Wirth Menorca, an 18th-century building complex
2013, https://www.wired.com/2013/09/
see-some-art-whir-you-can-google-will- on a small island in the Mahon port of Menorca, Spain. The
eventually-replace-museums/. gallery, in reality, is currently undergoing conservation work
5.  “About Google Arts & Culture,” Google
Arts & Culture, https://about.artsandcul- with an expected opening date in 2021.7
ture.google.com/. “Beside Itself ” starts in an open-air courtyard with one
6.  “ArtLab: Our new technology and
research division,” Hauser & Wirth, April 9, wall adorned with Lawrence Weiner’s BESIDE ITSELF
2020, https://www.hauserwirth.com/ (1970) and, next to it, another wall cut out to frame Louise
news/28281-artlab-new-technology-
research-division. Bourgeois’s Maman (1999) in the nearby garden. A white
7.  See “Virtual Reality Exhibition Hauser & circle on the ground by the entrance beckons the visitor into
Wirth Menorca,” Hauser & Wirth, https://
vip-hauserwirth.com/online-exhibitions/ a high-fidelity re-creation of the exhibition space, complete
hauser-wirth-menorca-in-vr/. with overhead spotlights hanging from white-painted wooden
rafters and patches of light cast on the floor from skylights.
Other white circles scattered across the polished concrete
floor become the steps one takes throughout the exhibition,
where even the smallest material aberrations are reproduced
in fine detail – the protective acrylic sheets in front of Charles
Gaines’s Librettos (2015), for instance, reflect and distort
Glenn Ligon’s silkscreen Come Out (2015) on the opposite wall.

193 Log 50
 When considering virtual reality technology’s commen-
dation as “the first step in a grand adventure into the landscape
of the imagination,” its application to the reproduction of an
ordinary exhibition space can appear to be an underwhelm-
8.  Frank Biocca, Taeyong Kim, and Mark ing use of resources.8 Yet in its insistence on the quotidian at
R. Levy, “The Vision of Virtual Reality,” in
Communication in the Age of Virtual Reality,
a time when most of the world is either unable or hesitant to
ed. Frank Biocca and Mark R. Levy (New leave their homes, HWVR demonstrates a potential of its soft-
York: Routledge, 1995), 6.
9.  Mike Pepi, “Is a Database a Museum?,”
ware platform rarely expounded upon by technologists.
in The Museum Is Not Enough, ed. Giovanna By rendering the textures of the architecture as faithfully
Borasi et al. (Montreal: Canadian Centre
for Architecture, 2019), 127. as those of its collection, HWVR rejects the scroll of the inter-
net browser for the greater material interconnectedness of
experiencing art. By modeling its affordances and limitations
after those of the building it emulates, as opposed to those of
the database in which its assets are digitally archived, HWVR
attempts to reembody its viewer using a platform that has a
tendency to do the very opposite. And unlike the in-person
experience imagined in Art Project 2023, made limitless (per-
haps exhaustingly so) by its virtual augmentation, basing the
constraints of “Beside Itself ” on those of the physical gallery
space compels the viewer to take pause and consider its cura-
torial decisions: the interplay between the artworks and the
lighting, or the isolation of a single set of paintings in a lone
corridor, or even the choice to drape a Max Bill–designed
cashmere blanket on an inconspicuous side chair.
“The museum,” writes art critic Mike Pepi, “is a symbolic
stalwart in an age of instant gratification, a commitment to
posterity in an age of presentism, and a structure that sup-
ports the value of humanity’s lone unique characteristic: the
penchant for narrative, uncertainty, and the unknowable in
the face of an algorithmic regime careening toward a society
built around optimization and measurement.”9 As COVID-19
pushes museums to increase their virtual outreach, Google
Arts & Culture and other websites with nonarchitectural spa-
tial organizations might, for good reason, increasingly become
the dominant platforms for art viewership through their easy
and inexhaustible accessibility. Yet the less popular choice to
model this outreach on the relationship between a curated set
of objects and their environmental context, even through the
relatively time-consuming processes of creating and viewing
a virtual simulation, can reaffirm the centrality of museum
spaces as sites of knowledge production.
Given the pandemic’s capacity to accelerate innovation,
and given its discontents, a stroll through “Beside Itself ” is
Shane Reiner-Roth is a PhD student at indeed a welcome respite.
the University of California, Los Angeles.

194
Tyler Survant
Modeling and
Remodeling
The Oval Office
On the evening of October 10, 1968, Gordon Bunshaft received
a phone call from the Oval Office, the official work space of
the president of the United States, located in the West Wing of
the White House. Bunshaft – a partner at Skidmore, Owings &
Merrill and already the celebrated designer of the Lever House
and Beinecke Library – was lead architect of the Lyndon
Baines Johnson (LBJ) Presidential Library, then under con-
struction on the campus of the University of Texas at Austin.
Bunshaft and his team had successfully deflected previous
attempts by First Lady “Lady Bird” Johnson to incorporate a
reproduction of the president’s office into the library’s design.
Now from that very room, President Johnson himself made
an appeal, citing the popularity of the life-size Oval Office
reproduction in Harry S. Truman’s presidential library and
the Kennedy family’s plan for a similar replica at the John F.
1.  Recording of telephone conversation Kennedy Presidential Library and Museum.1
between Lyndon B. Johnson and Gordon
Bunshaft, October 10, 1968, 8:57 pm,
Bunshaft was concerned with the propriety of an imita-
Citation #13533, Recordings and Transcripts tion Oval Office in an otherwise contemporary building. In
of Conversations and Meetings, LBJ
Presidential Library. an interview at the time, he called the drab paint color and
2. Gordon Bunshaft, “Interview I,” inter- staid furniture of the Truman replica “awful” and mused,
view by Page E. Mulhollan, Oral history
transcript, LBJ Presidential Library, June 25, “It’s really kind of nutty when you think about it, that all
1969, https://www.discoverlbj.org/item/ over the country: Oval Rooms!”2 But his primary objection
oh-bunshaftg-19690625-1-78-42.
3.  Recording of telephone conversation to a reproduction at the LBJ Library was pragmatic: with the
between Lyndon B. Johnson and Gordon Oval Office’s lofty 18-and-a-half-foot-high ceiling, he simply
Bunshaft.
didn’t have room for it between the floor slabs already being
poured on-site. President Johnson was undaunted, insist-
ing he had to have a replica even if it couldn’t be exact due
to space limitations. Few Americans get to experience the
thrill of entering the president’s office, he reasoned, and the
impression a reproduction gives visitors is far more impactful
than any text in a history book.3
On the library’s top level, unconstrained by floor-to-
floor heights, Bunshaft integrated a model of the Oval Office
at seven-eighths scale. This replica was more manageable in
size but still large enough to convey a sense of the original

195
Eric Gugler, Oval Office, The White space. Johnson would use the re-creation as a private work-
House, Washington, DC, 1934. Right: space for two years – from 1971, when the LBJ Presidential
Gordon Bunshaft, Oval Office replica,
Lyndon Baines Johnson Presidential
Library opened, until his death in 1973, when it became a
Library, Austin, Texas, 1971. Drawing: public exhibit. Despite the replica’s smaller stature – 18 inches
Tyler Survant. shorter than the original – its elliptical walls and domed ceil-
ing still project above the library’s modernist flat roof and are
visible even from certain vantage points on the ground, eight
stories below.

Replications
At the time, Johnson’s Oval Office was one of only a few
copies, but every US president since has commissioned a
full-scale replica of the room as furnished during his ten-
ure (all US presidents thus far have been men). Today, not
only does a president remodel the Oval Office when he
assumes the presidency but upon leaving office only four or
eight years later, he then models his remodel at 1:1 scale for
posterity. Ten complete Oval Office replicas (and two par-
tial ones) now appear in presidential libraries across the US.
Yet another is planned for the Obama Presidential Center in
Chicago, designed by Tod Williams Billie Tsien Architects.
Simultaneously political and biographical, presidential
libraries are equal parts archive, museum, and monument.
They serve as spaces of scholarship and public education as
well as sites of commemoration and propaganda.
Countless other Oval Office re-creations (with varying
degrees of accuracy) also dot the country, housed in private
homes, television and film studios, and public museums such
as the New-York Historical Society, which, as recently as

196 Log 50
 February 2020, and at a cost of two million dollars, debuted
its own replica modeled on Ronald Reagan’s second term in
4.  See Charles Passy, “New-York Historical office.4 The reproductions in presidential libraries, however,
Society Replicates Reagan’s Oval Office,”
Wall Street Journal, February 12, 2020,
stand out for their paradoxical authenticity: they are authentic
https://www.wsj.com/articles/new-york- fakes, so to speak, commissioned by former presidents, located
historical-society-replicates-reagans-oval-
office-11581550930. in buildings administered by the National Archives and
5.  See Judy Pesek, “Recreating the Oval Records Administration, and open to the general public. These
Office at the George W. Bush Presidential
Center,” GenslerOnCities, April 25, 2013, replicas are particularly uncanny because the line between
https://www.gensleron.com/cit- original and facsimile, genuine and fake, is often impossible
ies/2013/4/25/recreating-the-oval-office-
at-the-george-w-bush-presidential.html. for a visitor to establish. Faux furniture may intermingle with
6.  Ibid. the authentic pieces actually used by a president. LBJ’s replica
7.  Glenn Kessler, Salvador Rizzo, and
Meg Kelly, “President Trump has made is furnished with the president’s original mahogany desk, for
more than 20,000 false or misleading instance, but also with copies of other furnishings and art-
claims,” Washington Post, July 13, 2020,
https://www.washingtonpost.com/ work. Nothing less than the original or its exact imitation will
politics/2020/07/13/president-trump- do. Designers at Gensler, tasked with the interiors for Robert
has-made-more-than-20000-false-or-
misleading-claims/. A.M. Stern Architect’s George W. Bush Presidential Center in
Dallas, worked with vendors for a year and a half, through
five different mock-ups, to re-create the oval rug designed by
First Lady Laura Bush for her husband’s office.5 Gensler’s goal
for their reproduction was pure verisimilitude: “a replica that
even seasoned Secret Service officers have a hard time distin-
guishing from the original.”6
Like the 1:1 scale map of an empire in Jorge Luis
Borges’s short story “On Exactitude in Science,” Oval Office
replicas confound the map-territory relation, replacing the
original with its highly detailed, life-size representation.
The effect is one of “hyperreality,” identified by postmod-
ernist philosophers and historians as a melding of reality and
illusion where fact and fiction commingle. Perhaps it should
come as no surprise, then, that in 2017 the Oval Office would
be occupied by a reality TV star who has since run up a tab of
more than 20,000 false or misleading statements as president
– a “tsunami of untruths,” according to Washington Post
fact-checkers.7
Citing LBJ’s library in his 1973 essay “Faith in Fakes,”
later republished as “Travels in Hyperreality,” Umberto Eco
identified an American obsession with what he called the
“authentic copy” or the “Absolute Fake.” Eco cast Johnson’s
Oval Office replica in the same light as Disneyland, a mani-
festation of America’s penchant for illusion, fantasy, and
simulated reality, and observed that “constructing a full-scale
model of the Oval Office (using the same materials, the same
colors, but with everything obviously more polished, shin-
ier, protected against deterioration) means that for histori-
cal information to be absorbed, it has to assume the aspect

197 Log 50
8.  Umberto Eco, “Travels in Hyperreality,” of a reincarnation.”8 Indeed, the rise of Oval Office replicas
in Travels in Hyperreality: Essays, trans.
William Weaver (New York: Harcourt,
has paralleled the White House’s increasing historical con-
1983), 6–7. servation and curation. In the early 1960s, the building itself
9.  See US Public Law 87-286 and Executive
Order 11145.
became a historic house museum with the establishment of
10.  Eco, 53. the White House Office of the Curator, Committee for the
11.  Daniel J. Boorstin, The Image: A Guide
to Pseudo-Events in America (New York:
Preservation of the White House, and White House Historical
Vintage Books, 1992), 10. Association, as well as the commitment of the Smithsonian
Institution to care for all White House furniture, fixtures, and
decorative objects not in active use.9 Though the West Wing
is not held to the same standards of preservation as the State
Floor of the White House, presidents nevertheless work with
these committees and institutions to furnish their offices.
Ironically, then, Oval Office reproductions are steeped in the
historicity of the White House’s collection of furniture and
fine art but must project a sense of authenticity through imi-
tation pieces instead. While these replicas serve to honor the
past and commemorate the accomplishments of their presi-
dents, given their historical “look and feel” they may also
excuse their misdeeds, which can be explained away – like the
dated decor – as just a sign of the times.
Eco points to the geographic immensity of the US as a con-
tributing factor for its production of historical re-creations,
suggesting that “it is the spatial, as well as the temporal, dis-
tance that drives this country to construct not only imitations
of the past and of exotic lands but also imitations of itself.”10
With iterations spanning the country, Oval Office replicas
distribute and democratize access to the presidency, perform-
ing the premise of democracy – rule by the people rather
than by the elite – better than the real Oval Office itself (fre-
quently occupied by men of wealth and privilege). For most
Americans, an Oval Office model is an ersatz good, a substi-
tute for the real thing.
The increasing replication of the Oval Office also paral-
lels the growing power and prestige of the presidency in the
latter half of the 20th century and its consummation through
electronic media. In his 1962 classic The Image: A Guide to
Pseudo-Events in America on news production, celebrity cul-
ture, and other “bewitching unrealities of American life,”
political historian Daniel Boorstin ties increasing presiden-
tial power to the rise of broadcast news media, particularly
during the administration of Franklin D. Roosevelt – inci-
dentally the first occupant of the modern Oval Office and cre-
ator of the first presidential library.11 Photographer Nicolas
Grospierre, who documented Oval Office replicas around the
country in 2013, similarly points to the film industry and the

198 Log 50
Students take selfies in the replica
Oval Office, Lyndon Baines Johnson
Presidential Library, Austin, Texas,
March 5, 2016. Photo: Jay Godwin, LBJ
Presidential Library.

statistical correlation (if not a causal one) between the grow-

ing number of these reproductions and increasing airtime of
12.  See Nicolas Grospierre, “The Oval Oval Office film sets.12 Given its simultaneous physical and
Offices,” https://www.grospierre.art.pl/
portfolio/the-oval-offices/.
virtual proliferation, Grospierre concludes that the room
13.  Ibid. has “undergone a process of complete spectacularisation. It
14.  See Maureen Dowd, “Mr. Smith Went
to Washington . . . With Dimmers,” New
thus seems that the Oval Office has ceased to be a real place,
York Times, August 26, 2020, https://www. the actual room where the President works, and has in fact
nytimes.com/2020/08/26/style/michael-
smith-decorator-obamas-cindy-crawford-
become a symbolic set, the theater of power, quite literally.”13
rupert-murdoch.html. This observation is bolstered by more recent Oval Office
re-creations where visitors become actors in the space, in con-
trast to early exhibitions like LBJ’s, which are roped off like
artwork or artifacts – treasures to behold but not to touch.
Instagram and Flickr are filled with countless snapshots of
presidential library tourists in postures of mock seriousness,
often seated at a faux Resolute Desk, taking an important
call on the desktop telephone. Interactive replicas function
as presidential simulations, offering visitors an entertaining,
first-person perspective from the seat of American politi-
cal power. As a backdrop for selfies and social media posts,
they have a theatrical aspect not unlike the Oval Office sets
in the studios of Warner Bros., ABC, or YouTube. The White
House work of Los Angeles–based decorator Michael Smith,
who drew inspiration for Obama’s 2010 office makeover from
movies such as Dave (1993) and The American President (1995)
– among the first films to feature Oval Office sets – confirms
Hollywood’s influence.14
The spectacular character of the Oval Office means
the political performativity of the space is more essential
than its performance as a traditional workplace. In fact, for
Donald Trump, the “room where it happened” is often not
the Oval Office: he is said to prefer working in the White

199 Log 50
Oval Office furnishings since 1934.
Drawing: Tyler Survant.

House residence, away from the eyes of the press and junior
staff. For example, the president’s now infamous July 2019
phone call with Ukrainian President Volodymyr Zelensky
– a conversation that led to Trump’s impeachment – took
place upstairs in the private residence rather than in the
Oval Office, despite occurring during regular business hours.
Under Trump, the room has become less an office than a stu-
dio for photo ops. As Grospierre suggests, “It is more impor-
tant for the Oval Office to be shown as the place where the
public/the voters can imagine that power is being exercised,
15.  Grospierre. than what actually happens in reality.”15 In a society of the
16.  Associated Press, “Oval Office
Renovation Accents Symbols,” Los Angeles
spectacle, appearing trumps being.
Times, September 4, 1993, https://www.
latimes.com/archives/la-xpm-1993-09-
04-mn-31522-story.html.
Remodels and Role Models
Remodeling the Oval Office affords each new occupant of the
West Wing the opportunity to remodel the image of the presi-
dency. As then White House Curator Rex Scouten explained
after Bill Clinton’s renovation in 1993, “Most presidents come
in proposing change, and want their own offices to reflect
what they feel. They want to project their own image.”16
Regardless of their political ideology and party affiliation,
however, modern presidents have been conservative when it
comes to Oval Office remodeling, largely hewing to tradition
and privileging continuity over change. Subtle variations of
decor aside, the general placement of the furniture – a desk
at one end, in front of the windows, and a seating arrange-
ment at the other, in front of the fireplace – has been more
or less consistent since 1934, when Eric Gugler conceived the
modern Oval (itself modeled on Nathan C. Wyeth’s original
elliptical office design from 1909). Furthermore, several
objects, through their increasingly ritual reuse, have earned

200 Log 50
 the status of fixtures in the room: potted Swedish ivy has
sat on the mantel since 1961 (incidentally, the same year the
White House was declared a museum); a Seymour tall case
clock, now known as the Oval Office Grandfather Clock, has
remained in the space since the administration of Gerald
Ford; the Resolute Desk has come and gone since 1961 but has
stayed put since Clinton brought it back to the Oval in 1993.
While Trump has broken countless norms (and argu-
17.  See Andrew Weissmann et al., “Federal ably several laws17), in terms of interior design he has been
Criminal Offenses and the Impeachment
of Donald J. Trump,” ed. Ryan Goodman,
a model president, reinforcing the classic Oval Office “brand
Just Security, Reiss Center on Law and identity” by recycling the furnishings of his predecessors.
Security at New York University School
of Law, December 16, 2019, https://www. In a way, Trump’s office is already something of a replica.
justsecurity.org/67738/federal-criminal- His sofas, upholstered in a cream brocade pattern, are bor-
offenses-and-the-impeachment-of-
donald-j-trump/. rowed from George W. Bush’s Oval Office, and, in turn,
evoke LBJ and Reagan’s sofas. Trump’s gold drapery and
valences, matching yellow chairs, and porcelain table lamps
are from Clinton’s office. His sunburst oval rug – a furnish-
ing often commissioned anew by each successive president –
is Reagan’s bespoke design. When Trump unveiled his Oval
remodel in August 2017, several items stirred controversy for
their interpretation as coded political statements. Yet even
these were first displayed in the offices of past presidents.
Trump’s portrait of Andrew Jackson, a president infamous
for his support of slavery and treatment of Native Americans,
previously hung in Clinton’s office, among others. His US
military battle flags, arrayed ceremoniously around the room,
once graced LBJ and Richard Nixon’s offices. Contrary to
initial reporting, Trump kept a bronze bust of Martin Luther
King Jr., prominently displayed in President Obama’s office,
while adding to the space a bust of Winston Churchill. George
W. Bush was the first to display these sculptures in the Oval
Office. Only Trump’s wallpaper, a muted, floral damask pat-
tern with a repeating baroque rosette, is a major (if subtle)
update to the room. And yet, consciously or not, even this
choice was mimetic, an instance of life imitating art: another
off-white, damask wallpaper first premiered in September
2016 in the Oval Office of fictional president Tom Kirkman
from political thriller Designated Survivor.
Trump is certainly not the first to remodel the Oval on
historical precedent, but casting himself in the mold of past
presidents is especially fitting for a man evidently enamored
with the more ritualistic trappings of the office and sensitive
about his legitimacy (having lost the popular vote). His
design decisions also reflect an acute understanding of the
space’s symbolic power in mainstream media. For some past

201 Log 50
 presidents, at least, one perceives the Oval Office decor, with
its patriotic allusions and historical significance, as aspira-
tional: a call to moral leadership during national crises. For
Trump, however, decor and decorum are uncoupled. He has
continually defied the expectation that a president should
model virtuous behavior. George W. Bush, referring to the
intense public scrutiny and burden of presidential leadership,
once quipped that the Oval Office is round so there are no
corners to hide in. For Trump, the office itself is a rhetorical
refuge; its presidential appearance acts as a kind of red her-
ring, obfuscating his unpresidential conduct.

The Treachery of Images

Nearly 60 years before the Trump administration tear-gassed
peaceful Black Lives Matter protesters outside the White
House to clear the way for a photo op, Boorstin put forward
an early critique of the power of news and television images to
shape American politics, especially through “pseudo-events”
– synthetic media events rather than natural occurrences,
staged by politicians or the public relations industry for the
sole purpose of being reproduced. At that time, the great-
est threat to the United States, according to Boorstin, was the
“menace of unreality” or “the danger of replacing American
dreams by American illusions. Of replacing the ideals by the
18.  Boorstin, 240. images, the aspiration by the mold.”18
19.  Ibid., 43.
Writing shortly after the 1960 debate between Nixon and
20.  Ibid., 44.
Kennedy – the first ever presidential debate and also the first
televised debate, in which the candidates’ contrasting appear-
ances became national news – Boorstin critiqued the “quiz
show” format that privileged style over substance. Instead
of establishing relevant qualifications or policy differences,
he observed, the debate merely tested the “relative capacity
of the two candidates to perform under television stress.”19
Boorstin warned that such media spectacles were self-ful-
filling prophesies that could shape history. “In a democ-
racy,” he said, “reality tends to conform to the pseudo-event.
Nature imitates art.”20 Pseudo-events, he concluded, lead to
pseudo-qualifications.
By a similar logic, Oval Office replicas, as historicized,
synthetic spaces contrived for presidential aggrandizement,
model the office Trump has occupied these past four years.

Tyler Survant is an architect based

in New York.

202
Rizal Muslimin
Computational
Puppets
In March 2020, numerous simulations of the COVID-19 pan-
demic quickly emerged, recalling previous epidemic simula-
tors for malaria, SARS, and avian influenza in the last decade.1
To simulate virus transmission, people are represented as
points traveling and interacting in two-dimensional space
and identified as suspected agents (S), infected agents (I), and
recovered agents (R). This SIR model can then represent dif-
ferent states of each person’s condition based on their inter-
action with each other; as I collides with S, S turns into I, and
1.  See Harry Stevens, “Why outbreaks like I either dies or recovers (R).2 Recent simulations show, for
coronavirus spread exponentially, and how
to ‘flatten the curve,’” Washington Post,
example, that concentrated activities – such as buying grocer-
March 14, 2020, https://www.wash- ies and partying – are likely to cause many collisions between
ingtonpost.com/graphics/2020/world/
corona-simulator/. agents, increasing the number of I. These simulations are
2.  See Robert J. Gallop, “Modeling General zero-sum and highly oversimplified, but even in their most
Epidemics: SIR Model,” in Proceedings of
the 12th Annual NorthEast SAS Users Group abstract form, seeing the points interact in the simulator
Conference, Washington, DC, 1999. makes us more aware of how our behavior in the built envi-
3.  See Dirk Helbing, “A Mathematical
Model for the Behavior of Pedestrians,” ronment can place us in vulnerable positions.
Behavioral Science 36, no. 4 (October 1991): In architecture, agent-based modeling (ABM) has been
298–310; Michael Batty and Bin Jiang,
“Multi-Agent Simulation: New Approaches widely used by designers, planners, and engineers to model
to Exploring Space-Time Dynamics within an occupant’s behaviors, from exiting a stadium, escaping
GIS” (CASA Working Paper 10, The Bartlett
Centre for Advanced Spatial Analysis, a fire, or crossing a street to viewing a painting in a gallery
University College London, London, 1999). or window shopping at a mall.3 The agent in these models is
4.  See Craig W. Reynolds, “Flocks, Herds,
and Schools: A Distributed Behavioral analogous to a crash test dummy, only more active, virtual,
Model,” Computer Graphics 21, no. 4 (July and with a lower risk of impact. Though the modeled occu-
1987): 25–34.
pant typically has attributes – behavioral rules, memory, a
goal, detailed humanoid articulation, a self-learning capacity
– it can also simply be a point in space, moved iteratively by a
vector to simulate a certain behavior. We can easily track the
space these agents use and the time they spend on walking,
stopping, and reorienting in space. They work autonomously,
in a discrete manner, yet collectively they can portray social
behaviors. As demonstrated by Craig Reynolds in his 1986
artificial life program Boids, complex processes and behavior
such as flocking, avoiding, and steering in a swarm of fish or
birds can be simulated through ABM with simple rules.4 From
a two-dimensional grid, as in John Conway’s Game of Life, to
a volumetric environment, as in Reynolds’s Boids, the space
that the agents occupy is metamorphic in representation and

203
 translation. An agent can serve as anything from a person or
a vehicle to a smoke particle or a currency in the exchange
market. Indeed, agents are easily shaped in Euclidean space
and can fluidly adopt different contexts across scales and dis-
ciplines through changes in their variables.
In use for over 35 years, ABM has become sophisticated,
particularly in the gaming and filmmaking industries, where
the agents are given myriad abilities, tasks, and apparent
flaws to make them seem more human. They “see” through
programmed lines of sight, “hear” according to the values of
sound frequencies, and “move” along vectors governed by the
5. See Kirsten Moana Thompson, “Scale, laws of physics.5 With just a few character templates custom-
Spectacle and Movement: Massive Software
and Digital Special Effects in The Lord of the
ized through randomly applied variables, thousands of varia-
Rings,” in From Hobbits to Hollywood: Essays tions can be created to “act” in a carefully choreographed
on Peter Jackson’s Lord of the Rings, ed. Ernest
Mathijs and Murray Pomerance (Amsterdam; digital environment. Today, with deep learning algorithms,
New York: Rodopi, 2006), 283–99. agents can make an actor look younger and even assist one to
6.  See Mustafa Yarımbaş, “Netflix vs
Deepfake: The Irishman,” Predict, Medium, “act” posthumously.6
September 9, 2020, https://medium. Despite the advancements made in these fields, and the
com/predict/netflix-vs-deepfake-the-
irishman-1d4754de2701; Paul Kavanagh complexity of problems in the built environment, ABM is not
and Roger Guyett, “Star Wars: How We as progressive in architecture as in other industries. In ABM
Brought Carrie Fisher Back,” interview by
BBC, YouTube video, 4:03, posted by “BBC today, building users’ behavior is comparable to Reynolds’s
Click,” February 6, 2020, https://www. early Boids or Michael Batty’s agents in the late ’90s: they
youtube.com/watch?v=bDa5h3cj1DU.
walk from A to B. This simulation is important, but are users’
activities always so simple? Don’t users deserve more than to
be implied by the furniture in a floor plan or the entourage in
a perspective typically Photoshopped in at the end?
Storytelling is probably not the architect’s primary pur-
suit, but it is the architect who places walls, floors, stairs,
doors, roofs, and windows to preordain the users’ activities
and behaviors. To that extent, the architect is a puppeteer,
controlling the users’ access and obstacles – usually from a
top-down view – and hoping that the users will live as the
designer imagined.
It is this storytelling that needs to be refined in archi-
tectural ABM, especially in light of the current global pan-
demic. As people rethink their everyday routines and adapt to
online platforms, the role of the building is becoming more
contested and often challenged. For instance, WeWork, once
a model for repurposing old buildings with space-planning
algorithms for collective work spaces, is now struggling to
survive. Working and shopping from home are also rendering
the mainstream workplace and retail architecture increas-
ingly obsolete. As buildings are adapted for new practices such
as social distancing, there is a rising expectation of carefully
considering safety metrics in the same regard as structural

204 Log 50
Pierre Huyghe, This is not a time for precision and environmental appropriation. An improved
dreaming (still), 2004. Live puppet play architectural ABM is needed more than ever to assess how
and Super 16 mm film transferred to
DigiBeta (24 minutes; color, sound).
buildings meet the users’ needs with greater accuracy.
Photo: Michael Vahrenwald. Courtesy Of course, a person is much more than a traveling point
Carpenter Center for the Visual Arts on a screen. Nevertheless, we are constantly being unitized by
and Marian Goodman Gallery, New York.
others – as a number in a spreadsheet, a vote in an election, a
meal for “table 8,” or a chair in a floor plan. On social media,
everything is either liked or disliked, swiped left or right. We
generalize and simplify to process information and make deci-
sions, which makes us – even for a split second – comparable
to how a computer works. Perhaps that is enough to justify the
use of programmable agents for simulating multiple design
issues. To be more resilient in today’s climate, architectural
ABM must embrace diversity in design by involving factors
that were previously external to the simulation process, such
as public health, social values, and consumer trends. Now is
the time for architecture to become a test bed for new kinds of
programs, creative business models, solutions to public con-
cerns, and behaviors that emerge from letting agents interact.
If a pandemic chart from a point-based ABM was enough for
world leaders to close borders and put the economy on hold,
then there must be more that can be discovered from using
Rizal Muslimin studied design compu-
tation at MIT and teaches architectural more complex models in architecture.
computing at the University of Sydney.

205
Alice Loumeau
The World Model

Figure 26 The World Model, diagram In the summer of 1970, a team of international researchers at
and caption, from Donella H. Meadows MIT, commissioned by the Club of Rome, analyzed five basic
et al., The Limits to Growth: A Report
for the Club of Rome’s Project on the
factors responsible for shaping society: population growth,
Predicament of Mankind, 1972. agricultural production, nonrenewable resource depletion,
industrial output, and pollution generation. The interde-
pendencies of these five factors were represented and stud-
ied in a system dynamics model, called The World Model,
designed to integrate feedback loops from the various com-
ponents that affect the five factors – such as governmental
policies, changes in regimes, social trends, corrective mea-
sures, and bureaucratic delays – with the understanding that
it would constantly evolve along with changes in society. The
researchers came to the conclusion that the way we imagine
life on Earth is not sustainable and that the people in power
have to act quickly. They reasoned that humanity must move
away from its preoccupation with growth in order to avoid
scenarios that could lead to ecological collapse.
While The World Model did not anticipate some of
today’s major concerns – most notably the traces and impacts
of climate change – this report established the foundation for
a holistic approach to thinking about the imprint of human
behavior on the environment. It demonstrated how an action
or decision in one part of the world can trigger consequences
in another, relinking cause and effect despite time and dis-
tance. Widely overlooked – perhaps to legitimize global inac-
tion – this model should be reread today as an example of
faith in human ingenuity and the effort to pursue greater
planetary equilibrium.

Alice Loumeau is a practicing architect

and researcher at OMA, Rotterdam.

207
Before-and-after crime scene models presented in the courtroom during the
Ayyash et al. trial, Special Tribunal for Lebanon, The Hague, the Netherlands,
January 17, 2014. Photos courtesy Special Tribunal for Lebanon.
Jason Rhys Parry
Simulating an
Assassination: Four
Explosion Models
Before and After
Two scale models dominated the courtroom of the Special
Tribunal for Lebanon in The Hague on the morning of January
16, 2014. Both were models of the Rue Minet al Hosn in Beirut.
Precisely, the models depicted the street as it looked nearly a
1.  See Owen Bowcott, “Hariri killing decade earlier on February 14, 2005.1 The level of detail was
street-scene model dominates trial
courtroom,” Guardian, January 16,
exquisite. Small shrubs poked out of miniature planters, and
2014, https://www.theguardian.com/ weeds dotted a disused sidewalk. Despite the fact that they rep-
world/2014/jan/16/hariri-killing-street-
scene-model-trial-courtroom. resented the same street, the differences between the models
2.  See Steffen Siegel, ed., First Exposures: were stark. Where one featured an orderly line of parked cars,
Writings from the Beginning of Photography
(Los Angeles: The J. Paul Getty Museum, the other presented a pile of charred chassis. Modeled detri-
2017), 57, 282. tus filled the middle of one street, while the other showed a
smoothly paved surface. The models framed a key moment
in the history of modern Beirut: the massive explosion that
resulted in the death of the former prime minister of Lebanon,
Rafik Hariri, along with 21 others. Following preliminary inves-
tigations carried out by the United Nations and the Lebanese
police, the Special Tribunal for Lebanon was convened to deter-
mine the identities of those responsible for the assassination.
The tribunal sought to discover what happened in the moments
marking the transition from one model to the other.
These two scale models formed a classic before-and-after
dyad, a type of visual rhetoric dating at least to the origins of
photography. In 1839, Louis Daguerre, eager to earn recogni-
tion for his new invention, the daguerreotype, presented King
Ludwig I of Bavaria two views of the Boulevard du Temple in
Paris, taken at different times from the window of his studio.2
Documented in the two images were subtle changes in the
atmosphere registered by the play of light and shadow across
the street. Indeed, Daguerre’s experiments in capturing the
ephemeral changes of sunlight in a single space preceded the
debut exhibition of the impressionists at Nadar’s studio by
three and a half decades.
The before-and-after strategy would go on to find many
applications, from documenting the effects of treatment on

209
 mid-19th-century psychiatric patients to highlighting the
3. On the use of before-and-after photogra- efficacy of diet pills and exercise equipment.3 But, in each
phy in 19th-century psychiatry, see Susan
Sidlauskas, “Before and After: The Aesthetic
case, the crucial event – the moment of transformation –
as Evidence in Nineteenth-Century remains unrepresented. It is attested to only by its traces. As
Medical Photography,” in Before-and-After
Photography: Histories and Contexts, ed. Jordan Eyal and Ines Weizman write: “The gap between before-and-
Bear and Kate Palmer Albers (London: after images might also be considered as a reservoir of imag-
Bloomsbury, 2017), 15–42.
4. See Eyal Weizman and Ines Weizman, ined images and possible histories.”4 Although they testify
Before and After: Documenting the Architecture to a transformation, establishing a clear basis for assessing
of Disaster (Moscow: Strelka Press, 2014), 10.
5.  See Public Transcript of the Hearing held changes occurring at some point between an earlier moment
on 16 January 2014 in the Case of Ayyash et al., and a later one, the precise cause of these changes requires
STL-11-01, Special Tribunal for Lebanon,
January 16, 2014 (opening statements). additional forms of explanation.
On that day in the tribunal courtroom, the additional
explanation came in the lengthy statement given by Senior
Trial Counsel Alexander Milne. While summarizing Hariri’s
trajectory through the city on the day of the explosion, Milne
referred to photographs and security camera footage docu-
menting the movements of Hariri’s motorcade as well as those
of a white van carrying over a ton of explosives. At the point
when the paths of the motorcade and the van intersected out-
side the St. Georges Hotel, Milne drew the court’s attention
to the preblast model.5 The more prominent buildings and
many of the individual vehicles in the model had been identi-
fied during the course of Milne’s statement, but when Milne
arrived at the point in the retelling at which the preblast
model described the moment of the blast itself, the model
ceased to figure into the narrative. The preblast model’s util-
ity was exhausted at that very instant in the presentation.
The force of the explosion disabled the camera that had
been tracking the van. The camera, which provided the visual
material for the preblast model, was destroyed at the same
time that the preblast model ceased to describe the conditions
of the street. Indeed, Milne turned the court’s attention away
from the scene, to another location – using a recording of a
parliamentary address one kilometer away, which was also
disrupted by the explosion. Earthquake-detecting equipment
15 kilometers away sensed the shock wave; a seismological sta-
tion 60 kilometers away detected it as well. Clearly, the scale of
the event exceeded the limited purview of the pre- and post-
blast models. But what scale model could contain such a blast?

A Controlled Explosive Experiment

At 5:00 pm on October 19, 2010, a “controlled explosive experi-
ment” took place at the Captieux military base in France. The
experiment and the necessary preparations were carried out in
secrecy. Although an official bulletin published by the tribunal

210 Log 50
 denied that the experiment constituted a full-scale reconstruc-
tion of the assassination, leaked details of the event published by
6.  “Hariri probe team re-enacts blast,” the newspaper Le Figaro hinted at the experiment’s vast size.6
Al Jazeera, October 20, 2010, https://
www.aljazeera.com/news/middlee-
Effectively, an entire section of Beirut was rebuilt in south-
ast/2010/10/2010102002449814660.html. western France – a curious event given Beirut’s historic claim
7.  See Georges Malbrunot, “L’assassinat
d’Hariri reconstitué en secret,” Le Figaro,
to being the “Paris of the East.” Replicas of the walls of the St.
October 19, 2010, https://www.lefigaro. Georges Hotel, the HSBC Bank, and several adjacent houses
fr/international/2010/10/19/01003-
20101019ARTFIG00560-l-assassinat-d- were constructed at the French military base so as to reproduce
hariri-reconstitue-en-secret.php. the effects of the explosion as closely as possible. Identical mod-
8. See Lindsay Freeman, “Digital Evidence
and War Crimes Prosecutions: The Impact els of Hariri’s armored car and the bomber’s van were brought
of Digital Technologies on International to the scene, and soil was even shipped from Beirut to repro-
Criminal Investigations and Trials,” Fordham
International Law Journal 41, no. 2 (2018): 311. duce the blast’s effect on Lebanese earth.7 The exercise was to
9.  Kazi K. Ashraf, “Post Mortem: Building assess the validity of the charge put forward by the defense that
Destruction,” in Architecture Post Mortem:
The Diastolic Architecture of Decline, the explosion was not caused by a bomb contained in a van but
Dystopia, and Death, ed. Donald Kunze, was the result of a bomb planted underground.8
David Bertolin, and Simone Brott
(London: Routledge, 2016), 84. The mushroom cloud could be seen from a distance, but
reporters were not allowed to see the reconstruction itself. The
Captieux reconstruction was even heavily pixelated on Google
Earth. Such mystery only enhanced the ritualistic elements of
the experiment. Considering the immense expense and effort
consumed in the reconstruction and destruction of the scene
at Captieux gives one a sense that the quest for forensic exac-
titude began to blur into Dionysiac excess. There is something
theatrical about modeling at urban scale. When the destruc-
tion of a whole street is being simulated, the gap between
scenography and science narrows. The model becomes a film
set and the film set’s destruction becomes a spectacular form
of evidence. At the end of Andrei Tarkovsky’s 1986 film The
Sacrifice, a house burns down to the swelling sounds of Bach’s
St. Matthew Passion. During the filming, Tarkovsky’s cam-
era equipment jammed as the house was engulfed. Distraught
but determined, Tarkovsky had to rebuild the house and film
it burning down again.9 A similar fate awaited the Lebanese
artist Marwan Rechmaoui’s Monument for the Living, a
2.3-meter-tall model of the Burj al Murr skyscraper in Beirut.
Left unfinished at the outbreak of civil war, the concrete
skeleton of the structure became notorious as a sniper’s nest.
Despite the cessation of violence, the unfinished building still
stands – too dense and tall to be demolished safely. Rechmaoui
completed Monument for the Living in 2000, but, tragically,
it was ruined in transit during a European exhibition tour.
Over the next several years, Rechmaoui built a replica of his
model of Beirut’s incomplete tower: a model of a destroyed
model of an unfinished original. Art history is littered with
lost originals and surviving copies of copies, but the tribunal’s

211 Log 50
 reconstruction at Captieux was more ambitious still. While
ancient regimes often hosted celebrations reenacting the
founding of cities, this was a grand reenactment of an act of
10.  For example, a statue of the Tyche of partial urbicide.10
Constantinople was brought forth each
year to celebrate the city’s birthday. See
Captieux had several qualities that made it an attractive
Clifford Ando, The Matter of the Gods: locale for the experiment. Not only did France support the
Religion and the Roman Empire (Berkeley:
University of California Press, 2008), 191. work of the tribunal but there are not many places in Europe
11.  See Kenneth H. Olsen, “Seismological where one can safely detonate over 2,000 pounds of TNT
Monitoring of the World’s First Nuclear
Detonation – The Trinity Shot of 16 equivalent. As it planned the reconstruction and destruction
July 1945,” in Geophysics in the Affairs of a city street, the tribunal was forced to select a zone deemed
of Mankind: A Personalized History of
Exploration Geophysics, 2nd ed., ed. L.C. safe for destruction – a geography of approved explodability.
(Lee) Lawyer, Charles C. Bates, and Robert This process of selection echoed other significant moments
B. Rice (Tulsa: Society of Exploration
Geophysicists, 2001), 355–58. in the history of controlled explosions, including the famous
12.  See Prosecution Motion regarding Trinity nuclear test in 1945. Spanish conquistadores named
Upcoming Expert Witnesses, Case of Ayyash
et al., STL-11-01, Special Tribunal for the region where the first nuclear test occurred Jornada del
Lebanon, May 26, 2014. Muerto due to its hostility to travelers – a hostility that would
13.  Marie-Louise Gumuchian and Saad
Abedine, “Lebanon’s Hariri trial opens be compounded in the 20th century by radioactive fallout.
without the accused,” CNN, January 17, Eager to discover the extent of the shock wave produced by
2014, https://www.cnn.com/2014/01/16/
world/meast/lebanon-violence/index. the nuclear device, the military brought in a Harvard seismol-
html; Andrew Coombes, “Split remains
over Hariri tribunal,” Al Jazeera, March
ogist, L. Don Leet, to arrange monitoring of atmospheric and
5, 2009, https://www.aljazeera.com/ seismic waves resulting from the test. During the blast, a seis-
focus/2009/03/2009337554205393.html.
14.  Antonio Cassese, The Human Dimension
mic station eight kilometers away from the test site detected
of International Law: Selected Papers a novel kind of seismic wave that Leet dubbed the “hydro-
(Oxford: Oxford University Press, 2008),
425. Emphasis original. dynamic (H) wave.”11 This wave type was unusual due to its
15.  Ibid. similarity to the waves produced in water. Effectively, the
force of the explosion had made the Earth’s surface ripple like
a liquid. Leet’s analysis of these waves helped form the basis
of the discipline of nuclear explosion seismology.
At the Special Tribunal for Lebanon, seismic data was
entered as evidence, and the tribunal accepted a report from
a geophysicist named Iskandar Sursock.12 But seismology
also provided metaphors for the tribunal at multiple lev-
els. Hariri’s assassination was likened by media outlets to a
“seismic event” in Lebanon’s history and described as the
cause of “seismic shifts” in Middle East politics.13 But in an
abstract sense, the activity of the tribunal in the landscape
of international geopolitics was also laden with seismic sig-
nificance. Judge Antonio Cassese – future president of the
Special Tribunal for Lebanon – once described international
law as “an edifice built on a volcano – state sovereignty.”14 For
Cassese, state sovereignty is an unstable force that “explodes
onto the international scene” from time to time, demolish-
ing “the very bricks and mortar from which the Law of
Nations is built.”15 International laws function as “devices”
for withstanding the “seismic activity of states,” and among

212 Log 50
Judges examine the postblast model
during opening statements by the
prosecutor, January 16, 2014. Photo:
Toussaint Kluiters. © Reuters.

the most prominent of these devices is the international tri-

bunal, which “must always contend with the violent erup-
16.  Ibid. tions of state sovereignty.”16 Although the Special Tribunal
17. See Karinne Coombes, “Universal
Jurisdiction: A Means to End Impunity or a
for Lebanon was called into being to determine the identities
Threat to Friendly International Relations?,” of those actors behind the lethal explosion of 2005, its charge
The George Washington International Law
Review 43, no. 3 (2011): 419–66.
could be undermined at any moment by the seismic activity
of states on whose support it depended. The formidable logis-
tics of rebuilding a destroyed section of Beirut in France, only
to destroy it again, hint at the international coordination that
was simultaneously the tribunal’s precondition for existence
and its greatest vulnerability.
In this sense, not only has the Special Tribunal for
Lebanon catalyzed the construction of models but the inter-
national tribunal itself is a model of universal jurisdiction. A
highly controversial concept in international law, universal
jurisdiction refers to the idea that crimes can be prosecuted by
any nation, regardless of where the crimes occurred or the
relationship between the accuser and the accused. Proponents
of the idea describe it as a means to end impunity, but as the
Special Tribunal for Lebanon makes clear, such a promise
must contend with the realities of the uneven geographies of
power.17 All of the individuals accused by the tribunal remain
fugitives, granted safe havens in places (many of them inside
Lebanon) where the tribunal’s authority does not reach.

Waves in the Machine

Among the hundreds of officials present at the controlled
explosion at Captieux was an Argentinian engineer named
Daniel Ambrosini. Five years later, Ambrosini would offer
testimony at The Hague to describe a computer model he had
developed with another engineer, Bibiana Luccioni. After

213 Log 50
 carrying out several experimental explosions in Argentina, the
two engineers created a digital program for simulating urban
blasts. Using the computer model, the engineers predicted
the path of the blast wave as it traveled across the street in
Beirut and compared this prediction with the results obtained
18.  See Public Transcript of the Hearing held through the Captieux reconstruction.18
on 24 February 2015 in the Case of Ayyash
et al., STL-11-01, Special Tribunal for
In his testimony, Ambrosini described the Kinney-
Lebanon, February 24, 2015. Graham equation, which is a mathematical formula used to
19.  Ibid., 25.
calculate the pressure caused by an explosion in a space with-
20.  Public Transcript of the Hearing held on
23 February 2015 in the Case of Ayyash et al., out obstacles. Because of the urban setting of the explosion in
STL-11-01, Special Tribunal for Lebanon,
February 23, 2015.
Beirut, and the various obstacles that the blast wave encoun-
tered in the course of its propagation, this equation was of lit-
tle value. To illustrate the equation’s shortcomings, Ambrosini
asked the court to imagine a hypothetical explosion unfolding
in the courtroom: “In this specific room,” he said, “an explo-
sion produced at this specific point would reflect from the
ground. It’s almost like a wave, like a water wave. It would
reflect on the ground, it would reflect on the wall behind me,
it would reflect on the other walls as well. So it is way more
complicated to calculate the pressure at various points.”19 As
much as an explosion transforms the space in which it occurs,
it is also shaped by it. While explosions destroy architecture,
there is also a morphology to each individual explosion that
the surrounding architecture helps produce. Buildings shape
the waves that raze them.
As the force of an explosion radiates outward and encoun-
ters obstacles, the wave is reflected back toward its point of
origin. Different materials cause different kinds of reflections,
thus Ambrosini’s model was careful to factor in the heteroge-
neous materials engulfed by the wave. Included in his report
were tables containing formulas and scientific data on “the
behavior of air,” “the behavior of TNT,” and that of “masonry,
concrete, steel, soil.”20 Using this data, Ambrosini’s model man-
aged to simulate the course of the explosion at intervals as short
as tenths of a millisecond. In this respect, one advantage offered
by the computational model is the ability to stretch out time.
The digital model slices the event into discrete units of time so
short that, when presented serially, they appear continuous –
much like the frames in a film. Changes in pressure at this tim-
escale were mapped throughout the digital model of the Rue
Minet al Hosn.
The other advantage of the computer model was its iter-
ability. Contrary to the reconstruction at Captieux, the com-
putational model could be exploded thousands of times with
slightly different parameters. When the dust settled in these

214 Log 50
 digital explosions, the parameters producing a crater most
similar to that observed after the explosion in Beirut and the
reconstruction at Captieux were consistent with a bomb ele-
vated a small distance above the ground.
In her analysis of the use of computer models by the
Special Tribunal for Lebanon, lawyer Lindsay Freeman
warned of the consequences set by this precedent: “The first
time this technology has ever been used before an international
21.  Freeman, 311. criminal tribunal.”21 After all, how does one cross-examine a
22.  Perma Research & Dev. v. Singer Co.,
542 F.2d III, 115 (2nd Cir. 1976) at 121 (Van
computer simulation? Due to the complexity of its mechanics,
Graafeiland, J., dissenting), quoted in the computer simulation possesses an aura of credibility that
Andrea Roth, “Machine Testimony,” Yale
Law Journal 126, no. 7 (2017): 2013.
is difficult for opposing lawyers to challenge. The problem of
23.  See Roth. the opacity of computer-generated evidence was already made
manifest in courtrooms by the 1970s. In 1976, a case involving
the patent of an anti-skid device, Perma Research & Dev. v.
Singer Co., led to an early computer simulation being entered as
evidence. Despite the simulation’s inclusion as evidence during
the trial, the algorithms and data underlying it were not dis-
closed, which the defendants claimed put them at a disadvan-
tage. Their motion to have the computational processes behind
the simulation shared for examination was denied by a major-
ity decision, though one judge did write in a dissenting opinion
that he was “not prepared to accept the product of a computer
as the equivalent of Holy Writ.”22
The dissenting judge’s reservations notwithstanding, a
glance at future trends indicates a significant likelihood that
evidence derived from computer models and machine-learn-
ing algorithms will continue to be used as evidence. Already,
select legal scholars have begun to decry the outsize influence
that computer simulations have on juries, who find their
computationally derived conclusions as difficult to discount
as opposing lawyers find them to discredit.
At this point, we may well await the digital equivalent of
Gibson’s law, the adage that for every PhD there is an equal
and opposite PhD. Gibson’s law describes the common situa-
tion in which both sides in a trial are able to find experts will-
ing to testify on their behalf. Already, there are cases where
slightly different computer programs will produce contrary
results.23 The Gibson’s law of the future could read: “For
every simulation there is an equal and opposite simulation.”
When court cases devolve into contests over the perceived
evidential superiority of computer models, judges and juries
will almost certainly need their own digital assistants, pro-
grammed to adversarially challenge simulations and expose
the biases concealed in their black boxes. Events unfolding in

215 Log 50
 microseconds will become subject to legal contestation in a
way that was previously impossible, as will phenomena cap-
tured at microcosmic or macrocosmic scales. Obscure cellular
mechanisms and vast storm systems, digitally rendered, will
both enter into this zone of juridical polemics.
Despite the difficulties posed by the presence of computer
models in the courtroom, it would be a mistake to meet this
situation with a retreat into epistemological relativism or a
Luddite ban on digital evidence. Computer models are not
merely depictions of events. They are a form of testimony –
a prosthetic reconstructed memory that can accommodate
those explosive forces liable to sunder the brittle matter of
human minds and bodies. Like human memories, they are
subject to errors. But particularly in moments when lethal
forces are unleashed, these “machine testimonies” stand
in for the testimonies of those who were too close to these
forces, who experienced the shock wave that the computer,
unharmed, can make within itself.

Modeling the Hole of the Real

In Jorge Luis Borges’s famed parable “On Exactitude in
Science,” an emperor demands the crafting of a map that is
the size of his territory. In the case of the Special Tribunal
for Lebanon, the models collectively demonstrate a varia-
tion on the ambition of Borges’s emperor: the goal was not a
map the size of a territory but a modeled explosion the size
of an exploded territory – a comprehensive reconstruction
of an instant of destruction. These models each give form to
the sundering of urban space by capturing its effects in table-
top models, by duplicating its destruction on a French firing
range, and by digitally iterating it across milliseconds. Craters
were compared; discrepancies were debated.
An explosion is perhaps the ultimate form of what
criminologists call transient evidence – that is, evidence that
immediately begins to deteriorate. Think of the heat of a cof-
fee cup or the fleeting odor of an incendiary device. Despite
the vast destruction it causes, and its lingering physical and
psychological effects, an explosion disappears in an instant,
exhausting itself in its own annihilation. An explosion is a
wave that remakes spaces as it expends itself. To model an
Jason Rhys Parry is a visiting assis- explosion is to model the behavior of air, concrete, soil, and
tant professor in the Honors College steel when exposed to overwhelming energy. It is to rebuild
and assistant director of the R1
a city and to destruct it, again and again, until the modeled
Learning Community at Purdue
University, and a fellow of the Future explosion and the modeled air, concrete, steel, and soil are all
Architecture Platform. in accordance with the gaping hole left in reality.

216
Cynthia Davidson


with Eyal Weizman
& Christina Varvia
Operative Models

At Forensic Architecture, a self-described “research agency” based

at Goldsmiths, University of London, a model is not the projec-
tion of an architectural idea but rather the culmination of inves-
tigations into human rights and environmental issues that also
have spatial and architectural dimensions. The multidisciplinary
nonprofit agency uses architectural analysis and digital model-
ing to piece together evidence – including cell phone videos posted
online – from disputed crime scenes to produce forensic models
that include techniques such as reenactment, photogrammetry, and
“real-scale” simulation. These are models that explain, or reveal,
behaviors more than project them, and they appear in courtrooms or
tribunals as often as they do art exhibitions. Forensic Architecture
director Eyal Weizman, professor of spatial and visual cultures
at Goldsmiths and founder of FA 10 years ago, and former deputy
director Christina Varvia, now an FA research fellow who is also
pursuing a PhD in biopolitics, joined me on November 11, 2020, via
Zoom for a transatlantic conversation about their work. An edited
transcript of our hour together follows here. – CD

CYNTHIA DAVIDSON: I want to talk about models and their

behaviors, but I’d like to start with your idea of operative mod-
els, or the term operative models, which for me suggests behav-
ior. How did you come to that term and what does it mean?

CHRISTINA VARVIA: While we were designing the ICA exhi-

bition in 2018, which is when this term came about, we were
thinking of how to structure the work we were presenting
along key concepts. We tried to figure out how to best describe
our work from different perspectives: from counter investiga-
tions, which was also the name of the exhibition, to the image-
data complex, which is the way we synthesize images in order
to analyze them, to the architecture of memory, operative models,
and forensic aesthetics. The idea of operative models was really
trying to address how we tend to use models differently than
a standard architectural practice and to open up a space to
discuss the role of models in practices of counter investigation
and in political practices.

217
Forensic Architecture, investigation of the murder of Halit Yozgat, 2017. Composite of physical and virtual reconstructions
of the internet cafe in which the murder occurred on April 6, 2006. All images courtesy Forensic Architecture.

EYAL WEIZMAN: We should say that the that sense, we were trying to reflect a little bit
term operative model is a reference to Harun on the work that we’ve been doing, to see how
Farocki’s term operative images. Harun’s work the models we are producing take on a life of
in film has been a huge influence on our prac- their own and somehow change the course
tice. Operative images are images that are of political narratives in the contexts we are
doing things in the world, that are operating investigating. We were referring specifically
beyond representation. So the term operative to models that have done this in our practice.
takes both images and models outside the Perhaps it would be good to start with the
realm of representation into a realm of dif- model of the West Bank to describe that.
ferent operations, whether as simulations or
as a way of testing various hypotheses, run- EW: One of the first cases litigated in Israel
ning scenarios, activating memory, teaching against the apartheid wall was brought by
machinic vision to see and to learn, and other a small Palestinian village. The wall had
things. So effectively, they function as opera- been built on its land, separating the village
tive devices. This is true for spatial models and from its fields. The judges were faced with
for models understood as theoretical formula- all sorts of claims about slopes and distances
tions. Both need anchoring – a process known and heights, and decided to respond like an
as “ground truth” that calibrates elements architecture teacher would to a student who
in a model with their equivalent elements in speaks too much, go home and come back with
physical space. models. So the trial became something like an
architectural crit, with the military building
CV: Farocki actually uses operational images its model and the human rights group that
rather than operative, but there are very small represented the Palestinian village build-
differences between the two terms. He intro- ing another model. And, in fact, when the
duced the term to refer to images that are models were presented in court, the judges,
part of a process, that operate within the kind who were sitting too high on the bench to
of machinic process that often doesn’t even see them, stepped down – for the first time
reach human eyes, that is, within a military ever – to stand around the models. There, the
complex. But to talk about them as operative physical materiality of the models started to
rather than operational is also to talk about interrupt the legal protocols of the process
how the images or the models themselves itself. There is a certain formality to the legal
instigate and change an action rather than just process that was broken by people making
become part of a process – rather than just claims and shifting things on the models. The
play within something preexisting, let’s say. In model thus became the operative language

218 Log 50
through which the law and the legal case view two seconds later, therefore we under-
were themselves articulated. stand how those two pieces of footage relate
When Forensic Architecture started using to one another. So there’s an actor that almost
architecture to interrogate all sorts of inci- stitches the videos together in space and in
dents that were captured by videos circulating time. Even if there’s no overlap between the
on social media, we realized that the archi- cameras, that person’s pace becomes what we
tectural model can operate like a way of see- use to lock the viewpoints into place.
ing, an optical device. There are dozens, and
sometimes, as in some of our more recent CD: So switching produces the spatial quality?
cases like the protests in the US or in Hong
Kong, thousands of videos capturing more or CV: Exactly. Images and videos are flattened
less the same incident and posted online. The mediatized reality. We try to unpack that in
only way to understand what is happening is its three and four dimensions to figure out
to arrange those videos within an architectural how those relations play out beyond the image
model and let the model help you navigate frame or beyond the duration of the video
from one video to the next. The model allows clip. It’s more about figuring out a chain in
you to keep the videos in their full length. We which one piece connects to another through
don’t edit our source material. Rather, using a temporal or spatial link. Slowly, this makes
the model, we allow for a navigational view- a model that is always partial – we can never
ing, a spatial edit if you like, in which you can actually see the whole image, it’s always
move from one video to the next, bypassing patchy. But at the same time, it’s almost an
the “cut” so crucial in film history and theory. honest way of modeling, meaning that we
leave gaps where we don’t have any infor-
CD: How do you move from what would be a mation. We keep those bits abstract within a
linear edit to a spatial edit? Perhaps this ques- city landscape, for example. We only add the
tion relates to the one you pose in your semi- detail and model out from the videos the parts
nar called Operative Models: “What are the that we do know from that particular footage.
multiple relations between models and their That model then becomes the analytical tool
depicted realities?” that we can use to, again, stitch together the
narrative of what has taken place, the events
CV: I guess the models that we’re referring to depicted in the videos.
begin in the process of trying to make sense
of whatever has happened in the case that EW: Our models strive for what Walter
we’re investigating – for our own sake first, as Benjamin calls Bildraum, or image-space,
opposed to presentation purposes. To do that, a term that brings together photography
we need to figure out how all of the pieces of and architecture. Benjamin contrasts the
evidence that come to us relate to one another. Bildraum to the plan and sectional views of
So beyond figuring out moments that are of a building. He thinks the first is experien-
interest and slicing them together in the clas- tial while the second is analytic. It is a useful
sical editing process, we’re trying to figure term because our relation to images today
out how pieces of footage link to one another, must be spatial through all types of photog-
even through parts of the video that might be raphy – from three-dimensional lidar scans
irrelevant to the story. For example, some- to photogrammetry. Photographic space is
one passes by in the background and then one that you move through and are immersed
we see that same person in another camera in. But unlike what Benjamin suggests, it is

219 Log 50
not only experiential but can also be analytic. digitally. Is there also a need to have a physi-
So the operative model emerges through the cal model of a physical place, a place that you
multiplicity of very situated, partial, located can actually go out and measure?
perspectives; the more participants there
are with their phones on, the more you can CV: Yes. This links nicely to another function
actually reconstruct what is going on. But of the models, which is to simulate certain sce-
the model as a Bildraum is not like a satellite narios and to ask certain very specific research
image, a view from nowhere, or a God’s-eye questions that have to do, for example, with
view; it’s a very situated perspective. The idea time or with speed. One of the cases that we
that we call open verification is one by which reconstructed as a 1:1 model was the investiga-
the model becomes the medium of rela- tion we did on the killing of Halit Yozgat by
tion, not only between all the videos that are the NSU [National Socialist Underground], a
synced up within it but also for other kinds German neo-Nazi group. Here, the question
of testimonials and material pieces of evi- was very specific: a secret service agent was
dence. The operative model thus becomes a present in the internet cafe where the murder
social diagram, the medium within which took place, but he claimed not to have heard
the different participants in an investigation the gunshots, seen the body of the person who
come together to locate and test things. Unlike died, or smelled the gunpowder.
police investigations, which are conducted In this case the question was about the
behind closed doors, our investigations are perception of this agent, as well as questions
carried out in the open. We’re bringing in that arose in terms of timing, whether it was
groups that experienced violence firsthand – possible that this agent was in and out of the
sometimes the activists, the lawyers that par- shop before the murder happened and there-
ticipate with them, sometimes volunteers, or fore could have missed it. We needed to ask
crowdsourced information – in an open pro- those questions spatially and temporally. We
cess that is meshed together within the model. couldn’t go back to the original internet cafe
The model can also highlight for us not because it had been remodeled; it didn’t look
only what links in but also what doesn’t link. the same and its material properties didn’t
Increasingly, we see fake videos. One of the act the same way. So we had to rebuild it as a
ways to determine whether a video is fake or 1:1 abstracted model. It wasn’t about re-cre-
not is to look at it vertically, to look at how it ating it as a theater set. The texture and the
is composed, how it is JPEG-ed, how the pix- look of it weren’t as important as its acoustic
els are organized. That is, you look at the file and visual properties, its accuracy in scale in
itself, its inherent properties. But horizontal terms of how far one would have to walk, for
verification is testing whether it agrees with example – seven meters from one end to the
other bits of evidence. A fake video would other – and how long it would take for the
actually not weave together with genuine ones. killer to come in, shoot twice, and leave again.
So, effectively, I think that the idea of an oper- That model became a way for us to use
ative model is bringing together and collapsing actors to simulate or reenact those testimo-
photography, videography, and architecture. nies, specifically the testimony of the secret
service agent, and to question whether it
CD: When you’re bringing all this material would be plausible. So it became about almost
together, at what point is there a physical adopting the perspective of the secret service
model versus a digital model? I would assume agent, accepting his testimony, and then test-
that in working with video, you’re working ing what conditions this could be possible

220 Log 50
in. And this is the way we discounted it, of experts and combine all of their expertise to
course. But in the process of making the phys- synthesize some results. It seemed natural to
ical model, we also had to work with acoustic be making models in this particular case. And
experts and fluid dynamics experts to be able then, of course, there are moments where
to simulate the event. How could we re-create we’ve done reenactments on the site itself.
a gunshot in that space? We had to ask weap-
ons experts to shoot the same gun and record CD: We could talk about this for hours, but
the sound, and then re-create that sound in I want to go back to another question you
the model, measuring it from a specific loca- posed in the Operative Models seminar, and
tion, and then simulate all of that digitally that is, “How are models activated to gain
so that we could corroborate the two models agency in different forums?”
and work out the relationship between the
physical experiment and the digital. To cali- EW: At Forensic Architecture we do not only
brate the two, we obviously had to take some work in the courts. I would say about 20 per-
ground measurements, so we also made a dig- cent of our cases are performed in interna-
ital model of the physical abstracted model to tional and national courts, always taking a
see how close it would be to the original. That state agency to account, always concentrating
latter ground truth model was used for mak- on alleged crime by the police or other state
ing sure the measurements are translatable. institutions. Increasingly, our work is being
shown in other forums, in citizen tribunals,
CD: When I hear you describe this, I have to in the media. I think you’ve seen that the
ask, what, as architects, do you bring to such technologies we have been developing over
a forensic investigation that a police detective the past 10 years are now becoming rela-
does not? tively mainstream, with units in the New York
Times, the Guardian, and the Washington Post
CV: First of all, the will to do all of that, using some of what we have developed. And
because the police investigators did not care in the art world, we have presented in various
enough to figure out those same questions exhibitions that make our models operative
when they had access to the crime scene to in another way.
begin with. To answer the question, we would need
to look at how our models operate in relation
CD: But I would think you have different cli- to the different kinds of forums that we are
ents, shall we say. A policeman doesn’t really discussing. When we present cases in the
have a client, but has authority – I would media, a model synthesizes videos found in
guess you each have a different authority. open-source investigations. In the courts, the
models operate differently. Christina could
CV: A different position, because we are inves- tell you about the Golden Dawn case, where
tigating the police that didn’t do their job or the Nazi party in Greece was banned after a
the secret service agency that should have been case that we presented evidence for. It is effec-
investigating right-wing murders, right? So, tively just like the model in the West Bank.
positionality is one factor. But at the same time, The introduction of architectural language
as architects we know how to build models. into a legal context is not something that the
Whether they’re scale models or 1:1, we know legal framework can simply digest. It’s a kind
how to think about the acoustics and materi- of Trojan horse that makes people see things
ality of space and how to work with multiple in a different way. So an architectural way of

221 Log 50
storytelling and figuring things out has been where the physical model is, where the wit-
quite influential in these domains. ness stands, and where the experts stand, and
Within the art context, production fees that in itself becomes a certain form of spatial
for our exhibition at the Tate in London, for rhetoric. So these are just some of the ways
example, were used to produce an investi- in which we use models to make our point, to
gation, in this instance into an Israeli police find things out, and to intervene in processes.
killing of a Palestinian Bedouin man in the
Negev – a case in which we exposed a police CD: Christina, did you want to add to that?
murder that the government has now apolo-
gized for, retreating from their initial narra- CV: Maybe only a little anecdote about the
tive that that person was a terrorist. In that case I mentioned before, the neo-Nazi kill-
particular case, we also worked with the vil- ing. In that case, the family wanted the judge
lage residents and activists to reenact the inci- presiding over the trial of one of the per-
dent on the ground on which it took place. petrators to go down to the internet cafe to
Reenactment turns real space into a see for themselves how the account of the
model. You know the police trope of reenact- secret service agent was completely absurd.
ment. As the last stage of somebody’s confes- They were really trying to get the police offi-
sion, they’re taken to the scene of the crime cers and the judge to see the space for them-
in a highly ritualistic manner that becomes selves. Demonstrating the impossibility of that
part of the confession. With us, reenactment account spatially is hard to do without a cer-
has become a mode of testing things on-site. tain type of model. Since the shop was com-
The Negev case had to do with the relation- pletely gone, we had to re-create it. When we
ship of a slope, several buildings, and a road. then brought in MPs and journalists and they
By testing and reenacting that incident again saw that place, when they were in that space
and again with actors, we were able to show themselves, even though it was re-created and
that the police narrative was wrong and that remodeled, it made a huge difference to the
the person killed could not have run his car understanding of what this incident was. The
at the police as they had claimed. The village police narratives that deny certain accounts
became a model of itself, if you like, for the are based on the fact that it’s really difficult to
duration of that reenactment. have a situated and spatial understanding.
In other cases, we have planned to create Providing a model often allows people to see
in court a certain spatial layout to demon- for themselves. I think that is the function
strate things to the judges – how something we’re trying to promote in the models we pro-
would have been seen and could not have duce, to allow people to actually see those rela-
been seen. Whenever we are invited into a tionships so that they can judge for themselves.
court or a citizen tribunal, we also want to
design the court because this is a site in which CD: Yes, the spatial is very different from, as
we operate. How a person stands in front of a Eyal said earlier, the flatness of the image.
piece of evidence, whether material or media, People read something completely different
becomes very important. This is a process of from it. I want to move into something that
making things speak, because things do not you, Eyal, wrote about last year, about the
speak for themselves, they speak through idea of post-truth, and ask you both how you
the expert witness that animates them. And can operate in this post-truth climate. You
that “prosopopoeia” is architectural, because said that post-truth is a new form of censor-
you need to design where the screen stands, ship that “blocks one’s ability to evaluate and

222 Log 50
Forensic Architecture, investigation of a killing in Umm al-Hiran, 2018. Projecting thermal footage from a police helicopter
establishes the spatial relationship of figures and vehicles, reflected in a photogrammetry 3D site model.

debate facts,” and of course this creates doubt. kinds of truth statements, to build something
Through your forensic work with operative new out of the ruins. Finding a much more
models, are you trying to change minds or can robust and creative way of truth telling may
counter forensics only cast doubt in the same work by bringing together multiple perspec-
way that post-truth casts doubt? tives and multiple points of view, different
skills, and working in an open rather than a
EW: This is an excellent question. When we closed manner. The perspective of people on
see the way in which mainstream institu- the ground is the most important, the situ-
tions react to the challenge of post-truth ated experience of communities and people
we are left a little bit uneasy, because we see that are the victims of state violence and
people trying to reestablish the pillars of repression. When combined and meshed with
power knowledge, to salvage our institutions technological know-how and expertise, this
from these nihilistic, epistemological wreck- perspective can be stronger. But sometimes
ing balls that some contemporary regimes amateur investigators who are scanning the
employ. So, against this “dark epistemology,” internet, finding clues, and assembling them
we think it’s wrong to reestablish the author- together are also crucial. So artists, scien-
ity of the police, the FBI, and the judiciary. tists, architects, and filmmakers working to
After all, we similarly often contest the bridge the huge disciplinary gaps that oth-
impartiality of the judicial process, the police, erwise exist between those frames is essen-
and organizations like the FBI. We’ve con- tial and what we need to be doing, I think,
fronted secret services, police, and militaries today. The veracity of our statements should
across the world for incorrect statements, for be in its making, rather than in the assumed
weaponizing doubt, and for weaponizing prestige of our institutions. Rather than say-
the law against the weak. To that extent, ing, Princeton institute of this and that declares
we do not think that it is a trench war in that, therefore it’s true, or, this government
which we need to reestablish the old order, think tank thinks that and therefore it’s true, we
but rather an opportunity to think creatively believe that we have to show, publicly, in a
again about epistemology and to think about very accessible manner, every step of the way
the way in which we can construct different a case is assembled.

223 Log 50
 EW: Each one of our investigations relies on
a very broad and very diverse network of
collaborations. When we were invited to the
Whitney Biennial, we decided that the only
thing that we could do was to investigate
the Whitney at the Whitney, to investigate
Warren B. Kanders, a board member, and his
dealings with tear gas. In order to do that, we
collaborated with activists in Palestine and
in Tijuana, and in multiple places around the
world where communities suffering from
tear gas were collecting information both
online and physically. The on-the-ground
aspect was paramount but it needed to be
complemented with architectural and then
Forensic Architecture, Model Zoo, 2020. Machine learning computer skills. Together, we created 3D
classifiers use rendered images of 3D models or “synthetic
models of those munitions in order to render
data” and improve when they are trained on extreme varia-
tions of the modeled object. These extreme variations of tens of thousands of photo-realistic images
texture on the modeled projectiles help the classifier better of them, which we then used as training sets
recognize their shape, contours, and edges.
for teaching computers how to see. To learn,
And truth is fabricated in both meanings the classifier sometimes needs 10,000 or so
of the word. For some people, it’s made up; iterations of a single object. It’s almost like
for us, it is actually built very patiently by teaching a child to see. You say, this is a head-
combining different perspectives, perspec- phone. [Eyal holds up black headphones.] This
tives that would otherwise be kept outside of is also a headphone. [He changes the view of
the authority of truth statements. A friend of the headphones.] If I put it on my head, it’s still
ours, Achille Mbembe, who is a Cameroonian a headphone. And even if it’s red, it’s a head-
philosopher, contests the term university in phone. And even if you see half of it, it’s still a
terms of its universal conception of monop- headphone. So, at some point, the machine
erspectival knowledge, which comes with a learning classifiers learn how to identify an
history of white supremacy and colonialism, object independently. But in order to rap-
and calls instead for a pluriversity. Pluriversity idly provide all those iterations of objects, we
is the idea of a multiplicity of perspectives, were building 3D models of the objects that
embodied and situated experiences that are a community found on the ground or that
clashing against each other and producing we found photographs of online. Because
different forms of knowledge that can actu- we are architects, we know how to build
ally stand and withstand cross investigation. the right environments in which the canis-
We do not contest the fact that there is truth ters are located, and we rapidly created tens
out there, that there is something left to it; of thousands of iterations of those images
we contest the way it is arrived at, and we are for the computer so it could run and find
bringing into the mix many different voices tear gas made by Warren B. Kanders or by
that are otherwise excluded from it. other manufacturers online. So this is what
we showed at the Whitney, and we were
CD: Can you give me an example of one of actually able to contribute, together with
your collaborations? other activists, to his resignation and to his

224 Log 50
Forensic Architecture, investigation of the killing of Mark Duggan, 2020. Researchers divided the time before and after the
shooting into nine distinct moments, modeling each from the perspective of more than a dozen different witnesses.

disinvestment from tear gas when he sold his condition that is a commons produced from
business in tear gas manufacturing. the meshing of our multiple perspectives.
That conception of verification is some-
CD: That’s a fascinating use of models. thing we try to amplify through our mod-
els. We can get closer to a determination the
EW: So sometimes you can show how the more the multiple perspectives agree with
space of a gallery is not just a critical space each step, but it is better if those perspectives
for making critical comments on culture but are ontologically different. For example, a
one that can bite back. For too long, we were military video, a testimony by somebody on
thinking, both as architects and artists, about the ground, a satellite image, a bit of blood
critical interventions in politics and cul- splatter, a confession, perhaps of a soldier or a
ture. But we want to gain operative agency policeman, these all need to lock in. We try to
through our work and actually show that lock in hundreds and hundreds of data points
there are consequences. in order to see if we can create the fabric that
can establish this horizontal verification.
CD: You have talked about the need for verifi- That is to say, verification is a spatial
cation and used the term verification, which I practice. It is what the architect does by bring-
find quite interesting since to “verify in field” ing together different types of material, not
is something architecture has to do during only different elements, but different types of
construction. What does verification, or the elements. And the more diverse they are, the
process of verification, mean in your work? more open verification works. The more the
perspectives of the psychologist, the artist, the
EW: We consider truth and verification as filmmaker, the scientist, and the curator can
a commons. We tend to think of the com- be negotiated into a malleable something, the
mons as water, as air – language is a com- more solid the commons that is arrived at.
mons, something that is not owned but used Another very important principle that we
and thus evolves in that way. Our percep- work on is to start from a controversy, to start
tion of the world is as a kind of metapolitical from a split second, to start from an incident,

225 Log 50
and then weave the fabric around it to tie it operate, and we are always skeptical about
back to the world in which it is a part. So the how much the government can do, whatever
cases that we choose are always cases where government that is, but this becomes quite
we think there are historical hinges. Take the crucial in cases that are ongoing.
Golden Dawn killing in Greece. You could For example, we’re looking now at push-
take that moment in Athens and weave a fab- back cases of migrants that come into Europe,
ric to implicate the entire Nazi party. Or we from Turkey to Greece. We’ve investigated
can start with the killing of Mark Duggan in seven cases of pushbacks or, let’s say, five
London, the site of protests and riots in 2011, cases of pushbacks, two cases of shootings
and actually weave it back to look at race rela- across the border. And we could keep going
tions and police brutality across the city. The because this is an ongoing practice, we could
weaving is kind of endless, because each thread keep going forever. But what’s important
can be pulled as far as you want to pull it. for us is to give enough evidence to convince
those governments – the Greek government,
CD: Weaving is a wonderful concept. As you the European governments that are respon-
say, you can just keep going. So, Christina, how sible for those actions and for the policies that
do you know when to stop weaving? How do deny those actions – to open up an indepen-
you know when an operative model that comes dent investigation and allow independent
out of your forensic investigation is finished? monitoring groups to check them so that they
stop that practice. The political point here is,
CV: That’s a good question. I guess it varies how do we create enough action to change
significantly across cases. Sometimes there’s a that modus operandi? That’s the key.
very limited amount of evidence and there- It’s very different across investigations,
fore you’re finished with the analysis when of course. Certain openings become apparent
everything is locked into place and there’s when working either with the lawyers for the
nothing else to input. That’s one case. But, victims’ families or activist groups. When we
of course, there are cases with thousands of work in exhibitions, there’s a moment at the
images and videos, thousands of narratives exhibition opening that becomes both the end
that could plug into the model, and then it’s of a cycle of work and the beginning of a dif-
about trying to distill the most important ferent kind of cycle, because when we present
political questions and research questions that our findings, that is the moment that a differ-
will unlock certain actions further on. ent audience, whether it’s a public that gets
So sometimes it’s not only about solving a vote or the investigators who are part of
the case, about figuring out exactly what the official investigations, comes and is con-
happened and when it happened. This might fronted with that evidence. That then spins
not be possible, from our perspective, because off to different types of work that have to do
we don’t always have as many materials or with advocacy, sometimes undertaken by our
resources, access to the scene, or access to the partners and not us. So I guess there’s never
witnesses as, for example, government bodies a clear-cut end to a project, it’s never… the
or institutions might have. Sometimes our model never finishes. But inasmuch as it has a
work is mainly about opening up a space political life, then that’s the moment to let it
so that we can invite those who have the out into the world.
resources and those who can then take that
political action further and determine policy
to participate. I guess it’s a strange way to Cynthia Davidson is the editor of Log.

226
Sean Anderson
Compound Tenses

April 19, 1993: McLennan County/Waco, Texas

Federal Bureau of Investigation
Case #10–1713/H. Rept. 106–1037
1-SAC/adr
2-89B-SA-38851 / (4)

Body of a White Male, Age 30–35. Found at location A045–

5.675, Mt. Carmel Center, Old Farm Road. Estimated weight
157 pounds; height 5’ 11”. Found on stomach. Pertinent find-
ings in the Medical Examiner’s report include: major damage
due to burns at 90 percent coverage. Blood samples taken at
the lab show a small dose of amphetamines and alcohol along
with traces of prolonged exposure to tear gas (chloroben-
zylidenemalononitrile – CS). Hair samples with blood spatter
indicate severe trauma on right occipital. Bullet wound found,
probable cause of death. Self-inflicted unclear. Bullet wound
may also suggest possible violent encounter before death.

The Branch Davidian sect originated as a splinter group of the

Seventh-day Adventist Church in California in 1934. Founded
by Victor Houteff, the group believed that Christ’s return to
Earth could not happen until a purer church was established.
After his death, Lois Roden, a self-proclaimed “prophetess,”
took over and gained many recruits from all over the country.
She believed that Christ and the Holy Spirit were feminine.
She was later “disfellowshipped” by the Seventh-day Adventist
Church for her extreme views. Upon Roden’s death, Vernon
Wayne Howell, a member, argued with the prophetess’s son,
who claimed he could raise the dead, and subsequently dueled
with George Roden in a gun match outside his home and won.
George Roden was committed to a state mental institution
soon after.
Would-be rock musician Vernon Howell then traveled
across the world in search of new recruits and members for
the group of Branch Davidians. After he visited Australia and
parts of Asia, it is estimated that there were over 200 mem-
bers. After Howell changed his name to David Koresh, and
later, Yahweh Koresh, he predicted that he was the seventh

227
and final angel who would bring about the end of the world in a fiery Armageddon, prefigur-
ing a raid by the FBI and ATF at the sect’s ranch on Old Farm Road, on the outskirts of Waco,
Texas, on February 28, 1993, after a 51-day standoff.

FBI Interview with Person A (Potential Witness)

Magazine: African Americans are taking a new look at cults and their danger. There were an
estimated 40 Black individuals out of 120 members in the cult at the time of the raid. Where
did they all come from?
Person A: From all over… England, the Caribbean, and the US. One went to Harvard.
Magazine: How many Black members were left when you were apprehended by the police for
accessory to harm a federal agent?
Person A: I don’t know. Who cares? They dead, ain’t they?!

He could take you to depths you wouldn’t know. If you studied a certain book or chapter or a
whole book – a whole book, 66 chapters! After you got that understanding, he could take you
to another level, and another level.

FBI Background Interview with Advisor A, Peoria, IL

WC: Could you describe your feelings about dealings with cults, past and present, as well as the
Branch Davidians in particular?
A: Well, because they view all outsiders as in league with Satan, no compromise is ever possible.
Resistance to the outside world is therefore unyielding and potentially belligerent. Zbigniew
Brzezinski [former President Jimmy Carter’s foreign policy advisor] wrote that “the prevail-
ing orthodoxy among intellectuals in the West is that religion is a waning, irrational, and dys-
functional aberration.”1

Field Report: Waco Chronicle Arts and Entertainment Section

Although the Branch Davidian group has been in the area for a long time, at least since the
1960s, a couple of people have asked me to comment on the state of the ranch and what has
been going on out there on Old Farm Road. Reached by a gated long driveway, the pink-and-
white wood structures have been painted recently, replacing a few older structures that were
torn down. A new swimming pool and gymnasium have been added. The primary space of the
entire complex appears to be residential or, perhaps, the church itself.
Centrally located, the chapel is lit by southern light from large windows. The interior is
said to be sparse with only a few pews and some bleachers. Hay bales are often brought into
the space for longer sermons, allowing people to sit for extended periods. The chapel is next
to the water tower and the “watchtower.” The only thing taller in the area is the flagpole that
was installed out front. Interviews with a few people outside revealed that the bedrooms for
women and children are located on the second floor, while the men’s rooms are on the ground

1.  Zbigniew Brzezinski, Out of Control: Global Turmoil on the Eve of the Twenty-First Century (New York: Scribner, 1993), 218.

228 Log 50
Matthew Wittmer, model of the
Mount Carmel compound, circa 2009.
Box 1250, The Wittliff Collections,
Texas State University. Photo cour-
tesy the artist.

floor. Mr. Koresh sleeps in the “watchtower,” which includes an extensive collection of stereo
equipment. There is a large shared kitchen and dining area. A very deep ditch, almost the size
of the building itself, can be found adjacent to the main building.

FBI Interview with Agent B (Name Withheld; GG)

WC: Could you tell us your impression of February 28, 1993, the day your forces attempted to
enter the compound?
GG: Within five minutes after our men moved in – I was assigned to the southern quadrant –
we’d fired our load and were pinned down. We lost radio contact because of the intense gunfire
and because we were on the opposite side of the command post. All I kept thinking was that we
had control while serving the warrants. But after that, it was like lambs to the slaughter.

Overheard Conversation [Olde Towne Pub, Waco; unknown subjects]

Person 1: Why a gun business? If you wanted to make money, why not art or a bakery?
Ranching? Why weapons?
Person 2: Well, the right to bear arms is constitutional. Remember the Second Amendment?

ATF / Recorded Telephone Conversation [February 27, 1993; 0907–0911]

ATF: Mr. Koresh, we just want to speak with you about the weapons you have stored at the
ranch. We know you got ’em. All we ask of you is to come outside and talk to us.
DK: If you come within a hundred yards of my ranch there will be hell to pay.
ATF: Mr. Koresh, we know that you have illegal weapons inside the compound. We also have
received reports of the abuse of women and children.
DK: It’s all lies. Whatever anyone told you, it’s all lies. [phone hung up]

Taped Description of Initial Raid (February 28) by an ex–Branch Davidian [Name Withheld]
The voice of Mr. Nice Guy had disappeared and been replaced by that of a fanatic. David ran
about the ranch telling us that the day was come and to be prepared. He said the Mighty Men
(our private militia at the time) were going to protect everyone. He told the children that if

229 Log 50
FBI Mount Carmel compound model, Waco, “The Strangers Across the Street,” episode 2, aired January 31, 2018.

they heard any loud noises, they were to hide under their beds until someone came and found
them. Each room in the ranch was then equipped with a gun and two gas masks hung on a
shelf. The underground rooms were not ready yet.
Our day-to-day activities continued. We didn’t see much of David – some say he was talk-
ing to the Lord. Others said he was recording an album for the rest of the world to hear. In
the early morning on 28 February, a few people broke in but were overpowered by the Mighty
Men. David’s talk of future calamities was right: six of ours perished before the intruders left.
We had to bury them out back, by the new buildings. They had also killed some of our dogs that
the kids always played with. When we were taking care of the house and preparing dinner, and
our men were in prayer, the dogs were the only playmates for the kids.
From then on, things were different. The only times we were allowed outside were to
empty the big plastic jugs we used as the toilets. You see, the nice world we were used to
as described in the Bible was altered by the outsiders. It seemed as if we were always being
watched. One of the Mighty Men told me that from the tower, he could see army tanks
and soldiers setting up camps in front of the ranch. Just down the road. David kept making
announcements on the loudspeakers, telling us the end was near, to read our Bibles more. The
kids were scared. And we were upset. We knew the time had come.

FBI Interview: Advisor C [JJ]

WC: What is your impression of David Koresh?
JJ: He was a man that gorged himself on sex, beer, rock ’n’ roll, guns, and religion. The only
holiness he saw was in himself. He justified sealing himself and his followers off in a poorly
built compound east of Waco by personifying himself as good and the rest of the world as bad.

230 Log 50
Press Conference with DD [attorney], after the initial raid (March 2)
“They shot them… they were the kids’ pets. They had some mongrels out there, too, that got hurt in
the razor wire. David will come out of the ranch only when the FBI plays a 58-minute recorded reli-
gious message that he has taped. He has told me personally that the shoot-out was ‘unnecessary.’”

So far 34 of our flock have left the compound…

\ ATF Field Report \ (Unidentified Agent) \ March 22 \

We got some video last week showing some of the children inside the compound. Koresh must
have possession of some nice video recording equipment because his message was clear enough
for the children to understand and for one to say in front of a camera, “Are you coming to kill
me?” I about cried.

Recovered Journal Entry (Name Withheld): March 25

…after that first raid, we had to rebuild some walls and block windows because the agents had
broke through most of them. David instructed us to cover windows with anything we could
find: wood, paper, sheets. They start flashin bright lights at us in order for us to go crazy I
think. Wont work cause we all believe in the will of David. Occasionaly we can here this bad
music being played. All night. Some us were wondrin whether that was agents bein bored or
the devil playin tricks on us. It’s all dark in here now, cept for some lanterns and a little light
shining through the chapel windows. We dont know whats gonna happen, so we just keep
reading Scripture and talkin to each other. I wonder if the ranch can handle all those army
tanks we see now.

Dallas Ledger, Arts and Entertainment {Special Report on Building Arts}

Down on Old Farm Road in Waco, the Cable News Network (CNN) channel is broadcasting
live 24 hours a day for the first time. The compound has now been termed Ranch Apocalypse
by some in the media and supposedly by Mr. Koresh himself. An onslaught of journalists and
newspersons have arrived, producing a worldwide audience through the television. They have
assembled around an area a few miles away where the ranch and its outbuildings are clearly
visible. The flagpole and front door are clearly visible.
I am reminded of the floor plans used by the author S.S. Van Dine in his dry mystery nov-
els – or that old widow Mrs. Winchester in Pasadena – as we always knew, as readers, where
the characters were in a space, just as we were sure of our locations within the novel. But
what about the narrator? Was the author the true explicator of the mystery? Images we keep
seeing in the papers and on the television have dislocated these events from the viewer/reader,
making us question the validity of what we are seeing and hearing. There are so many recon-
structed views of the property. Is there a difference between the broadcast image of the com-
pound and that of the photograph – or even what is being experienced within?
There are reports that a second “compound” is being constructed for a made-for-televi-
sion movie down the road. Some have said it is very “lifelike,” complete with the watchtower

231 Log 50
as well as some of the other features, to provide a “realistic” approach to the unfolding sto-
ries. They even brought in the school bus like we see at the original ranch. It’s like a full-scale
model for the actors to occupy and the filmmakers to use. But how do we reconcile imagining
life in these two places? Both are being used by and for television. For audiences. For viewers.
At the same time, the Branch Davidian compound, as well as the standoff itself, has been
relegated to the interpretations of a politically minded group of officials seeking an end to the
standoff. The outward manifestations of the original and its reproduction (e.g., described as a
“fortress” that the members of the sect are not allowed to leave) are being played out on tele-
vision and in the newspapers. People keep making inferences about contrasting meanings for a set
of buildings that were once considered just a “ranch.” This nonalignment has produced a rever-
sal of meanings: the ranch was originally intended to be a center that would be domestic in scale
for the sole purpose of housing followers who believed in Mr. Koresh’s words and his direct con-
nection to the divine. Now it has become an artillery compound, able to withstand an impending
attack and stocked with enough supplies to last an “extended period of time” for those inside to
survive the Apocalypse predicted by Mr. Koresh. When did the end become a beginning?

The Dallas Daily Message Interview: Special Agent (SS)

DN: Could you tell us why you are using the loud music and the lights shining all night on the
compound?
SS: The audio assault is a progression in terms of negotiation designed to keep [the cultists]
from being able to sleep and to break down some of Koresh’s control over them.
DN: What is the approximate chronology up until now?

Day 1: Initial raid by ATF agents alongside the FBI results in four dead on our side and
six dead on theirs. Koresh is supposedly injured.
Day 2: The National Guard and other military branches called in. Koresh releases 21
children and calls the raid “unnecessary.”
Day 4: Some Davidians leave the compound and we immediately apprehend them.
Koresh pledges over the telephone that he will surrender after he hears from God.
Day 10: The Davidians hang a banner outside an upper-story window that reads, “FBI
BROKE NEGOTIATIONS WE WANT THE PRESS.” That thing could be seen from our com-
mand post two miles away.
Day 23: We begin using psychological warfare. The recorded noise you hear is a combi-
nation of Tibetan chants, unidentified rock music – some say Barry Manilow – and the
song “These Boots Are Made for Walkin’,” which would not be to the liking of Koresh,
and the recorded screams of animals. We also project intense beams of light into the
compound’s windows.

…I wore earplugs at night…

232 Log 50
Banner hung from an upper-story
window of the Branch Davidian com-
pound, 1993. Dick Reavis Papers, Box
41, Folder 11, The Wittliff Collections,
Texas State University. Photo courtesy
The Wittliff Collections.

Local Television Interview with a Neighbor (Name Withheld)

SP: Did you have a sense of what a media circus this has become?
RR: Oh yeah, we could see the “satellite city,” as they call it outside the gates. The roads are
blocked by all these cars and vans from places we have never heard of.

Instructions on the Limited Release of Tear Gas Within a Confined Perimeter

Before injecting tear gas into a confined area, the Bureau recommends securing any member
of the Department or other person that may be affected with a protective mask with a respira-
tor. This mask must be worn at all times to provide a safe breathing environment.
It is helpful to have a device (TGUX-67732) that provides some distance from the actual
site and its distribution center. Insert the device through a window or another opening before
releasing gas. There will be a loud popping noise because of the release of the pressurized con-
tainer. More gas canisters may be needed unless occupants are subdued. It is safe to enter the
area within 15 to 25 minutes, depending on the amount of gas released.

\ FBI Interview with ATF Agent C (Name Withheld; LL) \ April 09 \

SP: How long can they last, given the new conditions?
LL: If they got enough food, maybe six months. But no matter. All our positions are chip
shots, an easy head shot. Everyone said that the most provocative thing to do was to torch that
fence. We stood a chance that there would be a violent response. So we thought long and hard.
But we removed it, and there was no action.

We all kept ourselves very clean and we had plenty of food to eat.

Excerpt from Koresh Notes (Written by Kathy Andrade): April 11

My hand made heaven and earth. My hand also will bring it to the end. Do not find yourselves
to be fighting against Me. Why will you be Lost?

233 Log 50
\ FBI Interview with ATF Agent D (Name Withheld; MM) \ April 29 \ 1622–1624 \
SP: Any movement?
MM: The only rise we ever got out of David Koresh was last Sunday when one of our vehicles
towed his black Chevrolet Camaro to make room for the next day’s planned raid.

\ Minutes \ 041793 (April 17, 1993) \ [Confidential; Secrecy Level: Secret] \

XX: A couple of men down there have determined the safest places to target are the extreme
corners and central locations around the northern face of the central structure. If our sources
are correct, there should be no lives lost due to building failure. Everyone should come out of
that place, even the children.
XX: What is the risk?
XX: Look [name redacted], I’ve heard that members of the cult were dumping their daily waste
from white buckets into the pool as well as the bodies of those that could not survive the severe
nutritional disaster.
XX: Any updates on the remaining children?
XX: Koresh has been using “the helper” lately.

\ Recorded Telephone Conversation [Confidential; Secrecy Level: Top Secret] \

XX: Have you carefully considered it all? From every angle? Is it the best way to go?
XX: Yes, it is my responsibility and I think it is the best way to go.

April 19 0555
Loudspeaker announcement: There is going to be tear gas injected into the compound. This is
not an assault. Do not fire. The idea is to get you all out.

David said, “Everyone grab your masks!” We had been up all night reading the Good Book. I
was awake enough to help lead the flock to heaven.

Just after 6 am, the first armored vehicle approached the southern corner of the main struc-
ture. Over the next three hours, armored vehicles slammed into the wooden frame building,
pumping in tear gas, severely damaging the exterior.

Helicopters firing, lots of firing. You can see how they were all ready to come in by the way they
did. Just… they were fully armed… they knew that’s what they were going to do. They were just
going to come on. We stayed down on the ground for a long time. The noise was unbearable.

April 19 0631
Loudspeaker announcement: IF YOU CAN’T SEE, WALK TOWARD THE LOUDSPEAKER,
FOLLOW OUR VOICE.

234 Log 50
Aerial view of the Branch Davidian compound near Waco, Texas, 1993. Right: Branch Davidian compound in flames on
April 19, 1993, following a 51–day siege by the FBI and law enforcement. Photos: Federal Bureau of Investigation.

I could hear crackling noises over the speaker. It was hard to understand what they were say-
ing. The next thing we knew, they kept ramming the building and it was being knocked around.

April 19 0729
Loudspeaker announcement: THIS IS NOT THE WAY TO END IT.

Television Interview Excerpt with Bystander (Female; Age: 40s)

SS: Who ever heard of using American tanks against Americans on American soil?

…there was fire above me, across the hall… I guess on both sides too… Where they
rammed the buildings was where the stairs were. And when I realized the fire was
so close, that’s when I decided to jump.

\ ATF Report \ TX Field Division \ Not for distribution \

Word from Waco Command Center was that one or two people were spotted at different loca-
tions outside the compound. ATF agents, protected by their body armor and suits of fire-resis-
tant Nomex, left their armored vehicles and went after the individuals.
By noon, whole sections of the exterior had been demolished. Portions of the roof had col-
lapsed. Cult members inside had been forced into an ever-narrowing circle of interior rooms on
the upper floor. A vehicle rammed the upper story on the northeast side of the building, a flicker
of orange could be seen a few moments later…. Not far from the compound, the flagpole with
the Star of David flag fell into the ensuing flames.
Seven individuals [names and sexes withheld] jumped from windows and were rescued.
After receiving medical attention, each was taken into custody for processing.

235 Log 50
Dallas Ledger, Arts and Entertainment {Special Report on Building Arts}: May 11, 1993
All we have are the before-and-after images of the buildings at the Mt. Carmel Center, aka the
Branch Davidian compound. We can continue to question whether there was enough room to
get out or more people would have survived if the place hadn’t been infiltrated in the manner
it was. Some have examined the video footage and noted that places of egress, like the central
stairs, were either compromised or blocked by the insertion points of the tear gas and other
methods. Some say the fires were started from the inside while others speculate exploding tear
gas containers sparked the blaze in an already precarious building and structure.

The response from locals as well as the nation has been one of quiet rage.

From Time to Jet magazine, contrasting images of the Branch Davidian compound (Ranch
Apocalypse) have been presented to the public. But there is always something missing in each
drawing, each reconstruction, each model presented. In some, the complex is a vast, factory-
like complex and in others, depending on the mode of representation and where it is published,
the buildings are just an agglomeration of postmodern pitched roofs and a pool. These differ-
ences and absences are evocative of stories being told, of narratives that bespeak fear, compla-
cency, and disregard for ground truths. The perspectives are even more deceiving, and because
we do not have any real evidence supporting what the complex looked like on the inside
(except for the made-for-TV NBC Movie of the Week In the Line of Duty: Ambush in Waco), the
public can only speculate what went on for 51 days inside the Mt. Carmel Center on Old Farm
Road. The images in the magazines and newspapers, even those on the television news, will
never approximate a real version of the compound; they are deficient doubles, like most are.
If a model is ever produced for the congressional investigation commencing, what could
be deduced from its attention to details (or lack thereof )? What remains absent is often that
which we speak about the most. Aren’t models, after all, a way to “see” inside the now-
destroyed compound built for the movie and then reused by the ATF? Not necessarily a twin
but something that is in excess of the real? All of this devastation – seen on TV and around here
– seems essential to rethink in these days of disposable humans, communities, and buildings.
How does one comprehend the so-called model community that the Branch Davidians
imagined themselves to be? We have all heard of the horrific things that were allegedly hap-
pening within the walls of the compound. We have the testimonies of those who remained
inside until the last moments. And yet, what can be said of the building of the second “com-
pound” so close to the ranch for target and infiltration practice by the ATF? Each opening
labeled with exaggerated yellow-and-black signage including “Front Door,” “Window,” and
“Chapel.” Some speculate it is soon to be burned down as well, before becoming a tourist des-
tination. Are there new models of imagining, occupying, and building spaces – beyond the
various arms of the media that routinely dissect and reconfigure them?
Today, we only have the orange flags planted by the FBI retrieval unit to indicate where
bodies were found. They tell us that there was adequate space to escape the conflagration – an
archaeology of yet another truth.

2.  John J. O’Connor, “Critic’s Notebook; On TV, Truth is More Instant than Fiction,” New York Times, June 3, 1993.

236 Log 50
 All of us watched the standoff for 51 days. We watched the compound and all those inside
destroyed by fire on television. We saw the jumpers and those that ran. I was posted for much
of those final weeks on Holy Hill, near the compound, on the opposite side of where the media
and the FBI had established their operations centers. People had started assembling around
there wearing T-shirts that had Mr. Koresh’s face printed on them. Audio cassette tapes with
David’s sermons were being sold. Others prayed.
I keep thinking of privacy. What does it look like today? And something I think I once read
by a philosopher: “There is nothing but power… And nothing is more horrifying than the
absence of an answer.”

Review of In the Line of Duty: Ambush in Waco, NBC Movie of the Week (1993)
Ambush in Waco may be as close as TV can get to producing a truly instant docudrama. The
actors were on set near the Branch Davidian compound at the same time as the 51-day siege
was unfolding. Many spoke of the “reality” that they were forced to confront while perform-
ing their scenes, the sounds of Tibetan chants echoing on their set. “At least we could head
back to Waco at the end of a day’s shooting,” one actor said, “and not be stuck out there in
between two realities.” (Ambush ends with the February 28 initial ATF assault; the fatal con-
flagration is dealt with in a somber postscript.)
The soundtrack of the 51-day siege is now being made into mixtapes and sold in and
around the Waco area with titles like Excessive Force. In 1989, a similar onslaught of music was
used in an attempt to lure Manuel Noriega out of hiding. It included songs by Judas Priest,
Black Sabbath, Styx, and Greenwood’s “God Bless the USA.” For those in Waco and those at
the Branch Davidian compound, when it was still around, it is unlikely that Barry Manilow
and Nancy Sinatra being played on repeat was necessarily a bad thing.

“Critic’s Notebook; On TV, Truth is More Instant than Fiction”

“But this particular movie, part of a generally acclaimed series called ‘In the Line of Duty,’
proved unusually absorbing as it separately traces the stories of the Davidians and the Federal
agents. Played skillfully by Tim Daly, Koresh emerged as part charismatic leader, part sex
charlatan. In this instance, there was no effort to do a triumph spin. The lingering picture of
innocents led to slaughter and ill-prepared law-enforcement agents painted a singularly
disturbing picture of life in these United States.”2

Sean Anderson is Associate Curator in the Department of Architecture and Design at the Museum of Modern Art. A Fellow
of the American Academy in Rome, he has practiced as an architect and taught in Afghanistan, Australia, India, Italy,
Morocco, Sri Lanka, and the UAE.

237
Evolution of project delivery models and digital technologies, 1970–2020. From
Phil Bernstein, Architecture Design Data: Practice Competency in the Era of
Computation, 2018. Courtesy the author.
Phil Bernstein
Canonical Models
Of Architecture
Making models is a tradition as ancient as the practice of archi-
tecture itself. Models are a strategy for projecting the future
state of a building before it is built. Whether to demonstrate
“how [something] might work”1 or to “assist calculations and
predictions,”2 they give designers a way to organize, describe,
and communicate their intentions. A combination of models
and drawings conveys a design conception that, at least until
the modern age, existed mostly in the head of its designer.
A half century before Alberti declared that thinking
about building is a separate act from actually constructing
it, designers were competing for commissions with physi-
1.  Collins Dictionary, s.v. “model,” https:// cal models.3 The proposal for the Basilica di San Petronio in
www.collinsdictionary.com/dictionary/
english/model.
Bologna was made with an enormous facsimile, almost 60 feet
2.  Oxford English Dictionary, s.v. “model,” long.4 In 1418, when Brunelleschi proposed a dome for the
https://www.lexico.com/en/definition/
model.
Duomo in Florence, he won the competition with a wooden
3. See Leon Battista Alberti, On the Art of model and a daring proposal to construct the dome without
Building in Ten Books (Cambridge: MIT Press,
1988) and Mario Carpo, The Alphabet and the
scaffolding but provided no proof that such a feat would actu-
Algorithm (Cambridge: MIT Press, 2011). ally be possible.5 Physical models, like their latter-day digital
4.  Ross King, Brunelleschi’s Dome: How a
Renaissance Genius Reinvented Architecture
counterparts, are largely an attempt to abstract the enormous
(New York: Penguin Books, 2001), 7. complexities of the finished product, prioritizing selected
5.  Ibid., 44.
characteristics (like organization and shape) while ignoring
others (like construction sequence or performance).

Model Hierarchy
In the broadest sense, architects deploy and operate in a hier-
archy of models, all of which dictate the architect’s behavior,
and within which they exercise various degrees of agency,
authority, and autonomy. Construction (and its servant,
design) is a constituent part of macroeconomic models that
convert capital into assets. Projects – at least those that result
in actual buildings – are realized within delivery models that
establish the relationships and exchanges of risk and value
between consumers (clients) and purveyors (architects, engi-
neers, constructors). The structure of project delivery rela-
tionships and the roles of designers and builders are evolving,
from the original conflation of both responsibilities, which
was the ideal of the “master builder,” to the role of design
architect versus executive architect and the reintegration of

239
 “thinking” and “making” in today’s design-build firms and
verticalized companies like Katerra and WeWork. And under-
pinning the work of the architect are various representational
models, originally analog drawings and physical, small-scale
depictions, then with computer-generated drawings with
CAD, now with the addition of three-dimensional building
information modeling (BIM), and on the cusp of a future
with simulation, analytical models, and machine learning.
These representational structures are strongly interrelated;
they bound authority, sometimes limit autonomy, and may
offer opportunities for new roles, responsibilities, and potency.

Macroeconomic Models
Comprising 13 percent of global gross domestic product, con-
struction is a fundamental contributor to global economic
6.  See Filipe Barbosa et al., McKinsey well-being.6 After having largely instantiated the produc-
Global Institute, Reinventing Construction:
A Route to Higher Productivity, McKinsey &
tivity gains of the industrial age, the building industry has
Company, February 2017. been slow to adopt the economic acceleration seen by other
7.  Ibid.
8.  See Daniel Susskind, A World without
enterprises that have leveraged the information age, realiz-
Work: Technology, Automation, and How We ing increases in productivity of only one percent per annum
Should Respond (New York: Metropolitan
Books / Henry Holt & Company, 2020), 37.
over the last 20 years, compared to 2.8 percent for the overall
9.  See Erik Brynjolfsson and Andrew economy and 3.6 percent for manufacturing.7 The increased
McAfee, The Second Machine Age: Work,
Progress, and Prosperity in a Time of Brilliant
productivity across the digital-age economy has benefited
Technologies (New York: W.W. Norton & both design and construction workers – at least as consumers
Company, 2014).
– through a macroeconomic construct known as the “canoni-
cal model,” in which positions destroyed by automation are
replaced by newly created jobs and enhanced overall produc-
tivity raises all economic boats.8 You might have lost your job
laying bricks by hand to a robot, but you replaced it with a
job in the robot factory, and the increased productivity on the
jobsite made your house less expensive. The canonical model
suggests that as both design and construction are digitized,
some jobs will disappear (drafters, for example) but they will
be replaced by others (CAD operators or BIM specialists).
Some argue that we are entering a second incarnation
of the information age in which machines replace not just
manual labor but cognitive work.9 Just as heavy machinery
and cranes revolutionized the labor of building, devices
powered by artificial intelligence are predicted to replace the
knowledge work of building. Whether the canonical model
of creative destruction and job replacement obtains remains
to be seen. But it is clear that the service economy model, in
which designers deliver their knowledge and responsibility in
exchange for value, will transform as a result of the transition
from human insight and judgment to computational output,

240 Log 50
 where models of all kinds will be created and deployed with-
out human intervention.

Delivery Models
During the design and construction of the Duomo,
Brunelleschi embodied a simple model of delivery by con-
solidating all responsibility for both design and construc-
tion under a single source: himself. The distinctions between
design and construction decisions and information were
irrelevant because he generated and controlled both. The
design idea, construction approach, even the design of the
boats that brought stone from the quarry to the project site
originated with Brunelleschi. The resulting notion of the
“master builder” has been romanticized by architects ever
since, but this yearning for consolidated control was largely
left behind, at least in the US, in the late 18th century. The
division between design and construction, in the Albertian
sense, became institutionalized permanently during the
period of Reconstruction after the Civil War. By the turn of
the 20th century, the professionalization of the practice of
architecture, combined with the establishment of the role
of the general contractor, was a necessary result of the speed
10.  See George Barnett Johnston, Assembling and complexity of construction as America modernized.10
the Architect: The History and Theory of
Professional Practice (London: Bloomsbury
The corresponding delivery model established adversarial
Visual Arts, 2020). roles and responsibilities for designers and builders, and con-
11.  Ibid., 16.
12.  See Johnston, who traces the history
tentious behavioral dynamics that persist to this day. As early as
of this dynamic and argues that architects 1888, American architects and contractors drafted a model con-
in fact catalyzed the creation of the entire
discipline of general contracting by
tract form – the so-called Uniform Contract – that attempted
deciding to forgo their responsibilities to to normalize the operational and legal terms of design and
manage and coordinate subcontractors and
construction laborers on their jobsites. construction, but they failed to define, in clear terms, the level
of detailed information the architect needed to transmit to the
builder to assure a building was completed in accordance with
a budget and a schedule.11 Builders argued that architects were,
in fact, unable and unqualified to manage such parameters and
that the responsibility was best left to them, ascribing this fail-
ure to architects’ lack of understanding of the cost of building
rather than to their inability to tell the future – a skill clearly
necessary to accurately predict the uncertainties of construc-
tion, particularly as builders were held to fixed bids based on
inexact documents from architects.12
Underpinning this design-bid-build approach, which
quickly became the canonical model of construction, was the
economic proposition of lowest first cost: a client “purchased”
the services of an architect for a fixed percentage of the con-
struction cost and then locked a contractor into a fixed price

241 Log 50
 based on the “design model” as represented by the drawings and
specifications provided by the architect, all the while transferring
the risks of execution – financial or otherwise – to the builder.
In the latter third of the 20th century, aspects of this
design-bid-build model were manipulated to address the
emergent complexities of modern construction. High inter-
est rates in the 1970s demanded faster schedules to reduce
the cost of borrowed money, and construction managers
and fast-tracked projects emerged. Architects had tradition-
ally delivered a single set of drawings and specifications to a
contractor; under the fast-tracked model, design work was
exploded into a series of packages of overlapping component
parts (foundations, superstructure, enclosure, interiors, etc.)
to be drawn, bid, and built in pieces. The contractor, now
rebranded as “construction manager,” was paid to manage
this more complex – and allegedly faster – delivery model.
Architects worked hard to coordinate the pieces and assumed
more risk for the resulting mistakes, but they largely failed
to raise fees to reflect the pain. Meanwhile, the American
Institute of Architects, which established a committee to
write and manage the model Uniform Contract at the turn
of the century, began using standard model contracts of all
forms to broadcast putative practice standards – that is, mod-
13.  I was a member of this group, known els of behavior – to the entire US building industry.13
as the Contract Documents Committee,
from 1994 until 2006. By 2006, there were
Various delivery models continued to evolve into the 21st
almost 100 separate model contracts and century. The liability crisis of the 1980s yielded design-build,
forms in the AIA Contract Documents
system, reflecting the variety of delivery in which construction companies subcontracted architects
models and conditions faced by architects and engineers under a single agreement with the client, giving
at that time.
said client one entity to sue when projects failed, rather than
both the architect and the builder. Very large projects were
organized under the program management model, where one
superpower (like Bechtel, Jacobs, or Parsons) would consoli-
date the hiring of a series of design and construction firms to
build airports or military bases with multiple component proj-
ects. The information complex of such projects – with myriad
players, drawing and specification submittals, administrative
infrastructures – was vast, laying the groundwork for the
design and construction data systems of the internet age.
Yet distress about the efficacy of the industry has lingered,
despite a veritable smorgasbord of delivery models from
which to choose. As the 20th century ended, the Australian
government, unable to find any delivery model that could
manage the inherent risk of large-scale infrastructure proj-
ects, attempted a radical departure called the project alli-
ancing model, in which the parties in a cooperative venture

242 Log 50
 comprised of the government, engineering firms, and build-
14. By the mid-1990s, no design or con- ers sign a single agreement and share the risks and rewards.14
struction firm in Australia would undertake
the risks of building, for example, a
Project alliancing formed the basis of a new delivery model in
large transcontinental highway, as it was the US that emerged simultaneously with BIM called inte-
impossible to predict the schedule, costs, or
potential pitfalls of such an undertaking. grated project delivery, or IPD. BIM supposedly provides a
See Australian Government Department of data platform that is sufficiently robust and trustable that the
Infrastructure and Regional Development,
National Alliance Contracting Guidelines: canonical design-bid-build model can be abandoned for a sys-
Guide to Alliance Contracting, Commonwealth tem in which decision-making, dispute resolution, resources,
of Australia, September 2015.
15.  Phillip G. Bernstein, “Project Delivery risks, and rewards are jointly shared by all project partici-
Methods,” in American Institute of pants. The IPD model continues to evolve today, underpinned
Architects, The Architect’s Handbook of
Professional Practice, 15th ed. (Hoboken: by developing digital technologies that provide detailed (and
Wiley, 2013), 517. presumably more reliable) information to project partners.

Representational Models
Brunelleschi won the Duomo competition with scale models
and the assertion that he could build the dome without scaf-
folding, thereby allowing construction to continue without
interfering with the liturgical activities of the church. Much
like its modern graphic counterpart (construction drawings),
Brunelleschi’s model gave no indication whatsoever of how he
was to accomplish this feat. In fact, architectural representa-
tion has been traditionally understood as representing some-
thing called the architect’s design intent – that is, the state of
a building after it has been completed – without any detailed
instructions for how it is to be procured, assembled, or, in cir-
cumstances where shop drawings are required, built.15
This assumption is deeply embedded in project delivery
models and especially the canonical design-bid-build approach.
Since the architect’s representation only depicts final intent,
the builder is entirely responsible for the means and meth-
ods of construction. This principle was established to protect
the architect from responsibility for construction site safety:
if the builder is entirely on the hook for such means and
methods, the architect can’t be held liable if someone is killed
enacting those methods. But this separation exacerbates the
tensions established with the 1888 Uniform Contract, giving
builders further ammunition to claim that the information
provided by architects, which is incorporated into the con-
tract between the owner and general contractor, is incom-
plete and inept. Ironically, this perception was magnified
by the transition from manual drawing to computer-aided
drafting and the first three-dimensional digital models in the
1990s, as computer-drawn output looked deceptively com-
plete and precise but in fact was no more so than its hand-
drawn predecessors.

243 Log 50
 Even though a CAD representation of a building was a
crude sort of digital model, in that certain underlying geome-
tries were represented computationally in the guts of the digital
drawings, the coherent understanding of the building was reli-
ant on the coordinated orthographic abstractions depicted by
plans, elevations, sections, and details, all drawn in two dimen-
sions. A certain geometric virtuosity resulted, as it was much
easier to draw (and fabricate) a curvy form in a computer-
generated diagram. A generation of architects on the cutting
edge of design thinking started using shape-making tools from
aerospace and moviemaking – the “Blobmeisters,” as described
by Robert A.M. Stern. The drawings and digital models of the
blobs, however, were still only thin simulacra of the built arti-
fact. A series of dramatic transformations, from the architect’s
conception to the working drawings, from those drawings to
a bid project, and from the contract for construction to the
building itself, remained necessary despite the potential advan-
tages of the CAD digitization of the data, even as that data had
no awareness of their tectonics or performance.
That awareness was the promise of building information
modeling when the term was coined in 2002, and the transi-
tion from traditional CAD drafting to digital modeling began
16.  I was an employee of Autodesk and in earnest.16 The term distinguished between digital three-
involved in the strategic initiative the
company named BIM that began with
dimensional modelers then commonly in use (Piranesi,
the acquisition of Revit Technology Form-Z, even AutoCAD’s z-axis) and a tool (Revit) that had
Corporation in 2002. That tool has
subsequently become widely adopted “building awareness.” A building information model under-
among AEC professionals. stood not just the building’s three-dimensional geometry but
also its component parts and their relationships. The result-
ing data was not just an assembly of digital lines, arcs, and
circles but a database of geometry, behavior, and metadata
that described characteristics of the component elements
of the building. For example, a window was a data element
that resided in this database, along with its geometry, various
parameters related to its size and configuration, and addi-
tional characteristics such as its cost, insulation value, and
glazing. Just as important, that window understood its rela-
tionship to the wall in which it resided, and its presence in
that data-object wall adjusted the configuration (and result-
ing characteristics, such as area and insulation value) of that
wall. Walls expected to be grounded on a floor (data object)
and topped with a ceiling (data object). The designer could
create and configure these elements in whatever view of the
building (database) she chose. Whereas in a CAD drawing
set, the window is a graphic representation of a window seen
in plan, elevation, section, and a window schedule, in BIM,

244 Log 50
 those graphics are reports from the underlying database, con-
sistently described across any view.
There were immediate benefits to such an approach:
drawing reports were automatically coordinated with their
counterparts, eliminating the contractor’s complaint of mul-
tiple, conflicting representations of expensive elements in
the drawings. Simple things like accurate dimensioning and
drawing numbering were accomplished with ease. Field coor-
dination of systems, which often conflicted when built from
two-dimensional drawings, was simplified. The cost of such
errors, which could be as much as the architect’s total fee, was
dramatically reduced; however, locked into canonical delivery
models, architects once again failed to convert the value of
this improved performance into higher fees. The only realized
economic benefit of the transition to BIM was the marked
efficiency of producing documents with fewer staff, but this
advantage was likely erased by the fee pressures of the 2008
financial crisis during which BIM was widely implemented.
A building information model is an epistemological asser-
tion – a knowledge structure – from which other kinds of
design and construction reasoning can extend. In its early,
visionary incarnation, BIM was seen as the model-based cure
for the ills of the building industry, a data model that was also
a platform for all constituents of the process to clearly under-
stand the evolution and intention of a design. The information
infrastructure of the model would support construction, make
cost estimating and energy analysis easy, form the basis of facil-
ities management systems, and perhaps even reform the indus-
try’s challenged delivery models and value propositions.
But the ambitious hopes for BIM remain largely unrealized,
neutered by the failure of architects to venture past the limits
of traditional delivery models and software vendors’ timidity to
invest in powerful expansions of the data potential of the tool,
17.  Recently, architects have begun resulting in frustrations on both sides.17 Almost two decades after
demanding better software and value from
Autodesk, and the company has admitted
the concept of BIM was introduced to the market, architects and
some failures to deliver that value. See engineers have replaced their drafting platform (AutoCAD) with
Tom Ravenscroft, “Listening to concerns
‘top priority’ says Autodesk following a better drafting platform (Revit), with little other result.
architects’ criticism of BIM software,” If CAD was the first technological push from behind for
Dezeen, July 30, 2020, https://www.dezeen.
com/2020/07/30/autodesk-revit-bim- architects, and BIM the second, perhaps a third may do the
software-criticism-architects/. trick. BIM arrived at a time when digital models of build-
ings demanded enough storage, computational power, and
graphic resolution to be useful in the enterprise of representing
the enormous complexity of even the simplest building. The
computers that architects could afford and that could make
BIM possible arrived at approximately the same time as the

245 Log 50
 internet. Two decades later, there is new digital infrastruc-
ture – cloud computing – making essentially infinite com-
putational power and storage available to both designers and
builders, with ubiquitous connectivity through the internet
from the design studio through the fabrication shop to the job-
site. Faster BIM, however, will not be the catalyst but rather
the potential of a new generation of computational models
derived from the platforms of the second information age:
machine learning and artificial intelligence.
Just as BIM purports to simulate building, early work in
artificial intelligence tried to replicate human cognition. It was
hardly a fair fight, however, as even the best machines of the
late 20th century were no match for the processing power of
the human brain. But as computers became dramatically more
powerful with the cloud, new strategies for capturing, repre-
senting, and reasoning about information emerged based on
data models called neural networks, where information is col-
lected by computers and organized based on probability models
of correlation. The resulting systems can absorb heterogeneous
inputs, create their own cognitive schema, and reason inferen-
tially within that context. Such systems today anticipate the text
input on an iPhone, evaluate radiological images, determine
18.  See Richard E. Susskind and Daniel credit scores, and even arbitrate routine disputes on eBay.18
Susskind, The Future of the Professions:
How Technology Will Transform the Work of
As the building industry becomes increasingly automated,
Human Experts (Oxford: Oxford University it is likely that some combination of BIM (as an underlying
Press, 2015).
organizer of building data) and these machine-learning
systems will form the basis of new constellations of models
underpinning design and construction. Right now, thousands
of building information models reside on the servers of archi-
tects, engineers, and contractors, and while those models are
hardly used to their full potential, they are a rich pool of data
from which machine learning systems can teach themselves
the basics of building. An algorithm could, for example, look
at a collection of BIM-based hospital designs and reason about
best practices, typical design strategies, and best operating pro-
tocols. Over time, as clients collect data about the performance
of their buildings, using their own machine-learning platforms,
these systems could be connected and become a priori predic-
tors of the success of a design before it is constructed or used.
It is even easier to see parallel examples in the more technical
and data-rich realms of engineering and construction.
In the interim, many of the routine tasks performed by
architects will be modeled by these systems, replacing the
human creators of working drawings and specifications, for
example, with artificial intelligence. Banal design analysis

246 Log 50
 chores – code checking, cost estimating, area computations,
coordination – will be automated, along with a long list of
engineering analyses and technical evaluations of buildings.
But the ultimate modeler – the brain of a talented designer
who is synthesizing all these inputs and coordinating all these
processes toward a compelling architectural vision – won’t be
replaced by a machine anytime soon.

Beyond the Canon

The canonical models of architecture will be under consider-
able pressure as cognitive technologies progress in the next
few decades, transforming the workplace and the economy
along with it. The building industry and its architects are
likely to be flung into a reconsideration of basic working
principles. The fundamental traditions – that professional
architects use their personal judgment and a lot of drawings
to represent their ideas and that the models of project deliv-
ery pit designers and builders against one another – will be
challenged. Digital representation and the resulting ability
to simulate, analyze, and predict the building enterprise will
demand that designers reconsider the meaning of professional
service and their relationship to the construction process.
Two distinct paths may emerge. In the first, the routine
work of design and construction could become heavily auto-
mated, with algorithms that generate preapproved designs
and digitally automated industrialized construction protocols,
making building much more like manufacturing and com-
moditizing the acts of both design and building. The occasional
special commission may require an artist architect or even an
innovative technician, but most buildings in the era of machine
learning would be created by vertically integrated building
19.  For an elegant description of the delivery systems derived from mass-customized templates.19
evolution of digitally derived, customizable
copies in architecture, see Carpo.
The more optimistic route sees architects designing pro-
tocols, deliverables, and relationships with the models of
delivery, where the power of predictive and analytical models
supports more ambitious architectural ideas and the brain-
power replaced by algorithms is redirected toward improving
the performative and aesthetic qualities of buildings. At first,
these capabilities would be applied to the technical aspects of
delivering a building – cost, schedule, coordination – that
have been at the center of the tension between designers and
builders since before the Uniform Contract. Beyond technical
analytics, new models might simulate aspects of a building
occupant’s behavior, inspiring architects toward novel and
innovative expressions of the art.

247 Log 50
 At this particular moment of social self-examination, it
has never been more important for architects to understand
the role of buildings in amplifying racist socioeconomic struc-
tures. Simulative technologies offer a chance to depict the
relationships between social, economic, and architectural sys-
tems, and provide new insights by which to revise them and
create new architectures of equity. For example, demonstrat-
ing how the investment of capital or the implementation of
infrastructure affects the economic health of a neighborhood,
starting with an individual building, connects architecture to
much broader concerns. The data models and algorithms that
create these evaluations will certainly never replace design
insight, but they can amplify it when applied appropriately.
These strategies – combining the analytical power of
next-generation technologies with the synthetic, composi-
tional, and artistic skills of a designer – will make the work
of architects more valuable, improving the results of archi-
tecture. That new value can then be reflected in refactored
relationships to models of delivery, benefiting both the pro-
viders and consumers of building. Early work in IPD projects
– where designers are rewarded for delivering specific project
outcomes rather than being paid in lump-sum fees – suggests
the potential of such an approach. The profession can tie its
compensation more closely to delivering results that sophisti-
cated modeling systems help predict during the conception of
a building, deploying design judgment in combination with
analytic support toward much-improved ends.
Unfortunately, some economists believe the canonical
model of technological creative destruction and job re-creation
may be ended by the age of machine learning. An alternate
theory called the Autor Levy Murnane (ALM) hypothesis
posits that the economy is largely comprised not of jobs but
of tasks, and that those tasks will be increasingly assumed by
smart computers. Jobs that will remain in the second infor-
mation age will be those comprised of “tasks that cannot be
substituted by automation are generally complemented by it”
20.  David H. Autor, “Why Are There Still – ALM predicts the “massacre of the Dilberts.”20
So Many Jobs? The History and Future of
Workplace Automation,” Journal of Economic
Architects are not mindless cogs in the machines of the
Perspectives 29, no. 3 (Summer 2015): 6; see information age but rather creators of the environment in
also Susskind, A World without Work, 42.
which it operates. Computation can augment rather than
replace the act of design and catalyze a new model of architec-
tural practice that delivers new value to the models of delivery.
Phil Bernstein is an architect who is
associate dean and professor adjunct
at the Yale School of Architecture.

248
Penelope Dean
Business, Actually

busi·ness mod·el
noun

1. a design for the successful operation of a business, identifying

revenue sources, customer base, products, and details of financing.1

It is commonplace that the key to a successful architec-

tural practice is a good business model. Multiple versions
abide, usually underpinned by some assumption of finan-
cial gain: ruthlessly undercut the competition by lowering
fees, as Montgomery Schuyler proposed in his 1897 spoof
“Architecture Made Easy”; double bill for services by incor-
porating new interior design departments, as Perkins & Will
did in 1960; base compensation models on results rather than
1.  Oxford English Dictionary, s.v. “business time, as Phil Bernstein suggested in 2018.2 Yet what is less
model,” https://www.lexico.com/en/
definition/business_model.
understood is that a successful career in architectural culture
2.  See Montgomery Schuyler, “Architecture today also requires a good business model. Areas long seen
Made Easy,” Architectural Record 7, no. 2
(October–December 1897): 214–18;
as antithetical to the manners of business – history, theory,
“Historical Timeline,” Perkins&Will, design, environmentalism, and social activism among them
https://history.perkinswill.com/
timeline/; Phil Bernstein, “Why the Field – have absorbed aspects of the model, sometimes consciously,
of Architecture Needs a New Business sometimes naively, sometimes inadvertently, without much
Model,” Architectural Record, June 1, 2018,
https://www.architecturalrecord.com/ expectation of financial gain. This recent expansion of the
articles/13462-why-the-field-of-architec- business model, from a delimited professional concern within
ture-needs-a-new-business-model.
3.  For recent definitions, see, for example, architectural practice to a framework for the production of
Joan Magretta, “Why Business Models architectural knowledge generally, represents a significant
Matter,” Harvard Business Review (May
2002): 86–92, Mark W. Johnson, Clayton development in architectural culture.
M. Christensen; and Henning Kagermann, What constitutes a business model is as elusive a question
“Reinventing Your Business Model,”
Harvard Business Review (December 2008): as what one accepts as an architectural model, even as the
51–59; and Andrea Ovans, “What Is a
Business Model?,” Harvard Business Review,
two concepts become indistinguishable (see, for example, one
January 23, 2015, https://hbr.org/2015/01/ form of this mutual apotheosis in BIM). For those in business,
what-is-a-business-model. For earlier
definitions, see J.E. Mulvaney and C.W.
a model can be a “story” that explains how an organization
Mann, Practical Business Models (New works, a description of how parts of an enterprise fit together,
York: John Wiley & Sons, 1976).
or simply a plan for how to make money.3 For those of us in
architecture, the model can more usefully be comprehended
through its underlying assumptions. According to manage-
ment consultant Peter Drucker, assumptions “shape any orga-
nization’s behavior, dictate its decisions about what to do and
what not to do, and define what the organization considers

249
4.  Peter F. Drucker, “The Theory of meaningful results.”4 Assumptions extend from the environ-
the Business,” Harvard Business Review
(September–October 1994): 96.
ment of an organization (its markets, technology, customers
5.  Ibid., 96, 99–100. and competitors, their values and behavior), its mission, and
6.  William Lazer, “The Role of Models in
Marketing,” Journal of Marketing 26, no. 2
the “core competencies” required to accomplish that mission.5
(April 1962): 14. As marketing scholar William Lazer observes, “Only by mak-
ing assumptions is a model molded to fit reality.”6 And that
includes architectural realities.
One of the more unlikely places the business model has
appeared recently is in the discipline of architectural his-
tory. Business models are analyzed as the subject matter of
history (for example, corporate organizations, employment
contracts, construction specifications, budget spreadsheets).
Business models supply quantitative techniques of analysis
(big data, performative criteria, what-if scenarios) for gen-
erating evidence. Business models even inform the organi-
zational structures of architectural history collaboratives
that simulate corporate practice with websites, boards of
directors, and mission statements (Aggregate Architectural
History Collaborative). And business models have furnished
the discipline with corporate language and biggering – as Dr.
Seuss’s Once-ler encouraged in The Lorax – ambitions to
become “innovative” and “new” and more interdisciplinary
and more relevant. The effects of such importations have not
been without consequences. Architectural history has frag-
mented into a landscape of niche products (archival, coun-
terfactual, critical, global, local, material, micro, minor,
New Global, oral, revisionist, et alia), and a separation has
emerged between what many historians say they do and what
they really do. While several still claim to operate under the
assumption of ideological opposition to the current politi-
cal economy (as in rejections of authorship, commerce, and
even design), so much of the scholarship is precisely, if para-
doxically, facilitated through business protocols. Contrary
to being opposed to business, the discipline is largely in tune
with it. A good business model serves even those who protest
too much. Resistance, Incorporated.
Architectural history is not the only place in the acad-
emy where aspects of the model manifest. For design prac-
titioners, especially those sustained by university paychecks,
the business model drives subject matter, professional behav-
ior, and design graphics as well: the look and sound of busi-
ness, in these cases, often compensation for not having any.
Business models provide content for practices to map (for
example, a corporation’s distribution of goods and services)
in the hope that it will yield some architecture. Corporate

250 Log 50
 strategies – “building alliances,” “collaborative partner-
ships,” “contractual business systems,” and so on – are pro-
moted as the means of “expanding modes of practice,” that
7.  See, for example, Bryony Roberts, is, more biggering.7 And attention models fuel practitioners’
“Expanding Modes of Practice,” and
Peggy Deamer for the Architecture Lobby,
need for constant visibility via tweets, likes, and followers,
“Cooperativizing Small Firms,” Log 48: proceeding from the basic business parameters of most inter-
Expanding Modes of Practice (Winter/
Spring 2020): 9–14; 99–106. net companies: attract crowds to a website and then advertise
8.  See Michael Lewis, The New New Thing: projects to those crowds.8 The first design commission for
A Silicon Valley Story (New York: W.W.
Norton & Company, 1999), 333. an office is no longer the “mother’s house” but the website
9.  Peter F. Drucker, “What Business Can and business card. Even the burgeoning scale of exhibition
Learn from Nonprofits,” Harvard Business
Review (July–August 1989): 89. and biennial culture relies on an attention-driven business
10.  See, for example, the debate “Hays, model to make architecture more “impactful” (whatever
Kwinter, Scott, Speaks: Design and Crime
Forum” and Hal Foster, “Reply to My happened to meaning?), more measurable, more performa-
Interlocutors,” Praxis 5: Architecture After tive, more justifiable. And finally, dry business graphics – pie
Capitalism (2003): 16–23.
charts, flow diagrams, bar graphs, and spreadsheets – now
routinely supplement plans, sections, elevations, axonomet-
rics, and perspectives to help pitch architectural ideas (or
are they?) with all the familiarity of a business strategy to an
audience that is presumably well versed in business. As with
architectural history, much of this work is earnestly justified
in disciplinary or social terms, but the ideals that motivate it,
that animate it, that convey it, that deliver it, are invariably
business ideals.
This widespread absorption of the business model repre-
sents a significant change in attitude. Twenty-five years ago,
business was a dirty word in the academy. It signified profit
making. It implied market culture. It meant selling out. As
Drucker observed of many working in the nonprofit sector,
academics too have “prided themselves on being free of the
taint of commercialism and above such sordid considerations
as the bottom line.”9 The few who have dared to suggest that
the business protocols of the “new economy” might provide
an alternative assumption for design theory have been vilified
for doing so.10 But today, all the evidence seems to indicate
that many in architectural culture now think business is okay
(at least in private, which is pretty much public these days as
well). Across history, theory, and design, antagonism toward
the business model is neither vehemently hostile nor clear-
cut: contradictions are everywhere as designers and historians
rhetorically reject parts of business in theory (the commer-
cial side) but instrumentally adopt other parts in practice (the
subject matter, marketing techniques, organizing formats,
and discursive language). Total opposition to the model has
given way to selective collaboration with it. That is, to yet
another business trope.

251 Log 50
 How, exactly, did this business turn come about? The
answer to that question is twofold. In the first place, the busi-
ness model’s dissipation is symptomatic of a more general
proliferation. Political theorist Wendy Brown has written
that over the last 40 years, economic conduct came to replace
political conduct as “economic values, practices, and metrics”
extended to all fields, activities, and individuals, remaking each
11.  Wendy Brown, Undoing the Demos: on its own terms.11 In this way, Brown says, “everything comes
Neoliberalism’s Stealth Revolution (New
York: Zone Books, 2015), 30.
to comport increasingly with a business model and business
12.  Ibid., 123. metrics,” irrespective of intention.12 Indeed, today the business
13.  Ibid., 36.
model cuts across and seamlessly permeates institutions, acad-
14.  Drucker, “What Business Can Learn
from Nonprofits,” 89. emies, corporations, individuals: like the mobility of theory
across multiple academic disciplines; like the interchanges of
design methods with the business world; like the proliferation
of “best practices” across various institutions; and like the
migration of digital software across diverse practices. Why?
Because the model is indifferent to subject matter. It traffics in
the universal language of data and quantities, and trades the
uncertain tendencies of the citizen for the measurable utility of
homo economicus. But above all, as Brown ominously observes,
it is because the ultimate aim of every act is always the same: to
achieve competitive advantage in the marketplace.13
That the fictions of architectural culture – whether
autonomous, critical, postcritical, formal, informal – have
been largely replaced by the realities of facts, metrics, and
performance is exactly symptomatic of the altered orienta-
tions Brown describes. Furthermore, the turn has been com-
pounded by the realization of many of the need to operate as
a nonprofit corporation of No. 1: individuals “need manage-
ment even more than business does, precisely because they lack
the discipline of the bottom line.”14 The reality today is that
“good intentions” are no longer enough. They are no proxy
for expanding reach, visibility, and impact; for attracting new
resources, supporters, and donors; for acquiring new commis-
sions, clients, and contracts. Those require a commitment to
marketing and self-promotion that in turn fulfills the demands
imposed by the business model that in turn provides the sub-
ject matter, techniques, and objectives for so many in history
and design today.
A second answer to the question lies in the triumvirate of
technology, design, and business that converged within archi-
tectural culture during the mid-1990s at the hands of prac-
titioners, theorists, and historians. It is no coincidence that
the beginnings of architecture’s business turn coincided with
the rise of business software – Microsoft Excel 5.0 (1993)

252 Log 50
 and PowerPoint 3.0 (1992) – or that the term business model
only entered common usage at the same moment after the
arrival (not to mention exponential sales) of those particu-
15.  On the recent appearance of the term lar platforms.15 The electronic spreadsheet opened up unlim-
business model, see Magretta, “Why Business
Models Matter”; and Anna Codrea-Rado,
ited possibilities in the automation of crunching numbers and
“Until the 1990s, companies didn’t have presenting data for businesses,16 and, according to manage-
‘business models,’” Quartz, April 17, 2013,
https://qz.com/71489/until-the-nineties- ment consultant Joan Magretta, “ushered in a much more
business-models-werent-a-thing/. analytic approach to planning because every major line item
16.  See “The History of Microsoft Excel,”
ExcelHelp, https://www.excelhelp.com/ could be pulled apart, its components and subcomponents
the-history-of-microsoft-excel/. analyzed and tested. You could ask what-if questions about
17.  Magretta, 89.
18.  See Kathryn A. Neeley and Michael
the critical assumptions on which your business depended. . . .
Alley, “The Humble History of the In other words, you could model the behavior of a business,”
‘Bullet,’” (paper presented at the American
Society for Engineering Education Annual even before it was launched.17 And it is exactly this demand
Conference & Exposition, Vancouver, June to model behavior that architecture was all too willing to
26, 2011).
19.  Brown, 35. supply. The behind-the-scenes numbers from business R&D
departments were turned back into narratives in PowerPoint,
the software that provided audiovisual support for front-
of-house business presentations ranging from marketing to
sales, allowing presenters to emphasize pitches through bullet
points and default visuals. The platform provided a new genre
of communication aimed at enhancing the clarity and effi-
ciency of information.18 It too came with business biases and,
along with Excel, was silently imported into architectural
culture before 3D animation software but without any of the
self-congratulatory fanfare, and without any subsequent his-
toricizing, despite archaeologies of the digital ad infinitum.
Of course, it’s hard to commemorate an ongoing dominant
paradigm. The digital is dead (or at least “post-”), long live
its business!
Brown has described this general dissipation of the busi-
ness model into daily life as “termitelike.”19 In architectural
culture in the mid-1990s, the arrival was more locustlike.
From Amsterdam to London, Los Angeles, New York, and
Rotterdam, practitioners and historians conspicuously, if
amateurishly, tried to copy business models – their meth-
ods and mindsets. The rhetoric of collaborative partnerships
projected in PowerPoint by OMA and UNStudio; Alejandro
Zaera-Polo’s shotgun marriage of a business model with
complicated disciplinary content in his text “A World Full of
Holes”; MVRDV’s what-if datascapes originating from rows
of numbers invented in Excel, extrapolated in two dimensions
in MiniCAD, and extruded into fictional cities in Form-Z; the
Architectural Association’s modeling of its Design Research
Laboratory’s pedagogy after a corporation; Greg Lynn’s
mass customization of houses: business models provided the

253 Log 50
 unlikely source for the lot of them. Moreover, they provided
the contemporary lens for historians like Mark Wigley to
reposition, retrospectively, Walter Gropius as the self-con-
fessed “coordinator” who “turned design into a form of man-
agement,” along the lines of Bruce Barton’s casting of Jesus
as the “business entrepreneur” who “picked up twelve men
from the bottom ranks of business and forged them into an
20. See Mark Wigley, “Whatever Happened organization that conquered the world.”20 This, indeed, was
to Total Design?,” Harvard Design Magazine 5:
Design Arts and Architecture (Summer
the case of the architect nobody knew being reinvented for
1998): 21; and Bruce Barton, The Man Nobody the dot-com ’90s.
Knows (Chicago: Ivan R. Dee, 2000), 4.
21.  Lewis, The New New Thing, 333.
Significantly, these enthusiastic imitations of business,
22.  On this aspect, see The Opinionator these eager efforts to capture the glamor of business, were
(R.E. Somol), “All Work and No Play,” Flat
Out 4 (Spring 2020): 3, 79–81.
more like the business models described by Michael Lewis in
his bestseller The New New Thing chronicling the rise and fall
of Netscape founder Jim Clark: “half-baked plans.”21 If the
half-baked business plan for a Silicon Valley start-up was how
one planned to make money, for architects, it was how they
might engage with contemporary culture, as no real money
was at stake for many of its late adopters. The critics who dis-
missed the architectural approaches as merely complicit with
capitalism completely misunderstood the role of the business
model here. These were not models to be taken at face value;
they were fictions. They were attempts to get buildings to
design at a moment when corporate architectural offices were
scooping up all the commissions. They were detoured instru-
ments of imagination.
But that was then. Today, the business model is taken
seriously. Too seriously. It is no longer a paradigm but a real-
ity, just like the larger shift from model as representation of
reality to model as reality. Furthermore, use of the business
model in architectural culture has changed: it has shifted
from being something novel into something overworked. And
as it has changed, the assumptions of architectural culture
have changed with it. Architecture’s obligations that were
formally understood to be external to business – to propose
alternative ways to inhabit the world – have been replaced
by obligations internal to business – collaboration, innova-
tion, expansion, metrics – as business protocols supplant the
evaluative criteria of architecture, regardless of whether
the work under consideration is condemned or celebrated.
Fictions have turned into measurable facts.22 Practitioners
are fixated on the design of their offices as historians emu-
late business protocols while dismissing them. It all begs the
Cedric Price question: If business is seen to be the answer,
what was the question?

254 Log 50
 The best way to respond may very well come from busi-
ness itself. Drucker suggested two preventive measures to
keep an organization alert and from the brink of stagna-
tion. The first measure was abandonment: every three years,
“challenge every product, every service, every policy, every
distribution channel with the question, If we were not in it
23.  Drucker, “The Theory of the already, would we be going into it now?”23 Could we start by
Business,” 102.
24.  David Byrne, “Sea of Possibilities,”
ditching the words “innovation” and “collaboration” from
in Envisioning Emotional Epistemological our discourse? The second measure was to study the noncus-
Information (New York: Pace/MacGill
Gallery, 2003). The book is unpaginated
tomers carefully: ask why those of us in the field are jaded by
but colored dots (bullets?) connect the facts, exhausted by the relentless self-promotion, and miss
chapter headings listed in contents to pages
containing text. The quote comes from the the fictions, imagination, and surprises that used to define
“Sea of Possibilities” chapter, three red architectural culture. On that note, unlikely PowerPoint
dots followed by an orange one. Yes, it’s
overwrought and confusing. artist David Byrne offers an exciting, counterintuitive way
forward: “If business is poetry, then numbers are words and
sales presentations, marketing meetings and conferences
are the salons and literary collaborations of our time.”24 In
other words, let’s take the business model and turn it into our
poetry, rather than the other way around. Nothing personal,
it’s just business.

Penelope Dean is associate professor

of architecture at the University of
Illinois at Chicago. She is the founding
editor of Flat Out.

255
Kechao Xiang, Log of Curiosity, from the project Forest Metamorphosis, 2019.
Carbon black ink on 120-GSM matte paper, 5.94 by 8.41 centimeters.