Space Tessellations

space tessellations
experimenting with
parquet deformations
space tessellations
experimenting with
parquet deformations
Edited by
Werner Van Hoeydonck
Christian Kern
Eva Sommeregger
Birkhäuser
Basel
Table of Contents
7 In Memoriam: William S. Huff (1927–2021)

From HfG Ulm to Louis Kahn’s Design Office
From Symmetry to Parquet Deformations with
Temporality and Flow
Dénes Nagy
15 Introduction
Eva Sommeregger
Research Perspectives Teaching Perspectives

23 Parquet Deformations 175 Presenting the Experiments’ Outcomes
A Subtle, Intricate Art Form Editor's note
Douglas R. Hofstadter
177 The Tiling and the Whole
39 Past and Future of William Huff’s Christian Kern
Parquet Deformations
Werner Van Hoeydonck 183 3D Parquet Deformation
183 Exercise 1: 2D Parquets, 2D Parquet Deformation
65 Grundlehre at the HfG 183 1. Transformation of the Basic Element
—A Focus on “Visuelle Grammatik” 196 2. Continuous Deformation
William S. Huff 206 3. Deformation of the Basic Structure
214 Exercise 2: 3D Parquets, 3D Parquet Deformation
91 Geometry of Structures and 215 Exercise 3: Design Concept
Its Philosophical Aesthetic Background 222 Exercise 4: Presentation Model
Cornelie Leopold
253 Cellular Space Sequences
109 The Tiles, They Are a-Changin’ 254 Exercise 1: Figure Ground
Craig S. Kaplan 262 Exercise 2: Solid and Void
268 Exercise 3: Composition and Design
127 Parametric Modeling of Parquet 274 Exercise 4: Presentation Model
Deformations: A Novel Method
for Design and Analysis
Tuğrul Yazar
143 Pattern Manipulation through

Hinged Tessellations
Jay Bonner
163 Parakeet3D: Algorithmic Re-Envisioning

of Geometrical Pattern Morphogenesis 301 Epilogue
Esmaeil Mottaghi 302 Acknowledgments
Arman Khalil Beigi Khameneh 303 Imprint
In Memoriam:
William S. Huff (1927–2021)
From HfG Ulm to Louis Kahn’s Design Office
From Symmetry to Parquet Deformations with
Temporality and Flow
Dénes Nagy
William S. Huff lecturing at the

Hochschule für Gestaltung, Ulm.
First published in William S. Huff,
An Argument for Basic Design, Ulm
12/13, March 1965, pp. 25–36.
© Photography Roland Fürst,
HfG-Archiv/Museum Ulm.
On 21 January 2021, William S. Huff passed away at the at the Department of Architecture, Carnegie Mellon
age of 93. We hoped very much that we could discuss University in Pittsburgh in 1960, but he also worked
with him the progress of this book’s preparation, but in Kahn’s office in the summers of 1961 and 1962. He
we now have a different and very sad duty: We should returned to HfG Ulm several times to teach courses
remember his life and work and pay tribute to him with throughout the 1960s. In 1974 he joined the State Uni-
the latest developments in connection with parquet versity of New York at Buffalo and taught there as an
deformation, a topic that was very special for him. associate, and later as a full professor, until 1998, when
William was born in Pittsburgh in 1927. He served he became a professor emeritus.
in the navy during World War II and was awarded the Since my background is in mathematics and the
Victory Medal and the Asiatic Pacific Area Medal in history of science, it was not obvious that we would
1945–1946. He completed his studies in architecture begin a collaboration from the late 1980s onward.
at Yale University, from which he graduated in 1952. Teaching at the College of Engineering and Applied
William, who also had Swiss roots, was eager to spend Science, Arizona State University in 1986–1988, I orga
some time in Europe, and a Fulbright Fellowship nized two local conferences, titled “Symmetry in a
made it possible. He went to the German city Ulm Cultural Context”, with the hope that artists and schol-
and studied at the Hochschule für Gestaltung, or HfG ars would come together and share some ideas that
(Ulm School of Design), in 1956–1957. After returning to could also be fruitful in different fields. Following the
the United States, he joined the office of Louis Kahn success of these two events, I decided to organize
in Philadelphia in 1958. He started his teaching career a larger international congress and exhibition in 1989,
Space Tessellations 7
for which I choose a site in Hungary, my home country, works were published in the past; moreover, there
in order to also benefit East–West cooperation. were two “holly years” of symmetry when important
Referring to C. P. Snow’s famous essay titled “Two books were published. The Dutch chemist F. M. Jaeger’s
Cultures”, I suggested a different terminology as well as Lectures on the Principle of Symmetry and Its Applica
a related approach: We have just one culture, but it is, tions in All Natural Sciences and the Scottish mathe-
using an analogy from brain research, a “split culture”, matical biologist D’Arcy W. Thompson’s On Growth and
and similar to the function of the corpus callosum Form were released in 1917. Then, in 1952, five books
that links the two hemispheres of the human brain, were published in five languages, specifically the Italian
we need some “bridges” between the two sides. Here architect C. Bairati’s La simmetria dinamica, Scienza
symmetry, which is used in various fields of art and ed arte nell’architettura classica, [Dynamic Symmetry,
science, could be one of the possible bridges that Science and Art in Classical Architecture], the Polish
helps the interactions between the two hemispheres logician and mathematician S. Jaśkowski’s O symetrii w
of culture. zdobnictwie i przyrodzie [On Symmetry in Decorative Art
Looking for international partners, I wrote to and Nature], the German-American mathematician and
William S. Huff because his series of booklets titled theoretical physicist H. Weyl’s Symmetry, the German
Symmetry, An Appreciation of Its Presence in Man’s chemist K. L. Wolf and historian of science D. Kuhn’s
Consciousness, published between 1967 and 1977, Gestalt und Symmetrie, and the Russian crystallogra-
excited me. The booklets were designed by the pher G. Wulff’s (Yu. Vul’f) Simmetriya i ee proyavlenie v
graphic artist Tomás Gonda (1926–1988). This series prirode [Symmetry and Its Manifestation in Nature].
was privately printed with support from the US Office This publishing coincidence in 1952 is quite amaz-
of Education and distributed in Northern America for ing. Among these titles, Weyl’s is the best known due to
those universities with design programs. William im- the author’s great reputation in mathematics, physics,
mediately became a strong supporter of the planned and even in philosophy. The book is available in ten
event and later became a founding member of the languages, including three different Spanish and three
International Society for the Interdisciplinary Study of different Chinese translations. Wulff’s book was actually
Symmetry (SIS-Symmetry), which maintains a duty in its third revised edition in 1952, and its original pub-
to organize triennial congresses and exhibitions. These lication in 1908 marked the start of the Russian tradi-
events, which have the overarching title “Symmetry: tion in this field, which was initiated by E. S. Fedorov’s
Art and Science”, reached many parts of the world: discovery of the 17 wallpaper groups and the 230 space
Budapest, 1989; Hiroshima, 1992; Washington, DC, 1995; groups in mathematical crystallography in 1890–1891.
Haifa, 1998; Sydney, 2001; Tihany at Lake Balaton, 2004; It was continued by, among others, A. V. Shubnikov’s
Buenos Aires, 2007; Artists’ City Gmünd, 2010; Crete, monograph in 1940 and its new edition by Shubnikov
2013; Adelaide, 2016; Kanazawa, 2019. Another event is and Koptsik in 1972, which was later translated into
currently planned in Porto, 2022. William did not miss English and given the title Symmetry in Science and Art
an SIS-Symmetry event until the last two, but he did (New York, 1974).
participate in those via Skype. In 2007, we elected him Thus, the novelty of SIS-Symmetry was not the
to the position of Honorary President, and he remained interdisciplinary interest in symmetry, but the fact that
one of the most active members of the Society. society had started to organize regular events and
During our events in Haifa and Tihany, William provided a forum for international and interdisciplinary
connected with Claudio Guerri, a leading personality in cooperation. William was always a great supporter of
semiotics and professor of architecture at the Univer- our activities, and he invited some very good artists and
sity of Buenos Aires. They became collaborators, and scholars to join us, including the Swiss geometric artist
William became a regular visitor of the conferences of Caspar Schwabe; the Australian linguist Lynn Arnold,
SEMA (Sociedad de Estudios Morfológicos de Argentina), who is the former prime minister (premier) of South
which is our partner organization in South America. Australia; and more recently, the Belgian, Vienna based
Incidentally, for our Buenos Aires congress in 2007, two artist and architect Werner Van Hoeydonck, the editor of
books were published: the bilingual English and Spanish this book.
edition of parts 2–3 of William’s series of booklets on From 1989 onward, we had a lot of opportunities for
symmetry and the reprint of the Spanish translation of personal discussions with William. He often spoke to
the German book by K. L. Wolf and D. Kuhn with the title me about his years at HfG Ulm and Kahn’s office. The
Forma y simetría. In 2011, Claudio Guerri also published Hochschule für Gestaltung in Ulm, which was named
the introductory part of William’s series of booklets on by the secondary name of the Bauhaus at Dessau
symmetry; the complete work has six parts. (Hochschule für Gestaltung in Dessau), was estab
Of course, we do not claim that our interdisciplinary lished in 1953 with initial support by the American High
approach to symmetry was new. Many fine scholarly Command for Germany, and it existed until 1968, when
8 In Memoriam: William S. Huff Dénes Nagy

William S. Huff‘s Student Card,
Academic Year 1956/1957.
© HfG-Archiv/Museum Ulm,
HfG-Ar, BDSA Hu.
the Regional Parliament withdrew the funding. William We have already mentioned the first book by
arrived at HfG Ulm, the first American student there, K. L. Wolf and D. Kuhn, titled Gestalt und Symmetrie,
in 1956, during a time of important transition in the which was published in 1952 and presented a system-
institution. The first rector, Bauhaus-trained Swiss artist atic survey of symmetric figures. It is the same work
and designer Max Bill, resigned in 1956, and Argentine that was later translated into Spanish, known as For
concrete artist, designer, educator, and philosopher ma y simetría, in Buenos Aires. The second book was
Tomás Maldonado took over the leadership. This move written jointly by K. L. Wolf and R. Wolff in 1956 (not
also meant that Bill’s arts and crafts focus in design was related, see different name spellings). Their book has a
replaced by an interdisciplinary art-science approach very long title: Symmetrie: Versuch einer Anweisung zu
with social and economic considerations, as well as by gestalthaftem Sehen und sinnvollem Gestalten, sys
direct connections with industry. tematisch dargestellt und an zahlreichen Beispielen
In this way, HfG Ulm pioneered the field of design erläutert [Symmetry: Attempt towards an Instruction
science. Maldonado invited some leading scholars in Creative Seeing and Meaningful Forms, Systemati-
to teach courses or to give lectures. The long list cally Presented and Explained with Numerous Exam-
included, among others, American designer-inventor ples]. This book has a text-volume and a figure-volume
Buckminster Fuller, the French pioneer of information with hundreds of well-selected illustrations, from sym-
science A. Abraham Moles, the American father of metric polyhedra to tracks of animal footprints, from
cybernetics Norbert Wiener, and German chemist Karl works of art to musical scores.
Lothar Wolf. William told me that Maldonado himself Wolf’s lectures on symmetry had a great impact on
gave lectures on symmetry, but he also invited K. L. William. He also shared with me a story about when
Wolf to present further details. I was pleased with H. Weyl came to give a lecture on symmetry at HfG, but
this piece of information because I knew two books I believe the guest was not actually Weyl, but perhaps
by Wolf and realized his special interest in linking someone else. Of course, Weyl could have also given a
science with art. lecture on symmetry in Ulm, since he had retired from
the Princeton Institute for Advanced Study in 1951 and earlier study of the works by K. L. Wolf and by D’Arcy
moved back to Zürich, which is not far from Ulm, but he Thompson, respectively. It is interesting to see the
passed away in December 1955, shortly before William’s scientific background of this topic. Periodic patterns and
arrival. When I checked the catalogue of the HfG Library, their symmetries were studied intensively in crystallo
which is kept at the Ulm Museum, I realized they had a graphy in order to describe the structure of solid-state
fine selection of books in both art and science, includ- materials in which equivalent atoms (ions or molecules)
ing a German translation of Weyl’s book on symmetry. are arranged. On the two-dimensional plane, there are
William also had a great time in Louis Kahn’s office. exactly 17 repetition types (wallpaper groups). In con-
I learned from him that many design offices, includ- trast, there are infinite possibilities to represent each of
ing Kahn’s, had a copy of D’Arcy Thompson’s book On those with an arbitrary repeated element. An interest-
Growth and Form. The original 1917 edition was already ing case is the study of periodic parquets (monohedral
a massive book, but the second edition, published tilings, tessellations, or mosaics) where equal copies of
in 1942, became a two-volume set with more than a given figure fill the plane without gaps and overlaps.
1,100 pages. Artists and architects were impressed by It is not surprising that such parquets were studied
Thompson’s presentation of organic shapes and his at HfG Ulm. Their library had books by Heesch, Speiser,
view on morphogenesis. He also presented interest- Wolf, and even an early paper by Buerger and Lukesh
ing problems for mathematicians, but biologists were that linked atomic structures and wallpapers. Crys-
less happy because Thompson suggested that the tallographers and mathematicians were interested
development of living organisms was determined by in extending the topic of periodic patterns in various
physical and mathematical laws, while Darwin’s natural ways—for example coloring these systematically, which
selection is just secondary. Indeed, biologists made a may represent different physical properties (magnet
drastically abbreviated edition of Thompson’s book, in ism, spin, etc.) or dealing with the three-dimensional
which those chapters that contradicted their view were cases, but it was usually required that the same unit
deleted, but they appreciated some parts of the book. be repeated. Wolf also presented patterns where the
It is interesting to add that Thompson was awarded the basic unit may grow with a similar transformation
Darwin medal in 1946. and combined this process with a rotation that led to
I guess that William’s strong interest in Thompson’s spiral structures. It is also interesting to consider the
work was due to his time in Kahn’s office. It is possible rearrangement of atoms in a given crystalline material
that when he returned to teach courses in the early due to heating or other physical effects, a process that
1960s, William suggested that the HfG buy a copy, as leads to a different structure. Artists and designers may
the HfG Library has only the 1959 reprint edition. study similar events by drawing a periodic parquet to
Incidentally, Kahn was often invited to give lectures at feel its rhythm, which is too monotonous, then using
universities, but he was usually very busy and, in some a continuous deformation in order to gradually change
cases, asked William to make a presentation. When the structure and the related rhythm, and finally reach
Kahn was offered a university position, he declined it a new periodic parquet. D’Arcy Thompson’s book devo
but nominated William. This led to William’s teaching ted an entire chapter to the theory of transformations,
career, first in Pittsburgh at Carnegie Mellon University, which he used for the comparison of related forms in
and later in Buffalo at the State University of New York. biology. In connection with the history of this idea, he
Until the closure of HfG Ulm in 1968, he was regularly referred to Vitruvius’s and especially to Dürer’s study
invited to Ulm and gave courses there. Thus, William’s of human proportions, where, for example, the facial
method of teaching basic design in the US was a con- expressions are transformed and modified by slight
tinuation of the tradition started in Ulm. variations.
Incidentally, some of the former members of HfG William gave credit to Maldonado for the devel-
were also invited to teach at leading universities. opment of the initial concept of parquet deformation
For example, Maldonado went to Princeton and then at HfG in the mid-1950s, and he was also inspired by
to Bologna and Milan, Moles became a professor in some of M. C. Escher’s drawings, but then William went
Strasbourg, Ritter in Berkeley, and Roericht in Berlin, on to elevate parquet deformation into a form of art.
though his tableware was taken into the collection of Since temporality plays a role here, it has an inter
the Museum of Modern Art in New York. Aicher de esting Oriental analogy. William compared Chinese and
signed the pictograms for the 1972 Summer Olympics Japanese handscrolls and parquet deformations. In
in Munich; the logo of Lufthansa, in collaboration with both cases, a temporal element exists when they are
Tomás Gonda; and worked for various companies. created and usually also during the process of observa-
William, while teaching basic design courses, united tion by the viewer. The direction is, however, different: in
his interest in symmetric patterns and the theory of the Far East as well as in the case of Arabic illuminated
continuous transformations, which were related to his manuscripts, the process goes from right to left, while

in Western culture it goes from left to right. In 1994 we spent some days together with the “heroes” of Ulm and
organized a conference in Tsukuba Science City, Japan, witnessed the survival of the spirit of the school. We
and invited William to present his East–West compar- should thank William for “transforming” parquet defor-
ison; his paper appeared in the proceedings Katachi U mation into an important field in basic design educa-
Symmetry (edited by T. Ogawa, K. Miura, T. Masunari, and tion and for the inspiration to see and feel the “flow”.
D. Nagy, Tokyo, 1996). William worked with a large num-
ber of students to present parquet deformations. Many Post Scriptum
of them became distinguished architects and designers, It was great publicity for parquet deformation when
and they remember William’s course. Douglas Hofstadter, who wrote the “Metamagical
Some people may ask the following question: Is Themas” column in Scientific American from January
hand drawing important in the digital age? Our answer 1981 to July 1983, devoted the very last column, except
is definitely yes. For the development of the human for a later half-column, on William’s parquet deforma-
brain and for keeping some parts active, it is essential to tions. In the subtitle he described parquet deformation
do sophisticated manual tasks. If someone would like as “a subtle, intricate art form”. Later, Hofstadter col
to draw a parquet deformation with a computer, the lected his columns into a book with additional notes
first task would be making a detailed study of the (New York, 1985), and this column became the tenth
possible transformation, and the best way to do so is chapter, just after his column on “Pattern, Poetry, and
to sketch some parts by hand. Last, but not least, the Power in the Music of Frédéric Chopin”. Symbolically,
person who draws a parquet deformation would feel Hofstadter and Huff met again for the cover design of
the flow of the composition in order to create a new our periodical Symmetry: Culture and Science in 1990.
parquet. Moreover, using the terminology of positive The layout was designed by Gunter Schmitz, a col
psychology, some people may even achieve “flow”—a league of William at both the HfG in Ulm and then the
mental state in which a person is energized and has university in Buffalo; the image on the front cover was
positive feelings. (M. Csíkszentmihályi’s book on flow designed by William; and on the back cover we used
became widely known in 1993, when coach Jimmy Hofstadter’s ambigrams “Symmetry”, “Hungary”, and
Johnson showed it to the camera during the Super “Budapest”, which he kindly made for us. Moreover, I
Bowl halftime show and claimed that he used it in signed the introduction, titled “Manifesto on (Dis)sym-
preparing his team, which then won; see Flow: The Psy metry”, with his ambigram expressing my name. There
chology of Optimal Experience, New York, 1990.) Inciden after, I often used this “signature” in our publications,
tally, in 2012, William participated, together with Claudio giving credit to Douglas Hofstadter and in recognition of
Guerri, in the Yale School of Architecture symposium the foundation of SIS-Symmetry in 1989–1990. I will do
and exhibition asking Is drawing dead? During this the same thing here.
event, many distinguished architects emphasized the But let us first jump to the proceedings of our last
importance of hand drawing. congress in Kanazawa, Japan, 2019. William, as our
William’s description of the discussions at HfG Ulm honorary president, authored an introductory essay
was often so expressive that I felt as though I were titled “Thanks, Dénes/In Gratitude: A Home for a Life’s
there, having a chat with Tomás Maldonado, walking Driven Diversions”. It is shocking to see that it became
to Nick Roericht’s house on campus, and discuss- his last paper. SIS-Symmetry and SEMA are very much
ing semiotic questions with Martin Krempen. Then, honored by these words, and we are glad he felt that
suddenly this fantasy became reality in 2003: William we could provide a “home” for his ideas. It is our duty
invited Claudio Guerri and me to the celebration of to help these ideas to flourish.
the 50th anniversary of the HfG’s founding in Ulm. We
This “signature”, where “Dénes” (five letters) flows into

“Nagy” (four letters) and vice-versa with twofold rotation,
was a gift by Douglas Hofstadter for the first SIS-Symme-
try congress in 1989.
Biography of the Author
Dénes Nagy is a Hungarian-Australian mathematician, historian of science, and theoretician

of architecture. He started his teaching career at the Eötvös Loránd University in Budapest.
His later appointments: Arizona State University (Tempe, 1986–1988), University of the South
Pacific (Suva, 1989–1993), University of Tsukuba (Tsukuba Science City, 1993–2000). In 2001,
he was appointed as an Honorary Research Professor of the Institute for the Advancement
of Research, ACU, Melbourne, Australia. He is the Founding President of SIS-Symmetry
(International Society for the Interdisciplinary Study of Symmetry), which organizes triennial
art-science congresses and exhibitions: (1) Budapest, 1989; (2) Hiroshima, 1992; (3) Wash
ington, DC, 1995; (4) Haifa, 1998; (5) Sydney, 2001; (6) Tihany at Lake Balaton, 2004; (7) Buenos
Aires, 2007; (8) Artists’ City Gmünd, Austria, 2010; (9) Crete, Greece, 2013; (10) Adelaide,
South Australia, 2016, (11) Kanazawa, Japan. In the framework of SIS-Symmetry, he worked
together with the architect William S. Huff for a longer period. He is also the Chair of the
Committee for Folk Architecture of the Hungarian Academy of Sciences, Veszprém. He
was given many honorary titles and awards, including Doctor honoris causa, Honorary
Membership of the SEMA in Buenos Aires, Award of Pro Scientia absoluta vera, and the
Katachi-prize in Japan. In 2012, the President of Hungary awarded him the Knight’s Cross.
He gave invited lectures in more than 20 countries in six languages. His list of publications
includes 26 volumes and around 200 papers in English, Hungarian, German, French,
Russian, Ukrainian, Slovenian, Japanese, Chinese, and Persian. He was the co-editor of the
Encyclopedia of Hungarian Scientists. The list of his co-authors and co-editors include,
among others, Nobel Prize winners Eugene P. Wigner and Danny Shechtman, physicist Kodi
Husimi, and historian László Makkai.

Introduction
Werner Van Hoeydonck and
Eva Sommeregger
Fig.: M.C. Escher’s Metamorphosis II. © 2021 The M.C. Escher Company–The Netherlands.
All rights reserved by mcescher.com.
Experiments the nineteenth century to sculpture in the twentieth
The Way Things Go is the title of an experimental century and art and design in the expanded field in the
work by artists Peter Fischli and David Weiss in 1987. twenty-first century. Most lecturers affiliated with the
In this piece, a chain reaction of a series of physical center have or had a firm background in the arts4—a
and chemical experiments takes place that is then fact that signals the university’s willingness to embrace
captured by an approximately 30-minute-long film. artistic exploration within the curriculum of architec-
In an inventive and playful manner, the piece involves ture. The center has, thus, always enjoyed a particular
self-made structures using water, fire, foam, fog, a status within the faculty, offering space for investiga
tire, a candle, a balloon, a banana, a chair, and a table, tions and design from an artistic perspective.
among others. These contribute to an ongoing course In addition to a number of free electives and op
of events that creates suspense over whether one of tional modules, the center runs a mandatory second-
the experiments might, at some point, abruptly end year course in the Bachelor of Architecture program
and thus interrupt the continuous connection of mu- that is rarely found within the obligatory subjects of
tually dependent settings. Shown at the Documenta 8 architectural curricula: a basic design course titled
exhibition, as well as in the permanent collections of “Three-Dimensional Design”. Freed from the usual de-
the Centre Georges Pompidou and MoMA New York, the mands regarding the ways in which a future architec
work gives the experiment a firm position within the tural design might be used or practiced, students in
field of the arts. this class deal with three-dimensional objects as they
Scientifically speaking, an experiment is an empir- are by exploring an object’s actual sculptural traits.
ical form of investigation to gain knowledge. Coined by During the course of a four-month semester and
the early modern philosopher Francis Bacon, the Latin guided by a brief, students design and build a physi-
expressions of experientia quaesita and experimentum cal three-dimensional object, sometimes using digital
describe a sought-after experience.1 Experimentation techniques, and explore its possibilities and limitations
involves observing and actively intervening in a process via sensory perception and while considering its mate-
by making intentional changes in its defining condi riality and the tools and production methods involved.
tions. In order to represent a valid form of knowledge, The student works realized in the three-dimension-
it is essential that the series of steps involved in the al design classes of the fall/winter semesters of 2017/18
experiment are repeatable so that its results can be and 2018/19 form the backbone of this book. The topics
witnessed as often as desired. of these two courses were the result of fruitful con-
Within that context, the artistic experiment pre versations and collaborations. In summer 2017, inde-
sents a special case: Unlike an experiment understood pendent researcher Werner Van Hoeydonck—architect,
in the strict scientific sense—that is, the attempt to specialist in geometry and ornamentation, and co-
attain measurable results—an artistic experiment is an editor of this book—contacted a long-time affiliate with
open research activity designed by those who conduct the center—architect, researcher, and educator Anita
it.2 Equally, an artistic experiment may not necessarily Aigner—regarding the notion of ornaments. Aigner had
be reproducible; in some cases, a process may even dealt with the topic in a series of free electives and
be deliberately set up in order not to comply with this in 2006 had organized a symposium, titled “Surface
requirement. In its urge to investigate and express, the Control”, that linked contemporary debates with the
making of art is experimental in itself—be that in its ornaments’ lesser-known historical, cultural scientific,
thorough examination of a medium or in its ways of and design-related practical dimensions.
seeing and doing things.3 Van Hoeydonck’s expertise is the field of geome-
try, and his particular interest lies in so-called parquet
Genesis of the Experiments deformations, which are best known through the
Conducted for this Book works of M. C. Escher. Parquets, also known as tiling
This book explores experiments from an artistic per- or tessellation, refer to regular, interlocking geometric
spective, using as its example the design and geome- patterns covering a surface, without gaps or overlaps.
try exercises conducted by students at the Center for Parquet deformation, accordingly, includes a transfor-
Three-Dimensional Design and Model Making, a subunit mation of the tiles, and parquet deformation draw-
of the Institute of Art and Design at Vienna’s University ings necessarily need to be viewed by observers from
of Technology. The university was founded in 1815 under one side to the other, as the involved shapes slightly
the name of Imperial-Royal Polytechnic Institute, and change from one tile to the next, with the form intro-
the Center for Three-Dimensional Design and Model duced on one end differing completely from the form
Making has been involved in the education of architects at the other. As a term, “parquet deformations” was
since as early as 1866, providing studies of three- coined by William S. Huff, professor of architecture at
dimensional form ranging from ornamental plastic in the State University of New York and at the Ulm School
16 Introduction Werner Van Hoeydonck and Eva Sommeregger

of Design. Huff started to systematically investigate ssignment and its backgrounds and methodologies at
a
parquet deformations in the assignments given to the beginning of the “3D Parquet Deformation” chapter.
basic design students in the 1960s. Unlike Escher, who Following the notion of the artistic experiment,
used manifold shapes in his artworks, such as ani- this book focuses on presenting investigations rather
mals, chess pieces, and landscapes, Huff concentrated than judging the results obtained by students as right
on geometrical forms only. In other exercises, such or wrong. The topics investigated by the students’
as the “trisections of the cube”, students built skillful works overlap in many instances, and they have been
physical models from wood, which are still hosted in grouped accordingly during the editing process of the
the archive of the Ulm School of Design, and according book. The “3D Parquet Deformation” chapter features
to Huff, were an attempt to take parquet deformations works that explore “Composition”, “Dissolving”, “Gradual
into three dimensions.5 Changes”, “in Motion”, “Materiality Matters”, “Multiplica-
The question of parquets in three dimensions in tion”, and “On Stage”. The “Cellular Space Sequences”
spired Aigner and Van Hoeydonck, and they developed chapter collects works examining “Balance”, “Crystal-
an idea for an experiment: As with Huff, the teaching line”, “Gradual Changes”, “Materiality Matters”, “Multiply”,
of a basic design course could be used as a productive and “Opening Up the Inside”. Themes do not follow any
realm for experimentation—that is, as a springboard (scientific) criteria chosen beforehand; instead, they
for designing novel geometries. This time, however, have arisen from carefully studying the students’ works
something should be examined that Huff was no lon- and distilling ideas from them—ideas that reoccur, that
ger able to investigate himself: The intended exper- blur into one another, and that do not present a sys-
iment would consist of taking parquet deformations tematic evaluation regarding a particular direction, be
from two to three dimensions. In other words, could it materiality or geometric rule sets. Rather, the topics
the planar, gradually changing drawings of parquet de- extracted from the works do not follow any given order;
formations be translated into spatial objects using the they stand next to one another to build relationships.
very same transformational logic? This would result Readers might be reminded of Jorge Luis Borges’s fic-
in a special case of a three-dimensional parquet, also tional “Chinese encyclopedia”, which excels in absurdity
called a tessellation of space (a gap-free “space filling”, and breaking the limitations of ordering systems. It is
“close packing” of polyhedra or “honeycomb”). After cited by Michel Foucault in the foreword to the Order
further discussions with the center’s head, Christian of Things publication. Animals are ordered after cate
Kern, the idea to experiment with parquet deformation gories such as “suckling pigs”, “sirens”, “belonging to the
in three dimensions was then used as the design brief Emperor”, “drawn with a very fine camelhair brush”, or
for the three-dimensional design course of fall 2017/18. “from a long way off look like flies”.6 The present book
In the following year, Peter G. Auer, architect, ed- might not be able to reproduce the absurdity of Borges’
ucator, researcher and also a long-term affiliate of Chinese encyclopedia, nor Foucault’s “shattering laugh”
the center, followed in the footsteps of the parquet when reading it for the first time7—rather the book
deformation brief and developed the so-called “Cellu- takes Borges’s encyclopedia as a point of departure not
lar Space Sequences” exercise that would once again to classify, but to present a deliberate choice.
make use of an experiment translating two-dimen Within the linear scope of this book, the student
sional geometric patterns into three-dimensional ob works can be found in the second part of the book,
jects. This time, however, students designed cellular preceded by a chapter titled “Research Perspectives”
arrangements made from solid volumes into which at the beginning. This chapter does pioneering work
connective void spaces were placed. in making the concept of parquet deformations more
accessible. It showcases important historical texts that
Presenting the Experiments’ Outcomes offer profound insight, as well as selected perspec
Approximately 450 students completed the experiment tives written by specialists in contemporary geometric
in each of the two years—a selection of their works discourse: Their articles explain methodological ap
form the core of this book and are displayed in the proaches and provide outlooks into possible future
chapters carrying the names of the respective experi- developments, proving that there exists continuous and
ments: “3D Parquet Deformation” and “Cellular Space renewed interest in parquet deformations. The chap-
Sequences”. Both chapters progress chronologically so ter also includes reprints of historic articles, as well as
that readers can follow the logic of the assignments contemporary contributions conceived by international
given to students. In both cases, initial two-dimensional scholars, carefully selected by Werner Van Hoeydonck
exercises are shown, leading to a selection of three- for this publication.
dimensional models built at the end of the semester. Douglas Hofstadter, theoretical physician and writer,
Christian Kern, also head of teaching at the center, wrote a monthly column titled “Metamagical Themas”
provides a thorough introduction into the teaching for the renowned Scientific American magazine and
made parquet deformations known to a wider audi- practical realizations in composition and design was
ence through a piece on parquet deformations in July one characteristic of the Ulm School of Design, where
1983. Hofstadter gave his kind permission for the inclu- Huff studied and taught. Leopold paints a picture of
sion of a reprint of the extended version from his 1985 the fertile climate that ultimately made it possible to
book Metamagical Themas, titled “Parquet Deforma- devise the parquet deformation assignment.
tions: A Subtle, Intricate Art Form”, supplemented with Craig S. Kaplan, from the University of Waterloo in
original images gathered from the Ulm archive for this Canada, writes from the perspective of a mathemati
publication. cian and computer scientist, albeit using code as a
Providing new material gathered from the Ulm creative medium. He has designed parquet deforma
archive, Werner Van Hoeydonck historically context tions himself and published papers on Escher’s
ualizes the invention of the first parquet deformation geometrical metamorphoses and parquet deforma-
in his chapter “Past and Future of William Huff’s Par- tions in Islamic design. His essay “The Tiles, They Are
quet Deformations”. Having been in conversation with a-Changin’” compiles his latest research on parquet
historic witnesses, he expands on Huff’s own research deformations.
and examines possible future developments and the Architect and professor Tuğrul Yazar from the Is-
role of geometric exercises in basic design courses. tanbul Bilgi University describes parquet deformations
In “Grundlehre at the HfG—A Focus on ‘Visuelle from a computational perspective. Having used a par-
Grammatik’”, William S. Huff, the inventor of parquet quet deformation assignment in teaching computer
deformations, provides an extensive overview of how aided design software, he explores novel methods to
he experienced basic design education at the Ulm analyze and design parquet deformations using state-
School of Design, both as a student and later as a of-the-art programs Rhino and Grasshopper.
teacher.8 In this essay, a reprint from 2003 and beau- Jay Bonner, independent scholar and author of
tifully illustrated by outstanding student works newly the book Islamic Geometric Patterns: Their Historical
selected by Van Hoeydonck for the present book, Huff Development and Traditional Methods of Construction,
describes the pedagogic work of Tomás Maldonado speculates on the combination of parquet defor-
and elaborates on historical and formal developments mation with Islamic design in “Pattern Manipulation
in the concepts of basic design courses—from the through Hinged Tessellations”.
preliminary courses at the Bauhaus directed by Josef Finally, in “Parakeet3D: Algorithmic Re-Envisioning
Albers to modern-day practices in basic design edu- of Geometrical Pattern Morphogenesis”, Esmaeil Mot-
cation—as witnessed by him. taghi and Arman Khalil Beigi Khameneh, lecturers at
In “Geometry of Structures and Its Philosophical the University of Tehran and inventors of the P arakeet
Aesthetic Background”, Cornelie Leopold, mathema software, open the field of systematically scripting
tician, philosopher, and professor affiliated with parquet deformations.
the University of Kaiserslautern, examines the rela- In the book’s first article, positioned at the very
tionships between mathematical, philosophical, and beginning of the publication, Dénes Nagy, mathemati
design approaches in teaching at the Ulm School cian, science historian, and architecture theoretician,
of Design, illustrated by examples of foundation provides an epitaph on William S. Huff, who sadly
courses. The strong connection between a theoretical passed away on january 21, 2021. We dedicate this
background in structural thinking and hands-on book to his memory.
References
1
Siegfried Gehrmann, Natur, Erfahrung, Experiment – Francis Bacon und die Anfänge der
modernen Naturwissenschaft. “Erfahrung” – Über den wissenschaftlichen Umgang mit einem
Begriff. Essener Unikate, 16/2001, pp. 53–63.
2
Nicole Vennemann, Das Experiment in der zeitgenössischen Kunst. Initiierte Ereignisse als
Form der künstlerischen Forschung, transcript, Bielefeld, 2018.
3
Bernd Löbach-Hinweiser. Kunst und Wissenschaft. Band 4: Experimentelle Kunst,
Designbuch, 2016.
18 Introduction Werner Van Hoeydonck and Eva Sommeregger

4
Helmut Kuhn, Walter Ritzer, 150 Jahre Technische Hochschule Wien, self-published,
1965, pp. 321–323.
5
Email from William S. Huff to Werner Van Hoeydonck, 31.08.2019.
6
Michel Foucault, The Order of Things. An Archeology of the Human Sciences, Pantheon,
New York, 1970, p. xv.
7
ibid.
8
This text first appeared in 2003. See [Ed.] Ulmer Museum and the archive of the Ulm
School of Design, Ulmer Modelle – Modelle nach Ulm: Hochschule für Gestaltung Ulm
1953–1968, Ostfildern-Ruit, Hatje Cantz, 2003.
Biography of the Authors

See p. 63
Eva Sommeregger has a degree in architecture from the Vienna University of Technology, a
master’s degree in architectural design from the Bartlett School of Architecture in London,
and a doctorate from the Academy of Fine Arts Vienna. At universities in Vienna and London,
she gained teaching experience at the bachelor, master, and dissertation levels and generated
third-party funding. In 2010 she received the Margarete Schütte-Lihotzky grant, in 2011 she
received a Schindler grant at the MAK Center Los Angeles, and in 2018 she co-founded the
MAGAZIN exhibition space for contemporary architecture in Vienna. Her architectural practice
includes writing, curating, and developing experimental (post)digital spaces. Her work has
been exhibited at MOMA New York and the Venice Architecture Biennale, among others.
Research Perspectives
21
Parquet Deformations:
A Subtle, Intricate Art Form
Douglas R. Hofstadter
What is the difference between music and visual art? but most of them struck me as being unsuccessful. One
If someone asked me this question, I would have no striking counterexample is the set of “parquet deforma-
hesitation in responding. To me the major difference is tions” meta-composed by William S. Huff, professor of
temporality. Works of music intrinsically involve time; architectural design at the State University of New York
works of visual art do not. More precisely, pieces of at Buffalo.
music consist of sounds intended to be played and I say “meta-composed” with good reason. Huff him-
heard in a specific order and at a specific speed. Music self has never executed a single parquet deformation.
is therefore fundamentally one-dimensional; it is tied He has elicited hundreds of them, however, from his
to the rhythms of our existence. Works of visual art, students, and in so doing he has brought a high degree
in contrast, are generally two- or three-dimension- of refinement to this form of art. He might be likened to
al. Paintings and sculpture seldom have any intrinsic the conductor of a fine orchestra. Although the conduc-
“scanning order” built into them that the eye must fol- tor makes no sound in the course of a performance, we
low. Mobiles and other pieces of kinetic art may change credit the person doing the job, to a great degree, for the
over time, but often without any specific initial state or quality of the sound. We can only guess at how much
final state or intermediate states. You are free to come preparation and coaching went into the performance.
and go as you please. So it is with William S. Huff. For 23 years, his stu-
There are, of course, exceptions to this generaliza- dents in Buffalo and at Carnegie Mellon University have
tion. European art has grand friezes and historic cyclo been prodded into flights of artistic inspiration, and it
ramas, and Oriental art has intricate pastoral scrolls is thanks to Huff’s vision of what constitutes quality
up to hundreds of feet long. These types of visual art that some beautiful results have emerged. Not only
impose a temporal order and speed on the scanning has he elicited outstanding work from students, but he
eye. There is a starting point and an end point. Usually, has also carefully selected what he thinks are the best
as in stories, these points represent states of relative pieces, and these he is keeping in archives. For these
calm, particularly at the end. In between, various types reasons, I will at times refer to “Huff’s creations”, but
of tension are built up and resolved in an idiosyncratic it is always in this more indirect sense of “meta-cre
but pleasing visual rhythm. The calmer end states are ations” that I will mean it.
usually orderly and visually simple, whereas the tenser I don’t wish to take credit from the students who
intermediate states are usually more chaotic and executed the individual pieces, but there is a larger
visually confusing. If one replaces “visual” by “aural”, sense of the term “credit” that goes exclusively to Huff,
virtually the same can be said of music. the person who has shaped this entire art form himself.
I have been fascinated for many years by the idea of Let me use an analogy. Gazelles are marvelous beasts,
trying to capture the essence of the musical experience yet it is not they themselves but the selective pressures
in visual form. I have my own ideas about how this can of evolution that are responsible for their species’ unique
be done; in fact, I spent several years working out a form and wondrous qualities. Huff’s judgments and com-
of visual music. By no means, however, do I think there ments have here played the role of those impersonal,
is a unique or best way to carry out this task of “trans- selective evolutionary pressures, and out of them has
lation”, and indeed I have often wondered how others been molded a living and dynamic tradition, a “species”
might attempt to do it. I have seen a few such attempts, of art exemplified and extended by each new instance.
Space Tessellations Research Perspectives 23

Fig. 1: “Fylfot Flipflop”, by Fred Watts. Basic Design
Studio of William S. Huff, Spring 1963, Carnegie
Mellon University (CIT). © HfG-Archiv/Museum
Ulm, HfG-Ar, BDSA, Hu P 01. 018.
All that remains to be addressed by way of intro visual interest must come entirely from the complexity
duction is the meaning of “parquet deformation.” and subtlety of the interplay of abstract forms. There
Actually, it is nearly self-explanatory. Traditionally, a is nothing to “charm” the eye, as there is with pictures
parquet is a regular mosaic made of inlaid wood on the of animals. There is only the unembellished perceptual
floor of an elegant room, and a deformation—well, it’s experience. Because of the linearity of this form of art,
somewhere in between a distortion and a transforma Huff has likened it to visual music. He writes:
tion. Huff’s parquets are more abstract: They are reg-
ular tessellations (or tilings) of the plane, ideally drawn Although I am spectacularly ignorant of music, tone-deaf
with zero-thickness line segments and curves. The and hated those piano lessons (yet can be enthralled by
deformations are not arbitrary but must satisfy two Bach, Vivaldi or Debussy), I have the students “read” their
basic requirements: designs as I suppose a musician might scan a work: the
1. There must be change only in one dimension, themes, the events, the intervals, the number of steps
so that it is possible to see a temporal progres- from one event to another, the rhythms, the repetitions
sion in which one tessellation gradually becomes (which can be destructive, if not totally controlled, as well
another; as reinforcing). These are principally temporal, not spatial,
2. At each stage, the pattern must constitute a compositions (although all predominantly temporal
regular tessellation of the plane—that is, there compositions have, of necessity, an element of the spatial
must be a unit cell that could combine with itself and vice versa—e.g., the single-frame picture is the basic
so that it could cover an infinite plane exactly. element of the moving picture).
(Actually, the second requirement is not usually adhered
to strictly. It would be more accurate to say that the unit What are the basic elements of a parquet deformation?
cell at any stage of a parquet deformation can be easily First there is the class of allowed parquets. On this
modified so as to allow it to tile the plane perfectly.) Huff writes: “We play a different (or rather, tighter) game
From this very simple idea emerge some stun- from Escher’s. We work with only A tiles (i.e., congruent
ningly beautiful creations. Huff explains that he was tiles of the same handedness). We do not use, as he
originally inspired, back in 1960, by the M. C. Escher does, A and A‘ tiles (i.e., congruent tiles of both handed-
woodcut “Day and Night.” In that work, forms of birds ness), although an exception to this rule is the example
tiling the plane are gradually distorted (as the eye called Dual. Finally, we don’t use A and B tiles (i.e., two
scans downward) until they become diamond shaped, different interlocking tiles), since two such tiles can
looking like the checkerboard pattern of cultivated always be seen as subdivisions of a single larger tile.”
fields seen from the air. Escher, of course, became The other basic element is the repertory of stan-
famous for his tessellations, both pure and distorted, dard deforming devices. Typical devices include length
as well as for the other haunting games he played with ening or shortening a line; rotating a line; introducing
art and reality. a “hinge” somewhere inside a line segment so that it
Whereas Escher’s tessellations are almost always can “flex”; introducing a “bump” or “pimple” or “tooth”
based on animal forms, Huff decided to limit his scope (a small intrusion or extrusion having a simple shape)
to purely geometric forms. In a way, that is like a deci- in the middle of a line or at a vertex; shifting, rotating,
sion by a composer to follow austere musical patterns, expanding, or contracting a group of lines that form a
totally eschewing anything that might conjure up a “pro- natural subunit; and variations on these themes.
gram” (that is, some kind of image or story behind the To understand these descriptions, you must realize
sounds). An effect of this decision is that the beauty and that a reference to “a line” or “a vertex” is actually a
24 Parquet Deformations: A Subtle, Intricate Art Form Douglas R. Hofstadter

r eference to a line or a vertex inside a unit cell, and get increasingly jagged, until in the middle the roles
therefore, when one such line or vertex is altered, are completely reversed. The process then continues,
all the corresponding lines or vertices that play the so that by the time you have reached the other side
same role in the copies of that cell undergo the same the lines are back to normal. If you could filter out the
change. Since some of those copies may be at 90 horizontal lines, you would see a simple pattern of
degrees (or other angles) with respect to the master quite jagged lines alternating with less jagged ones.
cell, one locally innocent-looking change may induce When these two extremely simple, independent
changes at corresponding spots, resulting in unexpect- patterns—the horizontal and the vertical—are super
ed interactions whose visual consequences can be posed, what emerges is an unexpectedly rich percep-
quite exciting. Without further ado, let us proceed to tual feast. At the far left and right the eye picks out
examine some specific pieces. fylfots—that is, swastikas—of either handedness
Look at the piece titled “Fylfot Flipflop”. It is an early contained inside perfect squares. In the center the eye
one, executed by Fred Watts at Carnegie Mellon in 1963. immediately sees that the central fylfots are all gone,
If you simply let your eye skim across the top line, you replaced by perfect crosses inside pinwheels.
will get the distinct sensation of scanning a tiny moun- And then a queer perceptual reversal takes place. If
tain range. At each edge, you begin with a perfectly flat you just shift your focus of attention diagonally by half
plain and then move into gently rolling hills, which be- a pinwheel, you will notice that there is a fylfot right
come taller and steeper, eventually turning into jagged there before your eyes! In fact, suddenly fylfots appear
peaks; past the center point these start to soften into all over the central section, where before you had been
lower foothills, which gradually tail off into the plain seeing only crosses inside pinwheels. And conversely, of
again. This much is obvious even at a casual glance. course, now when you look at either end, you will see
Subtler to see is the line just below, whose zigging and pinwheels everywhere with crosses inside them. No fyl-
zagging is 180 degrees out of phase with the top line. fots! It is an astonishingly simple design, yet the effect
Notice that in the very center that line is completely at catches nearly everyone off guard.
rest: a perfectly horizontal stretch flanked on each side This is a simple example of the ubiquitous visual
by increasingly toothy regions. Below it there are seven phenomenon called “regrouping”, in which the boundary
more horizontal lines. Thus, if one completely filtered line of the unit cell shifts so that structures jump out at
out the vertical lines, one would see nine horizontal the eye that before were completely submerged and in-
lines stacked above one another, the odd-numbered visible, whereas conversely, of course, structures that a
ones jagged in the center, the even-numbered ones moment ago were completely obvious are now invisible,
smooth in the center. having been split into separate conceptual pieces by
Now what about the vertical lines? Both the left- the act of regrouping, or shift of perceptual boundaries.
hand and the right-hand borderlines are perfectly It is both a perceptual and a conceptual phenomenon, a
straight vertical lines. Their immediate neighbors, how delight to the subtle combination of eye and mind that
ever, are as jagged as possible, consisting of repeated is most sensitive to pattern.
90-degree bends, back and forth. The next vertical line For another example of regrouping take a look at
nearer the center is practically straight up and down “Crossover”, executed by Richard Lane, also at Carnegie
again. Then there is a wavy line again, and so on. As Mellon in 1963. Something really amazing happens in
you move across the picture you see that the jagged the middle, but I won’t tell you what. Just find it yourself
lines gradually get less jagged and the straight ones by careful looking.
Fig. 2: “Crossover”, by Richard Lane. Basic Design Studio of William S. Huff, Spring 1963,
Carnegie Mellon University (CIT). © HfG-Archiv/Museum Ulm, HfG-Ar, BDSA, Hu P 01. 015.

By the way, there are still features left to be ex can’t we also imagine adding a few extra years to his
plained in “Fylfot Flipflop”. At first it appears to be lifetime to let him explore it? After all, the only divinely
mirror-symmetric. For instance, all the fylfots at the imposed (that is, absolutely unslippable) constraint on
left end are spinning counterclockwise, while all the Bach’s years is that they and Mozart’s years add up to
ones at the right end are spinning clockwise. So far, so 100, no? Hence, if we give Bach five extra years, then
symmetrical. In the middle, however, all the fylfots go we merely take five away from Mozart. It is painful, to
counterclockwise. This surely violates the symmetry. be sure, but not all that bad. We could even let Bach
Furthermore, the one-quarter-way and three-quarter- live to 100! (Mozart would never have existed.)
way stages of the deformation, which ought to be Although it is difficult to imagine and impossible
mirror images of each other, bear no resemblance at all to know what Bach’s music would have been like if he
to each other. Can you figure out the logic behind this had lived in the twentieth century, it is certainly not im-
subtle asymmetry between the left and right sides? possible to know what Steve Reich’s music would have
This piece also illustrates yet another way parquet been like if he had lived in this century. In fact, I am
deformations resemble music. A unit cell—or rather, listening to a record of it right now. Now, Reich’s music
a vertical cross-section consisting of a stack of unit really is conscious of perceptual psychology. All the way
cells—is analogous to a measure in music. The regular through he plays with perceptual shifts and ambigu-
pulse of a piece of music is given by the repetition of ities, pivoting from one rhythm to another, from one
unit cells across the page. And the flow of a melodic harmonic origin to another, constantly keeping the lis-
line across measured boundaries is modeled by the tener on edge and tingling with nervous energy. Imag-
flow of a visual line—such as the mountain-range ine a piece resembling Ravel’s Bolero, only with a much
lines—across many unit cells. finer grain size, so that instead of its having roughly a
Bach’s music is always called up in discussions one-minute unit cell it has a three-second unit cell. Its
of the relation between mathematical patterns and changes are so tiny that sometimes you can barely tell
music, and this occasion is no exception. I am re- it is changing at all, whereas at other times the chang-
minded particularly of some of Bach’s texturally more es jump out at you. What Reich piece am I listening to?
uniform pieces, such as certain preludes from “The Well, it hardly matters, since most of his music satisfies
Well-Tempered Clavier”, where in each measure there is this characterization, but for the sake of specificity you
a certain pattern executed once or twice and possibly might try Music for a Large Ensemble, Octet, or Violin
more times. From measure to measure this pattern Phase. Let us now return to parquet deformations.
undergoes a slow metamorphosis, meandering in the “Dizzy Bee”, executed by Richard Mesnik at Carnegie
course of many measures from one region of harmonic Mellon in 1964, involves perceptual tricks of another
space to far-distant regions and then slowly returning kind. The left side looks like a perfect honeycomb or,
by some circuitous route. For specific examples you somewhat less poetically, a perfect bathroom floor.
might listen to (or look at the scores of) Book I, No. 1 When we move to the right, its perfection seems in
and No. 2, and Book II, No. 3 and No. 15. Many of the doubt as the rigidity of the lattice gives way to shapes
other preludes have this feature in places, although not that seem rounder. Then we notice that three of them
for their entirety. have combined to form one larger shape: a superhexa-
Bach seldom deliberately set out to play with gon made up of three rather squashed pentagons. The
the perceptual systems of his listeners. Artists of his curious thing is that if we now sweep our eyes from
century, although they occasionally played perceptual right to left back to the beginning, we can no longer
games, were considerably less sophisticated about, see the left side in quite the way we saw it before. The
and less fascinated by, issues we now deem part of small hexagons now are constantly grouping them
perceptual psychology. Such phenomena as regrouping selves into threes, although the grouping changes
would have intrigued Bach, and I sometimes wish he quickly. In our mind we experience “flickering clusters”
had known of certain effects and had been able to try where groups form for an instant and then disband,
them out, but then I remind myself that whatever time their components immediately regrouping in new com-
Bach might have spent playing with newfangled ideas binations. The poetic term “flickering clusters” comes
would have had to be subtracted from his time for from a famous theory of how water molecules behave,
producing the masterpieces we know and love, and so in which the bonds are hydrogen bonds rather than
why tamper with something so precious? mental ones.
On the other hand, I do not find this argument 100 Even more dizzying, perhaps, is “Consternation”,
percent compelling. Who says that if you are going to executed by Scott Grady at SUNY at Buffalo in 1977.
imagine playing with the past, you have to hold the life This is another parquet deformation in which hexagons
times of famous people constant in length? If we can and cubes vie for perceptual supremacy. It is so com-
imagine telling Bach about perceptual psychology, why plex and agitated in appearance that I scarcely dare to

Fig. 3: “Dizzy Bee”, by Richard Mesnik.
Basic Design Studio of William S. Huff,
Spring 1965, Carnegie Mellon University
(CIT). © HfG-Archiv/Museum Ulm, HfG-
Ar, BDSA, Hu P01. 023.
Fig. 4: “Consternation”, by Scott Grady.

(CIT). © HfG-Archiv/Museum Ulm,
HfG-Ar, BDSA, Hu P05. 014.
ttempt an analysis of it. In its intermediate regions I

a that the unit cell giving rise to the tessellation occurs
find the same exciting kind of visual pseudochaos as both vertically and horizontally, so that flexing it one way
exists in Escher’s best deformations. induces a crosswise flexing as well, and the two flexings
Perhaps irrelevantly, but I suspect not, the names combine to yield this curious and unexpected pattern.
of many of these studies remind me of pieces by Zez Another deformation that shows the amazing
Confrey, a composer best known in the 1920s for his results of an extremely simple but carefully chosen
novelty piano pieces such as “Dizzy Fingers” and “Kitten transformation principle is “Y Knot”, executed by Leland
on the Keys” and my favorite, “Flutter By, Butterfly”. Chen at SUNY at Buffalo in 1977. If you look at it with
Confrey specialized in pushing rag music to its limits full attention, you will see that its unit cell is in the
without losing musical charm, and some of the results “Cucaracha” shape of a three-bladed propeller and
seem to me to have a saucy, dazzling appeal not unlike that the unit cell never changes in shape. All that does
the jazzy appearance of this parquet deformation. change is the Y lodged tightly inside the unit cell. And
The next parquet deformation, “Oddity out of Old the only way the Y changes is by very slowly rotating
Oriental Ornament”, executed by Francis O’Donnell at clockwise. Admittedly, in the final stages of rotation
Carnegie Mellon in 1966, is based on an extremely sim- this forces some previously constant line segments to
ple principle: the insertion of a “hinge” in one single line extend themselves a bit, but that does not change the
segment and the subsequent flexing of the segment at outline of the unit cell in any way. It is remarkable what
that hinge. The reason for the stunningly rich results is well-chosen simplicity can do.

Fig. 5: “Oddity out of Old Oriental
Ornament”, originally by Francis D.
O’Donnell, Spring 1966, reworked by
Kathleen Harrigan, Spring 1987. Note
by Huff: “Variant: keeping a constant
module.” Basic Design Studio of William
S. Huff, Carnegie Mellon University
HfG-Ar, BDSA, Hu 01. 027.
Three of my favorites are “Quirky Cogs” (Arne this one can be broken up into long, wavy horizontal
arson, Carnegie Mellon, 1963), “Trifoliolate” (Glenn
L lines and vertical structures crossing them. It is a little
Paris, Carnegie Mellon, 1966) and “Arabesque” (Joel easier to see them if you start at the right side. For
Napach, SUNY at Buffalo, 1979). They all share the instance, you can see that just below the top there
feature of getting increasingly intricate as you move to is a long snaky line with numerous little nicks in it,
the right. Most of the preceding deformations do not undulating its way to the left and in so doing shedding
have this extreme quality of irreversibility—that is, the some of those nicks, so that at the very edge it has
ratcheted quality signaling that an evolutionary process degenerated into a perfect square wave, as such a
is taking place. I can’t help wondering if the design- periodic waveform is called in Fourier analysis. Com-
ers did not think they had painted themselves into a plementing this horizontal structure is a similar ver-
corner, particularly in the case of “Arabesque”. Is there tical structure that is harder to describe. The thought
any way you can back out of that supertangle except that comes to my mind is that of two ornate, rather
by retrograde motion, namely by retracing your steps? I rectangular hourglasses with ringed necks, one on top
suspect there is, but I wouldn’t care to try to find it. of the other. You can see for yourself.
As a contrast, consider “Razor Blades”, an As with “Fylfot Flipflop”, each of these patterns by
extended study in relative calmness. It was executed itself is intriguing, but of course the real excitement
at Carnegie Mellon in 1966, but unfortunately is un- comes from the daring act of superposing them. Inci
signed. Like “Fylfot Flipflop”, the first piece I described, dentally, I know of no piece of visual art that better
Fig. 7: “Quirky Cogs”, by Arne Larson.

(CIT). © HfG-Archiv/Museum Ulm, HfG-
Ar, BDSA, Hu P 01. 005.
Fig. 8: “Trifoliolate”, by Glenn Paris.

HfG-Ar, BDSA, Hu P 01. 029.

Fig. 6: “Y Knot”, by Leland Chen. Basic Design Fig. 9: “Arabesque” — II, by Joel Napach. Basic
Studio of William S. Huff, Spring 1977, SUNY Design Studio of William S. Huff, Spring 1979,
at Buffalo. © HfG-Archiv/Museum Ulm, HfG- redone Fall 1980, SUNY at Buffalo. © HfG-Archiv/
Ar, BDSA, Hu P 01. 030. Editorial note: for the Museum Ulm, HfG-Ar, BDSA, Hu P 05. 017.
right orientation you have to turn the book 90° Editorial note: for the right orientation you have to
clockwise. turn the book 90° clockwise.

Fig. 10: “Razor Blades”, anonymous. Fig. 11: “Cucaracha”, by Jorge Gutierrez.
Basic Design Studio of William S. Huff, Basic Design Studio of William S. Huff,
1966, Carnegie Mellon University (CIT). May 1977, SUNY at Buffalo. © HfG-
© HfG-Archiv/Museum Ulm, HfG-Ar, Archiv/Museum Ulm, HfG-Ar, BDSA,
BDSA, Hu P 08. 019. Hu P 01. 031.
Editorial note: for the right orientation Editorial note: for the right orientation
you have to turn the book 90° clockwise. you have to turn the book 90° clockwise.

captures the feeling of beauty and intricacy of a Steve cusps, explicit squarish swastikoids and implicit circular
Reich piece, created by slow “adiabatic” changes floating holes. Rather than demonstrating my inability to analyze
on top of the chaos and dynamism of the lower-level the ferocious complexity of the design, I should like to
frenzy. Looking back, I see I began by describing this use it as the jumping-off point for a brief discussion of
parquet deformation as “calm”. Well, what do you know? computers and creativity.
Perhaps I would be a good candidate for one of The New Some totally new things are going on in this partic
Yorker’s occasional notes titled “Our Forgetful Authors”. ular parquet deformation, things that have not appeared
More seriously, there is a reason for this inconsis- in any previous one. Notice the hollow circles at the left
tency. One’s emotional response to a given work of that shrink as you move to the right; notice also that at
art, whether the work is visual or musical, is not static the right there are hollow “anticircles” (concave shapes
and unchanging. There is no way of knowing how you made from four circular arcs turned inside out) that
will respond the next time you hear or see one of your shrink as you move to the left. Now, according to Huff,
favorite pieces. It may leave you unmoved or it may thrill such an idea had not appeared in any previously created
you to the bone. It depends on your mood, on what has deformations. This means that something unusual hap-
recently happened, on what happens to strike you, and pened here: something genuinely creative, something
on many other subtle intangibles. One’s reaction can unexpected, unpredictable, surprising, intriguing, and,
even change in the course of a few minutes. And so I not least, inspiring to future creators.
won’t apologize for this seeming lapse. So the question naturally arises: Would a comput-
Let us now look at “Cucaracha”, executed by Jorge er have been able to invent this parquet deformation?
Gutiérrez at SUNY at Buffalo in 1977. It moves from the Well, put it this way: It is a naïve and ill-posed question,
utmost geometricity—a lattice of perfect diamonds— but we can try to make some sense of it. The first thing
through a sequence of gradually more arbitrary modi to remind ourselves of is that the term “computer”
fications until it reaches some kind of near-freedom—a refers to nothing more than an inert hunk of metal and
dance of strange, angular, quasi-organic forms. This semiconductors. To go along with this bare computer,
fascinates me. Is entropy increasing or decreasing in this this hardware, we need software and energy. The former
rightward flow toward freedom? is a specific pattern inserted into the hardware, binding
A gracefully spiky deformation is the one wittily ti it with constraints and yet imbuing it with goals; the lat-
tled “Beecombing Blossoms”, executed by Laird Pylkas ter is what breathes “life” into it, making it act according
at SUNY at Buffalo in 1983. Huff told me Pylkas struggled to those constraints and goals.
for weeks with it and at the end, when she had resolved The next point is that the software is what really
her difficulties, she mused: “Why is it that the obvious controls what the machine does; the hardware simply
ideas always take so long to discover?” obeys the software’s dictates, step by step. And yet
As our last study, let us take “Clearing the Thicket”, the software could exist in a number of different
executed by Vincent Marlowe at SUNY at Buffalo in 1979, “instantiations”: realizations in different computer lan-
which mixes straight lines and curves, right angles and guages. What really counts about the software is not
Fig. 12: “Beecombing Blossoms”, by

Laird B. Pylkas. Basic Design Studio of
William S. Huff, Spring 1983, SUNY at
Buffalo. © HfG-Archiv/Museum Ulm,

Fig. 13: “Clearing the Thicket”, by
Vincent Marlowe. Basic Design Studio
of William S. Huff, Spring 1979, SUNY
at Buffalo. © HfG-Archiv/Museum Ulm,
its literal aspect but a more abstract, general, overall sense of the momentum but would lose information
“architecture”, which is best described in a nonformal about the position. When you are looking at just a
language such as English. We might say that the plan, single work of art, however, there is no mental blurring
the sketch, the central idea of a program is what we of its style with that of recent works or soon-to-come
are talking about here, not its final realization in some works; you have exact position information (“What
specific formal language or dialect. That is something is the style now?”) but no momentum information
we can leave to apprentices to carry out after we have (“Where was the style and where is it going?”).
presented them with our informal sketch. Some years ago, A. Michael Noll, a mathematician
So the question actually becomes less mundane- and computer artist, took a single Mondrian painting
sounding and more theoretical and philosophical: —an abstract, geometric study with seeminly random
Is there an architecture to creativity? Is there a plan, a elements—and from it extracted some statistics on
scheme, a set of principles that, if it were elucidated the patterns. Given these statistics, he programmed
clearly, could account for all the creativity embodied a computer to generate numerous pseudo-Mondrian
in the collection of all parquet deformations, past, paintings having the same or different values of these
present, and future? randomness-governing parameters. Then he showed
Note that we are asking about the collection of the results to viewers who had no foreknowledge of
parquet deformations, not about some specific work. what he was up to. The reactions were interesting:
It is a truism that any specific work of art can be More people preferred one of the pseudo-Mondrians
re-created, even re-created in various slightly novel than preferred the genuine Mondrian!
ways, by a programmed computer. This is quite amusing, even provocative, but it also
For example, the Dutch artist Piet Mondrian is a warning. It proves that a computer can certainly
evolved a highly idiosyncratic, somewhat cryptic style be programmed, after the fact, to imitate—and imitate
of painting over a period of many years. You can see, if well—mathematically capturable stylistic aspects of
you trace his development over time, exactly where he a given work. But it also warns us: Beware of cheap
came from and where he was headed. If you focus on imitations!
a single Mondrian work, however, you cannot sense this Consider parquet deformations. Undoubtedly a
stylistic momentum, this quality of dynamic, evolv- computer could be programmed to do any specific
ing style that any great artist has. Looking at one work parquet deformation or minor variations on it without
in isolation is like taking a snapshot of something in too much trouble. There simply are not that many pa-
motion: You capture its instantaneous position but rameters to any given deformation. But the essence of
not its momentum. Of course, the snapshot might any artistic act lies not in selecting particular values for
be blurred somehow, in which case you would get a certain parameters but far deeper: in the balancing of

myriad intangible and mostly unconscious mental forc- game to check out all the logical consequences of a
es, a judgmental act that results in many conceptual move. Good chess moves spring from the organization
choices eventually adding up to a tangible, perceptible, of a good chess mind: a set of perceptions arranged
measurable work of art. in such a way that certain kinds of ideas leap to mind
Once the finished work exists, scholars looking at when subtle patterns or cues are present. The way
it may seize on certain of its qualities that lend them- perceptions have of triggering old and buried mem-
selves to being easily parametrized. Anyone can do ories underlies skill in any type of human activity, not
statistics on a work of art once the work is there to be only chess. It is just that in chess the skill is particularly
scrutinized, but the ease of doing so can obscure the deceptive, because after the fact it can all be justified
fact that no one could have said a priori what kinds of by a logical analysis, a fact that seems to hint the origi-
mathematical observables would turn out to be rele- nal idea came from logic.
vant to the capturing of stylistic aspects of the as-yet- Writing lovely melodies is another one of those
unseen work of art. deceptive arts. To the mathematically inclined, notes
Huff’s own view on this question of mechanizing seem like numbers and melodies like number patterns.
the art of parquet deformations closely parallels mine. Therefore all the beauty of a melody seems as if it
He believes some basic principles could be formulated ought to be describable in some simple mathematical
at the present time that would enable a computer to way. So far, however, no formula has produced even
generate relatively stereotyped yet novel creations of a single good melody. Of course, you can look back
its own. He stresses, however, that his students occa- at any melody and write a formula that will produce
sionally come up with rule-breaking ideas that enchant it and variations on it. But this is retrospective, not
the eye for reasons deeper than any he has been able prospective. Lovely chess moves and lovely melodies
to verbalize. In this way the set of explicit rules gets (and lovely theorems in mathematics) have this in
gradually enlarged. Comparing the creativity that goes common: Every one has idiosyncratic nuances that
into parquet deformations with the creativity of a great seem logical a posteriori but are not easy to anticipate.
musician, Huff writes: To the mathematical mind, chess-playing skill and
melody-writing skill and theorem-writing skill seem
I don’t know about the consistency of the genius of Bach, obviously formalizable, but the truth turns out to be
but I did work with the great American architect Louis more tantalizingly complex than that. Too many subtle
Kahn (1901–1974) and suppose it must have been some- balances are involved.
what the same with Bach. That is, Kahn, out of moral, So it is with parquet deformation, I reckon. Each
spiritual and philosophical considerations, formulated one taken alone is in some sense mathematical. Taken
ways he would and ways he would not do a thing in as a class, however, they are not mathematical. This
architecture. Students came to know many of his ways, is what is tricky about them. Don’t let the apparently
and some of the best could imitate him rather well mathematical nature of an individual deformation fool
(although not perfectly). But as Kahn himself developed you; the architecture of a program that could create
he constantly brought in new principles that brought all these parquet deformations and more good ones
new transformations to his work, and he even occasion would have to incorporate computerized versions of
ally discarded an old rule. Consequently he was always concepts and judgments, and those are much more
several steps ahead of his imitators who knew what was elusive and complex than mere numbers.
but couldn’t imagine what will be. And so it is that com- At this point, many critics of computers and arti
puter-generated “original” Bach is an interesting exercise. ficial intelligence, eager to find something “computers
But it isn’t Bach—that unwritten work that Bach never got can’t do” (and never will be able to do), often go too
to, the day after he died. far: They jump to the conclusion that art and, more
generally, creativity are fundamentally uncomputeriz-
The real question is: What kind of architecture is respon- able. This is hardly the implied conclusion! The implied
sible for all these ideas? Or is there any one architecture conclusion is that if computers are to be enabled to
that could come up with them all? I would say that the act human, we shall have to wait until we have good
ability to design good parquet deformations is probably computer models of such human properties as per-
deceptive in the same way as the ability to play good ception, memory, mental categories, learning, and so
chess is: It seems more mathematical than it actually is. on. We are a long way from that. There is no reason to
A brilliant chess move, once the game is over and assume, however, that those goals are in principle un-
can be viewed in retrospect, can be seen as logical, as attainable, even if they remain far off for a long time.
“the correct thing to do in that situation”. But brilliant In this article, I have been playing with the double
moves do not originate from the kind of logical analysis meaning of the term “architecture”: It means both
that occurs after the game; there is no time during the the design of a habitat and the abstract essence of a

grand structure of any kind. The former has to do with was a professor of architecture. But after I had con-
hardware and the latter with software. In a certain versed with him and his colleagues, my horizons were
sense, William Huff is a professor of both brands of extended about the nature of their discipline.
architecture. Obviously his professional training is in The architect Louis Kahn had great respect for the
the design of “hardware”, namely genuine habitats for work of William S. Huff, and it is with his words that I
human beings, and he works in a school where that is would like to conclude:
what they do. He is also, however, in the business of
forming in the minds of his students a softer kind of What Huff teaches is not merely what he has learned
architecture: the mental architecture that underlies from someone else but what is drawn from his natural
the skill to create beauty. Fortunately for him he can gifts and belief in their truth and value. In my belief
take for granted the complexity of a human brain as what he teaches is the introduction to discipline under
the starting point on which to build this architecture. lying shapes and rhythms, which touches the arts of
But even so, there is a great art to instilling a sense for sight, the arts of sound and the arts of structure. It
beauty and novelty. teaches students of drawing to search for the abstract
When I first met Huff and saw how abstract and and not the representational. This is good as a reminder
seemingly impractical were the marvelous works of order for the instructors/architectural sketchers (like
produced in his design studio—ranging from parquet me), and good especially for the student sketchers
deformations to strange ways of slicing a cube to without background. It is the introduction to exactitudes
gestalt studies using thousands of dots to eye-bog- of the kind that instill the religion of the ordered path.
gling color patterns—at first I wondered why this man
Fig. 14: One genuine Mondrian plus three computer imitations. Can you spot the Mondrian?
If you rotate the figure so that east becomes south, it will be the one in the northwest
corner. The Mondrian, done in 1917, is titled “Composition with Lines”; the three others,
done in 1965, comprise a work called “Computer Composition with Lines”, and were created
by a computer at Bell Telephone Laboratories at the behest of computer tamer A. Michael
Noll. The subjectively “best” picture was found through surveys; it is the one diagonally
opposite the genuine Mondrian!

Post Scriptum
“The religion of the ordered path”—a lovely phrase. for the seeing by an acute observer. This perceptual
Though it is ambiguous, it captures the spirit of the act gets you more than half the way to genuine cre-
dedicated quest after patterned beauty, and particu ativity; the remainder involves treating this new quality
larly after the reasons that certain particular patterns as if it were an explicit knob: “twiddling” it as if it were
are beautiful. a parameter that had all along been in the program
In this article, I repeatedly claimed that it is rela- that made the creation.
tively easy to make a computer program that creates That way, the perceptual process is intimately
attractive art within a formula, but not at all easy to linked up with the generative process: A loop is closed
make a computer program that constantly comes up in which perceptions spark new potentials and experi
with novelty. Some people familiar with the computer mentation with new potentials opens up the way
art produced in the last couple of decades might pick for new perceptions. The element lacking in current
a fight with me over this. They might point to complex computer art is the ion of perception with generation.
patterns produced by simple algorithms, and then add Computers do not watch what they do; they simply
that there are certain simple algorithms which, when do it. When programs are able to look at what they’ve
you change merely a few parameters, come up with done and perceive it in ways that they never antic-
astonishingly different patterns that no human would ipated, then you’ll start to get close to the kinds of
be likely to recognize as being each other’s near kin. insight-giving disciplined exercises that Louis Kahn
An example is a very simple program I know, which fills was speaking of when he wrote of the “religion of the
a screen with rapidly changing sixfold-symmetric dot ordered path”.
patterns that look like magnified snowflakes; in just One of my favorite parquet deformations is called
a few seconds, any given pattern will dissolve and be “I at the Center”, by David Oleson at Carnegie Mellon
replaced by an unbelievably different sixfold-symmetric in 1964. This one violates the premise with which I
pattern. I have stood transfixed at a screen watching began my article: one-dimensionality. It develops its
these patterns unfold one after another, unable to central theme—the uppercase letter “I”—along two
anticipate in the slightest what will happen next—and perpendicular dimensions at once. The result is one of
yet knowing that the program itself is only a few lines the most lyrical and graceful compositions that I have
long! I have seen small changes in mathematical for- seen in this form.
mulas produce enormous visual changes in what those I am also pleased by the metaphorical quality
formulas represent graphically. it has. At the very center of a mesh is an I—an ego;
The trouble is, these parameter-based changes touching it are other things—other I’s—very much like
—knob-twiddlings, as I like to call them—are of a dif the central I, but not quite the same and not quite
ferent nature than the kinds of novel ideas people come as simple; then, as one goes further and further out,
up with when they vary a given idea. For a machine to the variety of I’s multiplies. To me this symbolizes a
make simple variants of a given design, it must possess web of human interconnections. Each of us is at the
an algorithm for making that design which has explicit very center of our own personal web, and each one
parameters; those parameters are then modifiable, as of us thinks: “I am the most normal, sensible, com-
with the pseudo-Mondrian paintings. prehensible individual.” And our identity—our “shape”
But the way people make variations is quite dif in personality space—springs largely from the way we
ferent. They look at some creation by an artist (or are embedded in that network—which is to say, from
computer), and then they abstract from it some the identities (shapes) of the people we are closest to.
quality that they observe in the creation itself (not This means that we help to define others’ identities
in some algorithm behind it). This newly abstracted even as they help to define our own. And very simply
quality may never have been thought of explicitly by but effectively, this parquet deformation conveys all
the artist (or programmer or computer), yet it is there that, and more, to me.

Fig. 15: “I at the Center”, by David Oleson.
Note by Huff: “Original.” Basic Design
Studio of William S. Huff, Spring 1964,
Carnegie Mellon University (CIT).
© HfG-Archiv/Museum Ulm, HfG-Ar,
BDSA, Hu P 06. 011.

Douglas Hofstadter has taught for four decades at Indiana University in Bloomington,
where he holds the title of Distinguished Professor in the College of Arts and Sciences. His
undergraduate degree (Stanford, 1965) was in mathematics, and his doctorate (Oregon, 1975)
in theoretical physics. At IU, he teaches cognitive science and comparative literature.
Hofstadter is best known for his books on minds, thinking, and consciousness. They include
Gödel, Escher, Bach: an Eternal Golden Braid (1980 Pulitzer Prize for General Nonfiction);
The Mind’s I (coauthored in 1981 with philosopher Daniel Dennett); Fluid Concepts and
Creative Analogies (written in 1995 with members of his cognitive-science research group);
I Am a Strange Loop (a personal view of the human condition, published in 2007); and
Surfaces and Essences: Analogy as the Fuel and Fire of Thinking (coauthored in 2013 with
French cognitive psychologist Emmanuel Sander). In 1981–1983 he wrote the monthly
column “Metamagical Themas” for Scientific American, and his collected columns were
published as a book with that title in 1985. Starting in his early teens, Hofstadter plunged
into the study of languages other than English, and this led to a long-term interest in
literary translation. He has translated books and poems from many languages into English,
including, in 1999, a verse translation of Alexander Pushkin’s novel-in-verse Eugene Onegin,
and he has written two books about translation: Le Ton beau de Marot: In Praise of the
Music of Language (published in 1997), and Translator, Trader, 2009. Currently he is working
on a memoir titled My Wild Grace Chase.

Past and Future of William S. Huff's
Parquet Deformations
Fig. 1: “Circlewaves”, 2016,

design by Luisa Paumann
and Werner Van Hoeydonck.

William S. Huff, American architect and inventor of par- Look at what Hofstadter labels “Rule 2”: At each stage,
quet deformations, offered the following comment on the pattern must constitute a regular tessellation of the
our design, “Circlewaves”: plane. I think we’re meant to interpret that as saying that
the pattern must be decomposed into copies of a single
NNNNOOOOOOOOOOOO, this is not a parquet deforma- shape, whereas you have two different shapes—a distorted
tion. Can we call it a “parquet deformation interrupted by “I” and a circle. On the other hand, I have never adhered too
deforming circles”? You can design anything you want. Just strictly to those rules myself. I think there’s a wide range
don’t call it a parquet deformation. A parquet is a repeat- of aesthetic opportunities in patterns that evolve in space,
ing figure that fills space. A parquet deformation takes and see no need to limit myself if a new opportunity arises.
these repeating (space-filling) figures and slowly deforms So, I encourage you to continue with your experiments.2
them, according to a limited set of rules. Your design has
introduced into one set of repeating, but deforming figures, My fascination with parquet deformations started in
a new figure, a circle—that itself is repeating but deform 2016. While I was showing my geometrical patterns at
ing (though deforming, in this case, only in size, but not an international fair for visual communication, Farhad
shape). Incidentally, the beginning of your design is NOT Kay, an entrepreneur from Berlin, told me about his car-
even a “parquet deformation” either, because you have two pentry company, Ligas, and his plans. He had ordered
different sets of pieces: a set of right-handed pieces and a five-axis CNC milling machine to produce wall and
a set of left-handed pieces. And, incidentally, the end of ceiling panels with patterns. These would not be just
your design has three different pieces, because the original any patterns, however; he wanted to produce patterns
piece becomes two different pieces at the end. William1 that do not repeat. This was an interesting challenge
that made me search the internet using keywords
Craig Kaplan, the computer scientist from Canada who such as “transforming patterns” and “metamorphoses”.
has studied parquet deformations and Escher’s “Meta- Previously, I had been designing classical, symmetrical,
morphoses” for over 20 years, had already warned me repeating geometrical patterns. Within a short period
about Huff’s “strict” rules: of time I came across David Bailey’s website, which
is focused on Escher patterns, but has an interesting
Your “Circlewave” design is lovely. For the record, Huff section about parquet deformations and an extensive
would not consider it a genuine parquet deformation. bibliography, as well as references.3
Fig. 2: Marcela Quijano, curator at the HfG archive

in Ulm, August 2017; in front of her, “Axonometry
Cubed” by Thomas Breen. SUNY at Buffalo. Top
left and right: the boxes with the trisections of
the cube. Bottom left: three congruent pieces of a
cube’s trisection. Photos: Werner Van Hoeydonck.
40 Past and Future of William S. Huff's Parquet Deformations Werner Van Hoeydonck
Fig. 3: Maurizio Sabini, Venetian Net,
19 × 27.88 in, India ink. Basic Design
Studio of William S. Huff, Spring 1982,
State University of New York at Buffalo
(SUNY). © HfG-Archiv/Museum Ulm,
Trip to Ulm
The Ulm archive, as I soon learned, was the key archive colored with acrylics, inks, airbrush, or watercolors.
to visit. William S. Huff arrived at the Ulm School of Sometimes, one can see the underlying grids in sharp
Design in October 1956 on a Fulbright scholarship and pencil. The parquet deformations are rather impressive
completed basic teaching under Tomás Maldonado. in size; some are even up to 100 cm in length. They are
Beginning in 1960, he developed a basic design course composed of short line segments, and no sequence in
at the Carnegie Institute of Technology in Pittsburgh, at least one direction is the same, which implies that
based on Tomás Maldonado’s teaching model. He de- total concentration was required to draw them. Regard
veloped basic design courses at numerous universities, ing the graphical work, Huff preferred the classical ink
including the HfG, to which he returned repeatedly as a pen, as, in his opinion, no fineliner, rapidograph, or
guest lecturer from 1963 to 1968. Huff donated the best other kind of technical pen could start and end a line
results of his student’s works to the HfG archive and in the same sharp way or compete with it in precision.7
thus made them available for research.4 Maurizio Sabini, a former student of Huff’s, shared with
Marcela Quijano, who in 2017 was still curator at the me the following anecdote regarding Huff’s method of
HfG archive and was responsible for the Huff donation, encouraging his students to reach such levels of per-
showed me the collection of drawings and models fection: “Huff sat down at the drafting board, the stu-
made by Huff’s students between 1961 and 1998. The dents all around him, adjusted the ruler, took the pen,
collection contains the best results from all his student filled it carefully with ink, adjusted the line thickness.
assignments, resulting in nine portfolios (each of which Then he drew a long line from the left to the middle,
includes approximately 50 drawings) and 20 wooden abruptly stopped the line, then a second line from the
boxes containing models of different sizes and materi- right to the middle and stood up; the connection in the
als.5 Boxes 1 and 2 contain the “trisections of the cube”. middle was perfect. We were baffled …”8
As I discovered later, the “trisections of the cube” were Certainly, drawing by hand in a precise way was
Huff’s attempt to experiment with parquet deforma- necessary for all architects in pre-computer times. In
tions in three dimensions.6 hand drawing, there are different levels of perfection;
The portfolios consist of large sheets of archival, if the first draft in pencil is executed well, one can fully
best-quality drawing paper or fine-line illustration card- concentrate on the final draft in ink. Regardless, the
board with semitransparent sheets between them to level of perfection achieved by Huff’s students is amaz-
protect them. The precise, intricate final drawings are ing and impressive. Although these drawings are not so
traced with ruler, pen, and Indian ink or composed of “perfect” as those that can be plotted now, the per-
black and white only, some of them with only shades sonal touch—the small “errors”—make one aware that
of water-colored gray tones. Many works are brightly these are hand drawings in which every single line had

to be drawn. No “copy and paste” perfectionism was dated 1979, in which Huff defines the assignment in a
possible here. William Huff must have “drilled” his very concise way.11 Divided into three parts—the task,
students to get this right, as one can see in the Ulm the principle, and the pedagogic goal—the A4 docu-
portfolios. It is likely that not all students were able ment offers some clues and keywords regarding the
to achieve Huff’s high standards, and, obviously, only way that Huff wanted students to be introduced to this
the results judged to be the best were archived by assignment. The task is “to fashion a continuous series
Huff. Many of these student works are worthy of being of parquet events into a temporal composition of flow
displayed permanently in a museum. The Museum of ing (rather than static) rhythms; to execute the design
Modern Art in New York acquired 45 works by Huff’s with ruling pen and ink.” The analogy to music in the
students in the early 1970s; unfortunately, they seem to chosen terminology—“temporal composition of flowing
have been lost. While this is unusual for such a pres- rhythms”—is not a coincidence:
tigious institution, apparently the MOMA is no longer
able to locate them.9 One hopes that these works will When I was young, my mother made me take piano
resurface at some point. Fortunately, these missing lessons. I wanted to make sculpture out of soap and to
works, which include three parquet deformations, are make colored things. This experience made me hate mu-
well documented in the HfG archive. For the present sic—any music at all. It took me a long time to appreciate
book, Dr. Martin Mäntele, the current director of the it, but it still was something that I did not understand.
Ulm archive, sent me all nine of the portfolios digitally. Working with these parquet deformations had me learning
Seen in its totality, this collection is a treasure trove—for (or thinking that I was learning) more about the composi-
future research and publications as well. Each draw tion of music that I had ever dreamed that I could do.12
ing is documented by two high-resolution images.
The front shows the actual student work, along with a In the principle, Huff defines a “pure” parquet as an
10-centimeter scale; on the back, the lower right corner endless configuration of congruent (same) pieces (tiles)
presents Huff’s meticulous documentation of the work. that pack the plane without overlapping or leaving
To get a sense of the atmosphere and knowledge holes. The underlying structure of a parquet is a lattice.
offered to the students of the Ulm School of Design, I Huff refers here to the five planar lattices, named for
visited the school—which was designed by Max Bill the famous French physicist and crystallographer
—and studied the library. Among books related to all August Bravais. He also points out that three of these
fields of design, art, and architecture, a broad range of lattices—the oblique (parallelogram, rhombic) lattice
sciences are present, along with seminal works of the and the two rectangular lattices—have certain limita
time, which is typical of the school’s interdisciplinary tions concerning subdivision into equal parts and are
approach to design. The books on patterns, symmetry, limited in their rotational possibilities: A parallelogram
topology, and geometry by the specialists in the field at allows only a twofold rotational subdivision. An exami
that time are important to understanding the roots of nation of the approximately 100 parquet deformations
Huff’s parquet deformation assignment and his lifelong archived in Ulm leads one to conclude that Huff pre-
study of symmetry in all its varieties.10 ferred the square and the special rhombic lattice (60°,
120°), allowing both equilateral triangular and regular
A Search for Sources hexagonal tessellations. In his 2003 lecture to the
William Huff, having studied architecture at Yale and Sociedad de Estudios Morfológicos de Argentina (SEMA),
specialized in basic design at Ulm, was well equipped Huff explains that the five planar lattices are related to
with practical and theoretical knowledge from famous the three regular grids (triangular, square, hexagonal)
scholars, architects, and artists. His students recall but should not be confused with a grid, since a grid
that he could talk for hours about art, but he also liked defines the lines, while a lattice is a set of points:
to use images and poetic metaphors to instruct and
inspire the students. This was certainly also the case In the case of the parquet deformations, the grid be
for the parquet deformation assignment. Although this comes the lattice; and, therefore, by definition, it is invari
assignment spanned 40 years, documentation of his ant. Thus, the deformations take place on the contours
more elaborate writings on parquet deformations is of the tile that are nailed to two lattice points (or to one
rather rare, and most of these sources were written to- rotor), between which they can be elaborated to extreme
ward the end of his teaching career. I will now present lengths, kinks, or turns or in multi-fold contours that are
and discuss all the relevant sources and quote Huff anchored on centers or axes of rotation.13
extensively to allow the reader to hear his voice.
In 2017, when we assigned our students parquets Huff does not offer any strategies or methods in “Best
and parquet deformations, we had only one written Problems”; he merely states that variant shapes of any
source by Huff: a document called “Best Problems”, one lattice system can be subtly deformed to another
lattice system and that these continuous deforma- manner in which film is seen, poetry read, and music
tions can be deformed along “syngenometric” lines. heard:
Syngenometry is a type of lower symmetry, to put it
simply, in which there is similarity in form but also Music, linearly regulated in respect to time, does have
gradual change. These different levels and types of spatial elements (notes and chords). Even as entities
symmetry were an important research topic through with evident periods, these spatial elements, separated
out Huff’s career.14 by intervals of time, occupy no more than the zero-di-
If one follows the defining syngenometric lines on mensional moment of the present. The handscroll does
the border of a parquet deformation step by step and have a real, two-dimensional presence; one of its planar
from tile to tile—if it is not framed—one will better dimensions is, however, so extremely elongated that it
understand the different “parquet events”. In other induces the palpable temporal component that explicitly
words, one understands how the lines deform and at channels one’s experiencing of it. The young medium of
which speed (tempo), which connects to the topic of film constructed from a flow of captured spatial mo-
temporality. ments (discrete photographs), whose transience tricks
In “The Landscape Handscroll and the Parquet the mind into perceiving continuous movement—may
Deformation”, written for an important intercultural and be broadly viewed as illustrative of the special nature of
interdisciplinary conference in Japan15, Huff explains his temporal visual arts. In a manner arguably comparable
fascination with the factor of time in art: to film, the handscroll is unrolled and rerolled, frame by
frame—though at the full discretion of a solitary viewer.
The different arts are frequently classified as either I bolster this contention with the statement of an expert
spatial or temporal. The spatial arts (painting, sculp on Chinese art, Wen C. Fong: Working in the handscroll
ture, architecture) occupy two-dimensional or three- format, which unfolds one section at a time from right to
dimensional real space. The temporal arts (music, dance, left, the painter employs serial images, his focus moving
poetry, theater) occupy the dimensional space of the in cinematic fashion in the development of his narrative.17
instantaneous present, continuously advancing through
a dimension that unfurls in time—regularly regarded as Seeking comparable examples of temporal plastic
“the fourth dimension of the space-time continuum”. compositions in Western art, Huff suggests some
This classification is, to be sure, an oversimplification; examples: the Parthenon’s frieze, the Column of Trajan,
for there is, of necessity, temporality in all spatial art the Bayeux tapestry, Mantegna’s Camera picta, Monet’s
and spatiality in all temporal art. A two-dimensional Water Lilies series, and Jackson Pollock’s paintings.
easel-painting or wall-hanging reveals its full composi- Since both Western and Eastern art have made art
tion at once; yet it is not fully comprehended in a glance. in which temporality plays an important role, Huff
The eye flits, involuntarily, over the plane—from detail concludes “that the early landscape handscrolls have
to detail, from figure to ground, from homogeneity to an aesthetic structure—within the specific culture
heterogeneity—in a never ending, randomized sequence and age—that accords to the individual artist’s own
that is experienced through time. Absolute comprehen- sense of drama and that any canons, West or East, are
sion is, perhaps, never attained; but the aggregate of the destined to be contested and even renounced by suc-
viewer’s fitful perceptual impulses contributes to an ever ceeding generations, if that art form is to remain vital.”18
fuller comprehension. The third dimension of sculpture Furthermore, Huff found the same aesthetic modalities
extends the viewer’s interaction with this art form. The in Eastern and Western Art and quotes art historian
many aspects of a piece of sculpture are experienced Nelson Wu:
only by the viewer’s moving around it. Mobiles, of course,
themselves move, while the viewer remains stationary. The student will quickly recognize that all pictorial
Architecture is experienced not only by the viewer’s expressions have the same building blocks: line, area,
moving around it, but through it. The moving viewer (or color, space, movement, and all the other privileges and
moving mobile) and the viewer’s moving eye involve an limitations that are, part and parcel, the birthright of a
ingredient of time.16 two-dimensional art. These components in their analyzed
form, simple and pure, are universalities, behaving like
Huff found in the Sino-Japanese handscroll an art musical tones, favoring no particular culture or tradition
form that must be experienced in a different way, and belonging to all.19
section by section, to fully grasp what is happening;
this is a temporal visual composition. A parquet de- The same universalities are to be found in Huff’s
formation—in Huff’s analogy to the oriental scroll— parquet deformation assignment. The lattices, the
is not intended to be viewed spatially, but temporally, grids, the lines, and the possible symmetric opera-
as a sort of visual music. Viewing them is akin to the tions are universal, but the organization depends “on

the dynamics of the student’s chosen motifs—and the study of deformations and transformations con
the magic that can be extracted […] many parquet stituting perhaps the most crucial underlying motif. The
deformations follow the general classical structure of perceptual refers to the normal behavior patterns of our
beginning with simplicity and building to complexity. sensory receptors, i.e. our everyday touching of the world
Yet some begin with complexity and build down. Some and the meanings we construct out of these encoun-
have extensive lulls at their centers. Some have their ters. Analogous to the variant and invariant factors of the
points of departure at their centers. Some have mul physical situation, those things that possess identities (or
tiple climaxes.”20 grouping properties) and those that create contrasts are
The pedagogic goal of the parquet deformation, the basic concerns of perceptual phenomena.24
as stated in Huff’s 1979 “Best Problems”, was to “have
the student become totally familiar not only with the Identity versus contrast, symmetry versus asymmetry,
families of congruent figures (tiles, parquets) that fill order versus disorder—such general terms were not
a plane and their topological relationships, but with subtle enough for Huff to describe all possible levels of
the fundamental principle of continuous deformations symmetry in structures. A more precise vocabulary was
(after Dürer and D’Arcy Thompson) and to have him needed for the subtler levels of symmetrical structure.
design an aesthetically coherent composition that is In his paper “Ordering Disorder after K. L. Wolf”, Huff
essentially temporal in contrast to the spatial compo urges the universal adoption of the classification of
sitions more familiar of the history of our western these subtler symmetries and, in fact, the whole range
visual culture.”21 of structure, as formulated by the German chemist K. L.
Temporality is achieved through subtlety. Subtle, Wolf (1901–1969), an authority on molecular structure:
geometrically interrelated shape changes are achieved
through continuous deformation of one spatially In many of these recently organized interdisciplinary asso-
evolving geometrical basic idea, respecting the topo- ciations, the memberships have moved from a focus on
logical relationships in a certain chosen structural field rigid regularities—passionate preoccupations with crystals
(lattices and grids). In “An Argument for Basic Design”, and their models, the regular and semi-regular polyhe-
written in 1965—while Huff was already teaching at the dral—to lesser regularities. However, I have noticed that,
Ulm School of Design—he writes: “By structure I strictly while many scientists and artists do now speak in shared
mean: the relationship or arrangement of parts or ele- terms about highly regular symmetries, neither employs
ments. To design, then, is first of all, to structure; and a comprehensive, much less unambiguous, terminology
for me the study of structure (in the abstract) is the about entities that seem to be less than symmetrical. We
equal of that which has been known as basic design or use and hear such words as “disorder”, “broken sym
foundation studies.”22 metry”, “irregularity”, “randomness”, and even “chaos”.
Huff also points out that structure is the one term Many things that are called “broken symmetry” and the
that still has value, “whereas terms like unity, harmony like are, in fact, not. Rather they are often lesser symme-
and proportion, rhythm, scale, composition, form, even tries—that is, orders or degrees (in other words, different
truth and virtue are words that once held great mean- levels) of symmetry that are lower than isometry.25
ing, evoking an air of cultured sensibility, are not so
easily used anymore. Lacking firmness if such ‘eternal’ Huff completed Wolf’s classification—which deals with
terms are used, their meanings have also lost their a logical sequence of degrees of regulation—with two
usefulness for contemporary design problems, or at extra levels: autometry and hypometry. The adapted
least are inadequate to describe them.”23 list (see also Cornelie Leopold’s chapter) starts with
The parquet deformation assignment, though, was autometry (maximum symmetry) and moves to isom-
not a real-world design assignment, but an abstract etry, homoeometry, syngenometry, katametry, hypom-
exercise to improve the students’ aesthetic acuities and etry, heterometry, and finally katametry (the lowest
to familiarize them with the topological relationships level of symmetry). Every level is defined by invariants
among parquets, with the transformational possibilities and variants in relation to position, size, angle, shape,
of their symmetries and restrictions of their defining certainty, and rule. The parquet deformation assign-
structures. Huff considered two distinct areas in the ment was viewed by Huff as a project related to the
study of structure—the physical and the perceptual: symmetric level of syngenometry, one that was devel-
oped along syngenometric lines.26 These lines outline
The physical refers to how a structure, as far as we can the basic geometrical idea, as can be seen in Figure 4.
humanly determine, actually is (from the microcosmic Huff frequently quotes Louis Kahn in his paper
atom to the macrocosmic universe). In the physical “Ordering Disorder after K. L. Wolf”, in another context,
manipulation of structure, we are interested in those but Kahn’s words can also be interpreted, in my opin
things that are invariant and those that are variant, with ion, to explain syngenometry:
Fig. 4: Paul Randazzese, "Brocade". Basic Design Studio of William S. Huff, Spring 1989,
State University of New York at Buffalo (SUNY). © HfG-Archiv/Museum Ulm, HfG-Ar, BDSA,
Hu P 05. 023. Note the syngenometric lines.
Order is equivalent to natural law. Not only is the con- members neither too numerous, nor too small, nor too
summately symmetrical crystal an integral part of Order, large, nor too dissonant or ungraceful, nor too disjointed
but the cataclysmically exploding star is no less so. or distant from the rest of the body. Alberti leaves it to
Kahn refers to Order as “the possibility to be”, vis-à-vis aesthetic judgment to sort this out.28
his word Desire, “the will to be”: From the macro- to
the microcosmic scale, the dual components of Order Huff was a subtle man, yet harmony alone was not
and Desire—when, and only when, concordant—beget subtle enough. He assigned his students to explore the
Existence.27 lowest level of symmetry, katametry, in his programmed
design assignments, in which harmony does not fully
Order, the possibility to be, can be considered the disappear, but becomes indiscernible. Huff uses an
invariant grid, whereas Desire, the will to be, stands for other quote from Kahn, an analogy to music, as a tool to
the line in its desire to shift its shape, to become a dif- eliminate perfect harmony and support the acceptance
ferent—yet still recognizably similar—line. A successful of dissonance: “When a dish fell in Mozart’s kitchen, it
parquet deformation then comes into existence; when broke—made a terrible noise. The maid screeched …; but
the invariant grid guides a syngenometric line which Mozart said, ‘Ah, dissonance!’ And dissonance belonged
gradually takes over control, the grid is always there immediately to music.”29
but loses its dominance and can, in fact, disappear.
To describe aesthetic coherence in a general sense, Dissonances can jolt us from a too gentle lull that
Huff refers to the Renaissance architect Alberti in his harmony might produce. But how much dissonance
paper “On Regulation and Hidden Harmony”, offering a can be tolerated? Richard Strauss’s tone poems (to
perfect quote to describe a balanced parquet defor- my taste) strike a stimulatory balance. On the other
mation: hand, Bach and Händel, masters of harmony, make
glorious art with structural intricacy. Works of some
Compartition will be seemly when it is neither jumpy, nor moderns become tedious (to me) because of too
confused, nor disorganized, nor disconnected, nor com- much dissonance or too much intricacy. Tedium can
posed of incongruous elements; it should be made up of be created by bewilderment; one’s temper can dismiss

sensory onslaughts. Alberti’s instruction, anticipatory In the same “Handscroll” paper, Huff reveals how he
of Gestalt, did not say “no discord”, but “not too much became acquainted with parquets, discusses the
discord”. Thus, dissonance enters, and we exercise our underlying issues and mathematical backgrounds in
aesthetic judgment—as we do regarding any piece of full detail, and explains the main methods used to
art; for harmony in itself does not a work of art make. design a parquet deformation. Since his paper is highly
Gestalt does not formulate aesthetic judgment; it does instructive, accurate, and unique, I will quote Huff
expedite the articulation of situations (of mind and extensively without interruption:
of utterance) on which aesthetic judgment is based.
Gestalt appreciates harmony; Gestalt appreciates Among the variety of teaching assignments developed
dissonance.30 in my design studio, the parquet deformation offers, as I
have suggested, an uncommon experience in the realm
The parquet deformations that were selected by Huff of aesthetics—the challenge to deal with a visual art
and conserved in Ulm are all aesthetically coherent that derives its essential vigor from the factor of time,
designs, according to Huff’s subtle sense for aesthet- but that is fitted, nonetheless, with a full complement of
ics. Huff obviously liked Hofstadter’s description of his planar tangibles. Underlying the students’ task are rigor-
way of teaching as “a kind of performance”, since he ous two-dimensional mathematical determinants that
used a quote by Hofstadter on his work in the “Hand- involve particulars of symmetry and topology and that
scroll Paper” in 1996: must be obeyed in devising operable design strategies.
In the mid-1950s, the topic of the parquet pattern was
“Huff himself has never executed a single parquet defor- introduced by Tomás Maldonado into the design curricu-
mation. He has elicited hundreds of them, however, from lum of the Foundation Course at Germany’s Hochschule
his students, and in so doing he has brought this form of für Gestaltung. Finding little guidance in a search of
art to a high degree of refinement. He might be likened contemporaneous texts for an English equivalent, I angli-
to the conductor of a fine orchestra. Although the con- cized the term Parkettierung, which appeared in German
ductor makes no sound in the course of a performance, texts. If Martin Gardner’s column, “Mathematical Games”
we give much credit to the person doing the job for the for Scientific American, is a reliable indicator of the evo-
quality of the sound. We can only guess at how much lution of the nomenclature of this inglenook of geometry,
preparation and coaching went into the performance. it was not until the mid-1970s that the words tile and
And what about the selection of the pieces and tempos tiling took dominance over other expressions. Regular
and styles—not to mention the many-year process of tessellation was the preferred term of the 1960s; and
culling the performances themselves?”31 articles of the time routinely cited “floor and wall tiles” as
Fig. 5: Moritz Kreitz, “One-handed Two-handed

Dual”. Basic Design Studio of William S. Huff, May
1966, Carnegie Mellon University. © HfG-Archiv/
Museum Ulm, HfG-Ar, BDSA, Hu P 01. 024. The
version below is one of the few examples with A
and A‘ tiles, also called enantiomorphic tiles (tiles
of both-handedness).
merely some among a number of examples of this sort pieces, which, precisely, determine a pattern’s reticular
of planar configuration. A parquet pattern (or monohedral interplay. Two general mechanisms prevail: (1) the defor-
tiling—Grünbaum) is defined as a space-filling array of mation of the pieces’ boundaries between lattice points,
congruent pieces—precluding, on that account, any gaps involving an equal give-and-take of areas that have been
between pieces or any overlapping of them. A puristic dislocated by those altered boundaries; (2) the subdivi-
approach is to admit only pieces that are superposable. sion of pieces into yet smaller congruent pieces by the
Congruence does not, however, stipulate handedness; corresponding multiple (according to 2-, 3-, 4-, or 6-fold
and there are, indeed, patterns whose pieces conform to rotational symmetry) of alike, but otherwise capriciously
the three strictures of congruence, no gaps, and no over- shaped contours, anchored only at the original piec-
laps—but are also enantiomorphic (or nonsuperposably es’ axes of rotation. By virtue of their repetitive, even
left- and right-handed). Might these be called improper spread over the plane, the periodic parquet patterns are
parquet patterns or tilings?32 of a sort that most readily accommodates continuous
The age-old one-by-two brick can be arranged relatively deformation, the incremental molding of one piece into
randomly (i.e., nonperiodically) to cover the plane. Other a differently shaped piece; and the two stipulated design
parquet pieces can fill the plane in various irregular or mechanisms that play a crucial part in the fashioning of
semiregular ways. The remarkable “rep-tiles” build into novel parquet pieces are the selfsame mechanisms that
expanding replicas of themselves, and some pieces can facilitate the fashioning of a parquet deformation. It is
be set into circular or spiral arrangements. However, worth noticing that the wallpaper group of the pattern
the parquet patterns most encountered, owing to their can change as a deformation is evolved. For example,
frequent application throughout the history of ornament, square pieces with fourfold rotors and four mirror axes
are periodic; and periodic patterns come under the can be continuously deformed into pieces with only
regime of the 17 wallpaper groups, which, in turn, submit twofold rotors and two mirror axes (e.g., double headed
to the five planar Bravais lattices. It is well-known that axes) and deformed again into pieces with only twofold
there are only three regular parquet patterns: regular rotors (e.g., S- or Z-shapes).33
arrays of equilateral triangles, of squares, and of regular
hexagons. It is not as well-known that there is an infinity Huff concludes his paper by referring to Thompson
of other-shaped pieces that can fill the plane mono- and makes interesting comparisons with the artist
hedrally. The shapes of such pieces and their periodic M. C. Escher:
arrangements conform to the same geometric con
straints that govern crystals; and understanding those The intriguing possibility of the incremental deformability
constraints expedites the designing of odd and intricate of one parquet pattern into another came to our attention
Fig. 6: Parkettierungsaufgabe 1956/57,

Hochschule für Gestaltung Ulm; taught
by Tomás Maldonado, student: William S.
Huff. Three colors proceed successively
toward neutralization: a secondary of the
first order (orange); a tertiary of the first
order (yellow–gray); a tertiary of the sec-
ond order (green–gray). To see this work in
color: Ulmer Modelle–Modelle nach Ulm,
Ostfildern-Ruit, Hatje Cantz, 2003, p. 179.

in 1960 when it was recognized in one student’s designs a significant difference between the two concoctions is
of several very different looking patterns that there were evident in the aesthetics of their pieces’ respective modes
underlying, but far from obvious morphological relation- of rendition. The lure of Escher’s designs lies within the
ships between them. The instigation of our first continu- contours of his pieces—that is, how they are graphically
ous deformation between such patterns was buttressed treated, be that as fishes, birds, lizards, or horsemen.
by a writing of a morphologist and two designs of a Erase the rendered surfaces of his pieces—the leftover
driven graphic illustrator: D’Arcy Wentworth Thompson’s contours emerge insipid. Corollarily, Escher’s figurative
seminal chapter “On the Theory of Transformations, or the imaging of his pieces inhibits the entrancing potential
Comparison of Related Forms” in On Growth and Form; of unconstrained contours to create their own rhythmic
M. C. Escher’s two beguiling woodcuts “Day and Night” and sensations. In the case of our deformations, the contours
“Metamorphosis”. In the latter, Escher, taking an artist’s are the object; they make the music. Fill between these
license with geometry, cunningly shifted his lengthy rib- contours with colors, textures, or figurative subjects—the
bon of a design from a square grid to a regular hexagonal music crashes.
grid and back—two times in violation, of course, of the
principle of the invariance of the Bravais lattice type, once Pure Parquets, Indian Baskets,
one has been established in a design with periodicity. and Mathematical Pastimes
Resemblances between Escher’s designs and the parquet Tomás Maldonado’s Parkettierungsaufgabe in the basic
deformations oblige that attention be given to their sub- design course of 1956 in Ulm fascinated Huff as a stu-
stantial dissimilarities. Both Escher’s and our bodies of dent. Huff mentions that he was probably the only one
work, involving periodic patterns, conform, of course, to in his class who actually executed this assignment.34
the same geometry. Our design studio deals all but exclu- In 1960, when William Huff started teaching basic
sively with superposable pieces (A pieces only)—though design at Carnegie Mellon University, he assigned his
occasionally with enantiomorphic ones (A and A‘ pieces). students the task of developing three variations (see
Patterns of both sorts are found in Escher’s designs. On Figure 7) of a pure parquet. The pure parquet studies
the other hand, Escher is quite apt to employ two differ were important to Huff. They were archived just as
ent interlocking pieces (A and B pieces—fishes and birds). carefully and in the same way as the parquet defor-
These designs are clever, but not as difficult to achieve mations and are the forerunners that ultimately led to
as might be assumed: the two different pieces, taken them. That is why he included dozens of them in his
together, merge into one proper parquet piece; the divid gift to the HfG-archive in Ulm:
ing contours between the two sub pieces are under no
mathematical constraints and are, therefore, completely When I set up a basic design curriculum for the Carnegie
pliant to the command of the artist. In this way, designs Institute of Technology, first implemented in 1960,
that have the geometry of the simpler wallpaper groups parquetry or tiling was among my curriculum’s various
can appear to be far more complex than they actually topics. Before I assigned this topic to my class, however, I
are. While an Escher design and a parquet deformation made a study of how a variety of different parquet shapes
both exploit the trickery of interlocking congruent pieces, could be fashioned. In fact, my resolve to continue an
Fig. 7: “Parquetry Study in Three

Variants” by James D. Richardson,
Spring 1961, Carnegie Mellon University.
BDSA, Hu P 09. 030. Note the borders
between the parquets: The connection
is already in the spirit of a parquet
deformation.
Fig. 8: Museum of Indian Baskets, designed by William S. Huff (1966–1972). Collage by Werner
Van Hoeydonck. Ground floor plan by William S. Huff; front view photo by Harvey Kaplan;
interior photos by Clyde Hare. Source: Maurice L. Zigmond, Gotlieb Adam Steiner and the G.
A. Steiner Museum, Journal of California and Great Basin Anthropology, 1979, Vol. 1, No. 2, pp.
322–330. Courtesy of Malki Museum. Top right, upper basket: “Beacon Lights” by Datsolalee,
1904–1905. The basket collection now resides in the New York State Historical Association’s
collections at Fenimore Art Museum in Cooperstown, New York. The private museum building
was demolished.
exploration of parquetry had been stimulated/solidified baskets and patterned ceramics. One of Huff’s first
by a fascinating small book, which I had picked up for ommissions as an architect was to design a private
c
pennies in Philadelphia a year or two earlier, Major Mac- museum for his mother, Elsa Steiner Huff—who con
Mahon’s New Mathematical Pastimes.35 tinued her father’s hobby of collecting Indian bas-
kets—where this collection could be displayed. The
Before looking deeper into MacMahon’s book, one must floor plan of the museum is very structured, has a
ask if there could have been any other early influences twofold rotation, and looks like a pattern in itself. The
that made Huff more suited to Maldonado’s parquet museum displayed 555 baskets from at least 62 Native
assignment than the other students in Ulm. When I was American tribes. William Steiner Huff later became
studying Huff’s inventory of his archive, the William S. responsible for this collection and made inventories of
Huff Papers, at the University at Buffalo, State University each piece, including all the details known to him from
of New York, a particular architectural project fascinated the records, with the same perfectionism with which
me: Huff’s Museum of Indian Baskets.36 he made inventories of his students’ works. A master
William S. Huff’s grandfather Gottlieb Adam piece in the Steiner Huff Collection is a work from Dat
Steiner (1844–1916) was a collector of Indian woven So La Lee, the Washo name of Mrs. Louisa Keyser of

the Washo tribe, located around Lake Tahoe in Nevada. weaving techniques, adorned with motifs and geo-
She was said to be one of the most famous basket metrical shapes, can be seen in other ancient cultures
weavers of Native American art. Her Washo tribe’s roots on every continent. These basic forms—zig-zag lines,
may be traced back 9,000 years. Dat So La Lee made squares, triangles, rhombic forms (two triangles), and
this particular basket in 1904–1905; she worked on it hexagons—are indeed universal; they are the elemental
for 14 months, making 29 stitches per inch, or 80,000 building blocks in Huff’s parquet deformation assign-
stitches in all, and named it Beacon Lights. Its flame ment. “Leather of the Lesser Gator” is an excellent ex-
design is said to commemorate the large signal fires ample of this unique and subtle shapeshifting between
built on hills and mountains whenever it was neces- the fundamental shapes of the square, the triangle,
sary to call her tribe, the Washo people, together.37 The and the rhombus. Never before in the long history of
titling of a work with a memorable name reminds me geometrical patterns had there been a comparable
of the poetic names that Huff gave to every parquet attempt to subtly deform one elemental geometrical
deformation. I assume there is a connection between form into another, thereby showing fundamental rela-
Huff’s having been reared with an awareness of this tionships between basic forms. The history of geomet
basket collection and his interest in patterns. The rical patterns is based mainly on repetitions of the
Carnegie Museum in Pittsburgh displayed a portion same forms or motifs. Huff and his students opened a
of Steiner’s basket collection between 1913 and 1937. completely new chapter of design possibilities based
Young William was ten years old in 1937 when the col- on the structural interrelationships and shapeshifting
lection was placed in storage in his mother’s house. between basic forms.
In his final email to me, written shortly before he The possibility of shapeshifting, in the sense of
passed away, William Huff forwarded a link with infor- a “give-and-take process” on the borders of tiles,
mation on his grandfather, this special collection, and is clearly present in the aforementioned book New
the building he designed, writing that this building and M
athematical Pastimes by Major Percy Alexander
the collection were a source of pride to him. This is im- MacMahon.38 This book inspired Huff to further
portant, as it links the pure parquets and the parquet explore the topic of patterns in his Basic Design
deformations in relation to the long history of patterns. Studio. In Part I, MacMahon introduces geometri-
Similar basket patterns, derived from the structures of cal edge-matching puzzles, which are similar to the
Fig. 9: Thomas C. Davies, “Leather of

the Lesser Gator”. Basic Design Studio
of William S. Huff; 19 × 19 in, India ink.
Spring 1969, Carnegie Mellon Universi-
ty. © HfG-Archiv/Museum Ulm, HfG-Ar,
BDSA, Hu P 01. 007.
Fig. 10: “New Mathematical Pastimes” by Percy Alexander Mac Mahon; copyright Paul Garcia
and Tarquin Reprints, 2004; reprint of the original 1930 edition, pp. 56–58. On the left, the
transformation diagram with numbers for the System C1, 1, 2, or 1 to 1, 2 to 3, and 3 to 4.
Bottom left: the four chosen new boundaries. The piece each of whose compartments is
colored 4 vanishes, giving the 23 pieces (middle). On the right: the assemblage of the new
pieces according to the original diagram on the left. Redrawn by the author with kind per-
mission of Andrew Griffin, publisher of the Tarquingroup.
Fig. 11. Peter Hotz, The Original

(with frame). Basic Design Studio of
William S. Huff, Spring 1961, Carn-
egie Institute of Technology. © HfG
Archiv/ Museum Ulm, HfG-Ar, BDSA,
Hu P 01. 011. There are two other
unframed versions in Ulm.
well-known domino games, but he explores them in ompletely new set of tiles appears, in which none
c
an original way. Depending on how many numbers or of the pieces is the same. The pastime seeker then
colors are involved, he sets up different square and has to puzzle the pieces together until the original
hexagonal diagrams, challenging the pastime seeker diagram has been restored.
to arrange the pieces according to pre-established In Part III of the book, MacMahon examines what
connecting rules, such as “one has to connect to can be done with pieces that are similar in shape and
one and two to two”, or, in the colored variation, “red size to make designs of repeating patterns for deco-
has to connect with red, etc.” In the second part of rative work, based on triangular, square, pentagonal,
the book, “The Transformation of Part I”, he uses the and hexagonal pieces. Examining the drawings of
same square, triangular, and hexagonal diagrams, but Parts II and III, one can easily see correspondences to
this time the connections must be made according Huff’s students’ works. Even the so-called Cairo tiling
to a give-and-take formfitting, without numbers or is displayed, a tiling which is also present in the very
colors. The borderlines are deformed according to first parquet deformation by Peter Hotz.
a limited set of symmetrical straight, V-, or Z-lines, In his 2003 SEMA lecture notes, Huff recalls how
but duos of asymmetrical lines also occur, so that a this happened:

Peter Hotz, a gifted student, made the observation, when In On Growth and Form, D’Arcy Thompson uses the
I visited his drafting board (then, I did some board-to- terms transformation and deformation interchangeably.
board criticism in the studio during the development Both terms are used about 20 times each in Chapter
phase), that incremental changes could be affected from 7, “The Theory of Transformations, or the Comparison
one parquet variation to another. In fact, he had sketched of Related Forms”.41 Transformation definitely has a
not three variants of the original parquet, but five or six more positive connotation than deformation, which
[…] At the time that Peter put together his design, the also means distortion or misshaping. Huff probably
first parquet deformation, I considered it to be a one- preferred the term deformation because it better
time variant of my assignment. But after reading D’Arcy describes the active process of shapeshifting, with dif-
Thompson’s chapter/article [on] “continuous deformation,” ferent steps in a continuous permutation, as opposed
I had second thoughts and assigned to my whole class, in to transformation, describing more the end result of
the third year of my teaching, the assignment to produce the process.
parquet deformations. The second part of the term under The distortion of grid systems was not D’Arcy
discussion, deformation, came as already mentioned Thompson’s invention, but an artist’s:
from D’Arcy Thompson—the parquet deformation being,
of course, a continuous deformation. […] The recognized D’Arcy Thompson credited his use of continuous defor-
potential of innumerable variants of any given parquet (an mation as an analytical tool for classifying species ac-
array of congruent tiling) led to connecting variant tilings cording to a design tool devised by Dürer. Albrecht Dürer
to one another serially through continuous deformation.39 gridded the face, usually in profile, of the “ideal” man.
Then, he distorted the grid: the grid was elongated along
Huff’s fair acknowledgment that it was a student who its vertical, elongated horizontally, turned into a trape
invented the first parquet deformation—he always zoid in which the shorter cord was at the top, and into a
credited his students for their creations—is a good trapezoid in which the shorter cord was at the bottom.
example of his and Maldonado’s pedagogical method, Each resulting profile was a recognizable facial type. His
creating an assignment as a challenge for the students deformations allowed the painter facial varieties. Dürer
to experiment with, often leading to surprising out also drew his versions of the Vitruvian man and Vitruvian
comes. Huff mentions this: “The good pedagogue is on woman—both “ideal” Teutonic body types. He then
constant alert, knows when he sees it, and is ready to measured off critical measurements of the body along
promulgate something extraordinary that occurs under the vertical axis: chin, shoulders, breast, navel, crotch,
his instruction.”40 knee, ankle. Then he elongated the vertical measurements
Fig. 12: D’Arcy Wentworth Thompson, On Growth and Form, Chapter 7 “The Theory of Trans-
formations, or the Comparison of Related Forms”, 2nd ed., Vol. 2, 1959, Cambridge University
Press, figures newly arranged from pp. 1026–1095. Three faces top left: Thompson was in-
spired by Dürer’s grid-stretching technique; changes between the parts of the face result in
totally different characters. Two faces bottom left: Dürer changes the angle of the coordinate
system, another technique resulting in different characters. Two drawings right: Thompson,
inspired by Dürer, uses different transformations of coordinate systems to compare related
species, in this case a human skull in a regular Cartesian coordinate system transformed
into a chimpanzee skull. Collage by the author. Courtesy of Cambridge University Press.
proportionally, while preserving the horizontal measures. deformations more spatial and served as an import-
Presto! A Watussi’s body type. He compressed the ver- ant step toward our experiments with parquet defor-
tical proportionally. Presto, an Eskimo body type. D’Arcy mations in 3D. The students were allowed to deform
Thompson adopted Dürer’s device to explain morpholog- spatial grids in the 3D assignment as well. In some
ical differences among related species. He laid a grid on cases, this resulted in gracefully contracting or expand-
the profile of a human skull; then, he marked off similar ing structures, as the reader will see in the second part
features found in the chimpanzee skull and showed how of this book.
the grid had gracefully/continuously deformed.42
Huff’s “Classroom Tutorial”
In Huff’s parquet deformations, the grid itself was Nicholas Bruscia, assistant professor in the Depart-
not to be deformed or stretched. In 2017, grid defor- ment of Architecture at the University of Buffalo, the
mations were an extra experiment proposed to our same university where William Huff taught, works very
students at the Technical University Vienna. Thompson much in the spirit of Huff, experimenting in 2D and 3D,
and Dürer’s grid-stretching technique was integrated and makes original and contemporary assignments
into the assignment as an opportunity to make parquet challenging his students in a similar way to Huff’s. He
Fig. 13: M.C. Escher’s Liberation, April

1955, Lithograph, 434 × 199 mm.
© 2021 The M.C. Escher Company–
The Netherlands. All rights reserved by
www.mcescher.com.

sent me a 77-page document that is not dated but Blossoms”); seven examples developed planarly (in
must have been compiled in this form around 1983. more than one direction) upon the square lattice (“I at
The document looks like a portfolio or a classroom the Center”); two examples developed planarly upon
tutorial and is titled “The Parquet Deformations from the special (60°–120°) rhombic lattice. Furthermore,
the Basic Design Studio of William S. Huff at Carnegie the tutorial contains “two clumsy examples (by good
Mellon University, Hochschule für Gestaltung and students) demonstrating that not all deformation
State University of New-York at Buffalo from 1960 to attempts work out aesthetically; an attempt to deform
1983”.43 ‘rep-tile’ groupings (two examples); dual developments:
The tutorial contains eleven examples of pure par- (a) all same-handed congruent parquets (b) interlocking
quets—seven of them in black and white—followed handed and other handed parquets.” See also Figure
by 39 parquet deformations, which are all currently 5. At the end of the syllabus Huff shows a “parquet
archived in Ulm. This “sample map, manual, tutorial, deformation developed lineally in conventional versus
syllabus or presentation map” is the first collection boustrophedon modes of transcription”, followed by “a
of parquet deformations and served as the basis for temporal scroll of Escher (probably to become includ-
Hofstadter’s Scientific American article in 1983, bring ed in text portion)”.45 See Figure 13.
ing parquet deformations a broad scope of attention In Huff’s tutorial, there are only four pages with
in America’s most popular science magazine. Douglas short explanations. The first parquet deformation by
Hofstadter explained to me how he got to know Huff: student Hotz is described in the following way:
What happened was that Bill Huff wrote me a letter on The first parquet deformation, executed in 1961, consists
February 14th, 1982, telling me all about parquet defor- of a virtual catalog of familiar repeating tiles linked con-
mations and sending me a few of them as samples. I tinuously together: There is (a) the basic grid of squares,
replied with considerable interest, and Bill and I started (b) the semi-regular, equal sided, overlapping hexagons,
corresponding, and I became fascinated with parquet (c) the cross, (d) the brick, basket weave, (e) the pinwheel,
deformations. Eventually, in February of 1983, I visited Bill (f) the modified pinwheel with the accentuated swastika
in Buffalo for a couple of days, and during that visit I met crossing, and (g) return to the square grid. The transitions
some of his colleagues and his students (and I think I of this original development are not always meted as
gave a talk there, but that’s a bit blurry in my memory) harmoniously as are effected in its progeny of the studies
—and then, of course, as a result of my visit, I had tons of that followed. [See Figure 14.]46
new material, and my column in Scientific American was
published in July of 1983. I hope this helps clarify how the The syllabus leading to Hofstadter’s article mainly
connection was established. Yours, Doug.44 shows examples, text is scarce, and he describes only
“Crossover”, “From Five to Four and Two Halves”, “Dizzy
The parquets and parquet deformations are classified Bee”, and “Consternation”, in only a few sentences with
in groups according to their lattice structures. In paren- diagrams, as shown in Figure 14. The SEMA lecture of
theses I added the names of the works displayed and 200347 is a better source for additional information on
described in Hofstadter’s article: Strategies. Huff does not cover every possible situation,
Twelve examples developed lineally on the square merely the basic ones:
lattice (“Fylfot Flipflop”, “Crossover”, “Crazy Cogs”, “Razor
Blades”, “Oddity Out of Old Oriental Ornament”, and Strategies on the square lattice:
“Clearing the Thicket”); five examples developed lineally 1. opposite sides: paired with mirror concave and
across the diagonal of the square lattice (“Cucaracha”); with mirror convex.
13 examples developed lineally upon the special 2. adjacent sides: paired with mirror concave and
(60°–120°) rhombic lattice (“Dizzy Bee”, “Consternation”, with mirror convex.
“Trifoliolate”, “Arabesque”, “Y Knot”, and “Beecombing 3. all four sides; twofold S or Z-shape.
Fig. 14: Huff’s analysis of the

first parquet deformation by
Peter Hotz, 1961.
Fig. 15: Basic strategies on the square (1, 2, 3, 4 top left),
parallelogram (1 and 2 bottom left), special rhombic
(60°–120°) lattice (A1, A2), triangle (B1) and hexagon (C1, C2).
Note that the arrangement of elements serves only to illus-
trate the tiling. To turn them into a parquet deformation,
all lines must evolve in space. Drawing by the author.
4. What happens when the convex and concave This brings us to the second major strategy of parquetry
lines of Points 1 and 2 do not have mirror property? —one that subdivides certain parquetry designs. Those
In some cases, congruent, but opposite-hand- tiles that have 2-fold, 3-fold, 4-fold, or 6-fold properties
ed tiles are created. (Occasionally, I allowed these can be subdivided by 2-fold S or Z curved lines or lines
situations.) that pass or passes through the center of the tile. These
lines can be rotated incrementally or continuously within
Strategies on the parallelogram lattice: the cells as part of the whole strategy of deformation.48
1. opposite sides: any line repeated along the x-axis;
any line repeated along the y-axis. This reproduces Huff describes basic strategies, processes that can also
the most basic of the 17 wallpaper groups (two-di- be understood by looking, examining, or retracing the
mensional crystallographic groups). parquet deformations themselves, since there is much
2. opposite sides: twofold S or Z-shape. In this case, more happening in the more complex ones. To cover
tiles can be oriented in two directions, where in all the strategies in the approximately 100 parquet
Case 1 they are oriented in only one direction. deformations archived in Ulm, a catalogue of strategies
must be made ranging from conventional, basic strate
Strategies on the special rhombic lattice: gies to the more intricate “hybrids”, in which different
A. On the lattice itself operations are combined. Such a catalogue should also
1. opposite sides: paired with two different kinds give information regarding the syngenometric lines, the
of 2-fold S or Z curves rhythm (temporality/subtleness), vertical and horizontal
2. all sides: same 2-fold S or Z curves axes, symmetry groups, subgrids, and lattices involved.
B. Trace regular triangular grid over the special This is future work, and cooperation with mathemati
rhombic lattice cians is needed to help categorize them, always
1. all sides: same 2-fold S or Z curves keeping in mind that the descriptions and vocabulary
C. Trace regular hexagonal grid over the special should be accurate in a mathematical sense but, at the
rhombic lattice same time, understandable to the designer.
1. all sides: same 2-fold S or Z curve After this extensive overview of all relevant texts
2. Alternating sides: mirror concave and mirror by Huff available to me at this moment concerning
convex lines. parquet deformations, I want to share my thoughts on

Fig. 16: Robert Skolnik, “Opus Optimum”, India ink,
28 × 39.81 in. Basic Design Studio of William S. Huff, Spring Fig. 17: Loosening Huff’s rules. In this design two
1966, Carnegie Mellon University. © HfG-Archiv/Museum forms start to interact, a rectangle and a square.
Ulm, HfG-Ar, BDSA, Hu P 08. 020. (Editor's note: Original Design by the author. (Editor's note: Original
orientation is 90° clockwise) orientation is 90° clockwise)
some possible futures for parquet deformations as an them from getting lost in space. The accomplishments
exercise for designers, or “a subtle, intricate art form”, of the mathematicians and geometers who specialized
as Douglas Hofstadter describes it so well, but also on in this field are an open invitation for designers and
applications in design and architecture, the use of CAD architects to use this knowledge in their designs.
and the possibilities that arise when we loosen some In thinking of possible futures for the parquet
of the strict rules that Huff set up. deformation assignment, one can indeed find new
inspiration in books such as Tilings and Patterns, as
Possible Futures they are filled with ideas waiting to be explored. At
of Parquet Deformations least seven of the 28 methods in Flächenschluß can
In the long tradition of “small forms” in architecture be used to develop parquet deformations. Teachers
and design, otherwise called ornaments, a multifac need not be afraid: The topic is visual, and no for-
eted series of factors can be discerned in the course mulas are needed to experiment with patterns and
of the twentieth century. These include the widespread polyhedra. It is here that universities and art schools
reception of Adolf Loos’s notion of ornament, the ad- play a crucial role, although the reluctance to teach
vent of the International Style and the standardization about patterns and ornaments is still present, which
of building components, ultimately leading to the slow is understandable, because the topic was neglect-
disappearance of the visible use of these small forms ed—if not condemned—(see Adolf Loos’s “Ornament
in architecture and in the curriculum of architectural and Crime”) for such a long time. It is exactly through
education. The small forms became, at best, underlying, Huff’s renewed look at the “small” forms, available
invisible grids that helped to control standardization, or now through the ongoing research of mathematicians
they served in the background as proportion systems. and geometers, that this reluctance to small forms
Over the course of the twentieth century, mathemati- or ornaments can be overcome. Our students were
cians continued the work started by crystallographers completely open to and interested in the subject and
to systematically analyze and define spatial structures immediately fascinated by trying to find new solutions
in 2D and 3D. Huff knew Tilings and Patterns, which or new combinations of operations. The motifs in the
appeared in 1987, 26 years after the first parquet defor- history of patterns must be studied in a prospective
mation had been made. He also knew Flachenschlüß, way instead of a perspective way, seeing them as a
which appeared in 1963.49 Apart from his mentioning starting point for possible deformations and not as an
both books, they do not seem to have significantly endpoint or a result. Otto Antonia Graf, my personal
influenced the parquet deformation assignment. This mentor and professor at the Academy of Fine Arts
becomes clear through an arbitrary comparison of a Vienna, taught me this and always emphasized seeing
parquet deformation from 1998 with one from 1961. the “transformational potential” in forms, which is
During almost 40 years, Huff’s strict rules remained un- one of the great benefits of the parquet deformation
changed, although exceptions were allowed occasion- assignment. Another advantage is that students learn
ally (see Figure 5 as an example). This resulted in a very to think parametrically without the use of a computer.
consistent body of work of some 100 parquet deforma- Every deformation of one basic form is guided by a
tions, now archived in Ulm. In this book, one will find “transformational idea” which can be depicted in one
some of the finest examples. It would be of significant single drawing in which all the steps of the deforma-
benefit to bring them all online as study material to tion are seen at once. This is a very fruitful in-between
inspire future designers. The parquet deformation as step if one wants to use CAD or parametric model-
signment makes students more aware of the geomet- ing tools such as Grasshopper, since a computer will
rical patterns that surround them, they learn that these always need an “idea” in order to help in experimenting
patterns have all been classified (into 17 wallpaper with variations, improving subtleness and accuracy,
groups) and they grasp the underlying structural grids and so on.
and lattices that define them, as well as how these Another way to reinvent parquet deformations
are all related and how they can be transformed. The would be to loosen some of their strict rules, but then
notions of gradual change, geometrical coherence and one must find a new name for such an assignment, out
subtle transformations inherent in this assignment in- of respect for Huff’s legacy. This recalls Huff’s email at
crease the students’ visual acuities and spatial thinking. the beginning of this article: “You can design anything
The world of polygons and polyhedra, along with their you want. Just don’t call it a Parquet deformation.”
geometrical constraints and subtle transformational David Bailey’s phrase “geometrical tessellating meta-
possibilities, offer exactly that benefit. They are the per- morphosis”, Craig Kaplan’s “evolving patterns”, and Jay
fect playground for young students’ experiments. The Bonner’s “pattern manipulation” are good suggestions.
grids, lattices and space-filling structures offer them Which strict rules could or should be loosened? I
the necessary scaffolding to explore space, preventing have a few proposals: If one allows more than one tile,

any vertex can transform into a slowly popping-up poly- from Kuala Lumpur, founded Azmas Rugs, an initiative
hedron, circle, or any other form under the same rule that allows farmers and their families in India to gain
of temporality and subtleness. One could allow lattices an additional income on the loom in their backyard
other than the five Bravais lattices used by Huff. Many and started producing rugs with parquet deformations
of the diagrams in Tilings and Patterns could then serve designed by Craig Kaplan and myself. Azmas Rugs will
as a starting point, such as tilings based on star poly- also produce a carpet based on a work by one of Huff’s
hedra, as seen in the pioneering work of Craig Kaplan.50 students, and—this is important—Azmi Merican con-
At my invitation for this book, Craig Kaplan again made tacted the author, now a successful American archi-
new and pioneering research focusing on the possibil- tect, to ask for permission. Studio Nov24’s chief design-
ities of shapeshifting between Laves tiles, and he con- er, Hamideh Jafari, is currently developing a beautiful
centrates on the isohedral tilings described in Tilings collection of woven, high-quality carpets under my
and Patterns. If one allows more than one polyhedron guidance. They are manufactured in India and Iran and
from scratch—a feature which is already integrated into are based on parquet deformations. The young math-
the Parakeet software by Esmaeil Mottaghi and Arman ematician Edmund Harris decorated four walls around
Khalil Beigi—a new world of possibilities again opens. elevator entrances with very sophisticated 2D parquet
Jay Bonner, a specialist in Islamic design, shows the deformations printed as wallpaper.
possibilities of pattern manipulation through hinged In architecture, there is a need and demand for
tessellations, in which a device from a 3D operation is new, more sophisticated patterns that go beyond
used in a sequential way in 2D. We should also consider mere decoration. The potential of patterned screens to
allowing A and A‘ tiles, if the opportunity arises. provide shade for a façade, lend privacy to a balcony,
Finally, to bring new life to the assignment, one may or act as a divider in an office space is also obvious.
explore possible applications in the real world of archi- Kinetic façades and screens are another possible but
tecture and design to share this art form with the public. tricky challenge. The diagrams and the vocabulary
Applications in fields such as fabric and fashion design, used in kinetic design are, in any case, similar to the
graphic design, car design, and home decoration are parquet deformation vocabulary: flux, flow, movement,
obvious, and we have access to powerful laser cutters, temporality, sequences, framework, events, spatial
CNC machines, milling, molding, and casting techniques, transformation techniques, subtlety, and morphology,
as well as large-format printers, that may be used to among other terms.53 There are many more examples
integrate such designs into architectural projects. and many more applications to imagine, which I leave
Richard Lane’s parquet deformation “Crossover” to the reader’s imagination.
was used on a large scale by the famous architect The main reason that I see for a revival of morpho-
Emilio Ambasz in Pro Memoria Garden, a project for a logically transformed patterns, whether as a design
labyrinth in Lüdenhausen, Germany, in 1978. Although challenge for students or as real-life applications, is
Ambasz won the competition, the garden was for some that the concept behind them suits today’s challenging
reason not created, but Ambasz’s design is now in the times very well. Society must rethink and transform
Museum of Modern Art (MoMA) collection.51 Huff was many systems and structures. Experimenting with
not unhappy about the fact that a colleague of his had parquet deformations is an excellent illustration of the
used one of his student’s works, but what he could not transformation processes needed in the world. Space
tolerate was that neither Richard Lane nor Huff’s Basic tessellations is a general term selected as the main
Design Studio were credited or referenced, although title for this book, because it is about connecting the
the design was an exact copy of “Crossover”. Several unconnected through a new kind of spatial and struc-
letters to the MOMA and to the publisher, Rizzoli, are tural reasoning, exploring relationships and connections
proof of this.52 I mention Ambasz’s project, not only that have not been made before.
to show that Huff had nothing against applications if Space tessellations in architecture is ongoing
referenced properly, but also because of its scale: It research open to anyone. With the help of this book,
was a parquet deformation designed as a labyrinth we hope to reach architects, artists, and universities
made of hedges and at least as long as a football field. worldwide and invite them to the world of parquet
It was also a social project involving children in taking deformations and their inherent transformational
care of the garden. ideas. Three-dimensional corollaries of two-dimen-
Craig Kaplan shows in his chapter a very nice ap- sional geometric principles relating to the science
plication of a parquet deformation in low-relief casted of tiling are both fascinating and of high value to
bronze for the Museum of Mathematics in New York. geometric artists and architects alike. In particular,
Parquet deformations are also perfectly suited for wall the relationship between 2D parquet deformations
paneling (with acoustic properties) for spaces such as and 3D space-filling polyhedral networks is original
restaurants, hotels, and offices. Recently, Azmi Merican, research that has great potential for expanding the
design understanding and repertoire for architectural still in its infancy, apart from some important excep-
expression; this includes the architectonics of spatial tions.54 Space tessellations as a field has the potential
layout, vaulting design, and, especially, space-frame to introduce new realms of geometric inspiration and
structures. It facilitates and expands current under- beauty to contemporary artists and architects. In ad-
standing of the use of these 2D and 3D structures as dition to architects, geometric artists, and designers,
scaffolding for geometric patterns and architectural both mathematicians and computer programmers are
expression. Within the two-dimensional realm, this welcome to become key members of this research. I
has already been well documented, but, when extrap- truly hope that the research and experiments pre-
olated to three-dimensional geometry, this research is sented in this book will inspire you.
Acknowledgment
I would like to thank William S. Huff, his students, and his family for keeping Huff’s legacy
alive. The HfG Archive in Ulm, especially Marcela Quijano and Martin Mäntele for their support.
The staff at the Institute of Art and Design, TU Wien, especially Anita Aigner and Peter Auer
for setting up the assignments, the external lecturers, and all the students and student assis-
tants who participated in our experiments. I also want to thank Marie Reichel, the graphic de-
signer of this book, Eva Sommeregger and Christian Kern, co-editors and sparring partners in
realizing this project. For their support and feedback, I want to thank Claudio Guerri from the
University of Buenos Aires, Nicholas Bruscia and Rose Orcutt from the University of Buffalo,
Martin Aurand of the Carnegie Mellon University Architecture Archives and L
eslie Lubbers, di-
rector of the Memphis Museum. My gratitude goes to all the international contributors: Dénes
Nagy, Douglas Hofstadter, William. S. Huff, Cornelie Leopold, Tuğrul Yazar, Jay Bonner, Craig
Kaplan, Esmaeil Mottaghi, and Arman Khalil Beigi. Special thanks to David Bailey, who always
keeps me updated on parquet deformations. I am extremely grateful to Christian Kern for his
support and to the sponsors and private donors that made my work on this book possible:
Azmi Merican, Paul Mercelis, Kristof Morel, Irene Jochems, Greet Van Hoeydonck, Geert L M
Pauwels. Very special thanks to Sunnive Van Hoeydonck and Luisa Paumann for their ongoing
support. Finally, Birkhäuser’s team for this book: David Marold, Bettina. R. Algieri and Ada
St. Laurent, for making this book reality in a highly professional manner. A very special thank
you to Anita Aigner, who believed in my Space Tessellations project from the first second, and
to my mentors in the past, Gilbert Decouvreur and Otto Antonia Graf.
References
1
Email from William S. Huff to Werner Van Hoeydonck, 31 August 2019.
2
Email from Craig Kaplan to Werner Van Hoeydonck, 12 May 2017.
3
David Bailey’s webpage: http://www.tess-elation.co.uk/.
4
https://hfg-archiv.museumulm.de/en/hfg-collection/collection-hfg-stiftung/.
5
The nine major assignments that William S. Huff regularly gave to his students in his basic
design classes were 1. Symmetry or Programmed Design; 2. Two-Fold Mirror-Rotation (or Inversion)
Symmetry; 3. Mirror-Rotation-Dilatation; 4. Parquet Deformations; 5. Trisection of the Cube; 6.
One-Sided Surfaces: Variations on the Möbius Band; 7. Depth Cue; 8. Figurative–Ground; 9. Raster.
6
In 2015, William Huff in an email to Marcela Quijano, curator at the HfG Ulm archive in an
email to me in August 2018.

7
Online talk with Professor Maurizio Sabini, former student of William Huff, April 2021.
8
Ibid.
9
William S. Huff Papers, University Archives, State University of New York at Buffalo, MS
139.2. Item descriptions provided by William S. Huff, pp. 12–17. On page 17 Huff writes, “I
only now realized that the collection was considered to be lost.” https://library.buffalo.edu/
archives/pdf/ms-139-2-public-partial-inventory.pdf.
10
A full catalogue of the HfG library is available online as a downloadable 1,005-page pdf;
the “small” library contains around 6,000 books. https://hfg-archiv.museumulm.de/wp-
content/uploads/2019/01/f_05_bibliothek.pdf.
11
“Best Problems”, 1979. Tim McGinty, professor in the architecture department of the
University of Wisconsin, collected 24 assignments in a “Best Beginning Design Projects
Collection” and shared this compilation among his 23 colleagues working at different
universities across the US. In the introduction Mr. McGinty writes, “If you use any of these
projects, remember that they are the creative fruit of your peers and they deserve credit.
If you print or republish them, you should ask their permission.” For this compilation Huff
selected the parquet deformation and one of his other assignments, the mirror-rotation
symmetry assignment.
12
Lecture notes for SEMA, 2013; SEMA stands for Sociedad de Estudios Morfológicos de
Argentina. These lecture notes were sent to me by Claudio Guerri, honorary president
of SEMA.
13
Ibid.
14
In handling symmetry, I have moved my students from isometry (e.g., wallpapers),
through homoemetry (e.g., spirals) and syngenometry (e.g., deformations), to katametry
(e.g., programmed design). Katametry involves the lessening of regulation; nonetheless,
regulation remains. In William S. Huff, “On Regulation and Hidden Harmony”, Harmony of
Forms and Processes, Lviv, 2008, p. 3.
15
Selected papers from the international interdisciplinary symposium entitled Katachi
U Symmetry, held at the University of Tsukuba in Japan, 21–25 November 1994.
Two interdisciplinary concepts, katachi and symmetry, born in the East and West,
respectively, came together to further advance intercultural cooperation. The scope
of topics covered included: 1. Science on Form; 2. Geometrical Arts and Morphology;
3. Invisible - Visible I Viewing Invisible Images by Comparing them to Visible Forms; 4.
Sensing Order; 5. Symmetry, Dissymmetry, and Broken Symmetry in Art and Science.
Huff’s paper is titled “The Landscape Handscroll and the Parquet Deformation”, in
Katachi U Symmetry, T. Ogawa, K. Miura, T. Masunari, D. Nagy (eds.), Springer-Verlag,
Tokyo, 1996, pp. 307–314.
16
Huff, William S. “The Landscape Handscroll and the Parquet Deformation”, In Katachi
U Symmetry, T. Ogawa, K. Miura, T. Masunari, D. Nagy (eds.), Springer-Verlag, Tokyo, 1996,
pp. 307–314. For some good examples of handscrolls, see Willmann, Anna, “Japanese
Illustrated Handscrolls”, Heilbrunn Timeline of Art History, The Metropolitan Museum of
Art, 2000. http://www.metmuseum.org/toah/hd/jilh/hd_ jilh.htm, November 2012. Huff also
references Wen C. Fong, Beyond Representation: Chinese Painting and Calligraphy, 8th–14th
Century, Yale University Press, 1992, pp. 87–88.
17
Ibid.
18
Ibid.
19
Ibid.
20
Ibid.
21
“Best Problems”, 1979.
22
William S. Huff, “An Argument for Basic Design”, Ulm 12/13, Journal of the Ulm School of
Design, 1965.
23
Ibid.
24
Ibid.
25
William S. Huff, “Ordering Disorder after K. L. Wolf”, Forma, 15, Proceedings of the 2nd
Katachi U Symmetry Symposium, Tsukuba, 1999, Part 2, 2000, pp. 41–47.
26
See Note 12.
27
Huff, “Ordering Disorder”, pp. 41–47.
28
William S. Huff, “On Regulation and Hidden Harmony”, Harmony of Forms and Processes,
Lviv, 2008. References to Alberti: Leon Battista, “On the Art of Building in Ten Books”, trans.
from Latin by Joseph Rykwert, Neil Leach, and Robert Tavernor, MIT Press, Cambridge,
1485/1988, p. 183. References to Kahn: What Will Be Has Always Been, R. S. Wurman (ed.),
New York, 1986, p. 77.
29
William S. Huff, “Grundlehre at the HfG—With a Focus on ‘Visuelle Grammatik’”,
Ulmer Modelle – Modelle nach Ulm, Ostfildern-Ruit, Hatje Cantz, 2003, p. 196.
30
Huff, “On Regulation”, p. 77.
31
Huff, “The Landscape Handscroll”, pp. 87–88; Douglas R. Hofstadter, “Metamagical
Themas. Parquet Deformations: Patterns of tiles that shift gradually in one direction”,
Scientific American 249, July 1983, pp. 14–20. See also: Douglas R. Hofstadter, “Parquet
Deformations: A Subtle, Intricate Art Form”, Metamagical Themas—Questing for the Essence
of Mind and Pattern, Basic Books Inc., 1985, pp. 191–212. The integral version of 1983 with 14
examples of parquet deformations is reprinted in this book with Douglas Hofstadter’s kind
permission.
32
Huff, “The Landscape Handscroll”, pp. 87–88.
33
Ibid.
34
Huff, “Grundlehre”.
35
Huff, “The Landscape Handscroll”, pp. 87–88.
36
William S. Huff Papers, State University of New York at Buffalo; Maurice L. Zigmond,
Gotlieb Adam Steiner and the G. A. Steiner Museum, Journal of California and Great Basin
Anthropology, 1979, Vol. 1, No. 2: pp. 322–330. Zigmond quotes Huff’s offering an analysis
of the architectural concept which prompted his design of the building. Huff points out
that the individual cells of the floor plan “are interlocked in what is known as the ‘basket
weave’ pattern, but that this arrangement has nothing to do with the contents of the

interior. It was in the ’50s at the Alhambra (in Spain), that I reflected on the applicability of
such a tile pattern in structuring a building; it was not a nod to the Indian basket—though
not an unhappy coincidence! Also many suppose the building’s silhouettes reflect Pueblo
Indian dwellings. Again, I had no such thought; rather, the design is a result of my own
particular grasp of principles of architecture—proportion, composition, outline, etc. The
plan possesses 2-fold, rotational symmetry—to be distinguished from bilateral, or mirror,
symmetry. The front, though asymmetrical of itself is consequently the same as the back. It
is made up of 10 interlocking, repeating cells (and two towers) so laid out that they provide
a variety of spatial experiences: the cluster, the vista, the dead end.”
37
Ibid.
38
Major P. A. MacMahon, New Mathematical Pastimes, Cambridge University Press, 1921.
39
Lecture notes for SEMA, 2013.
40
Huff, “Grundlehre”, p. 187.
41
“The Theory of Transformations, or the Comparison of Related Forms”, D’Arcy Wentworth
Thompson, On Growth and Form, 2nd ed., Vol. 2, Cambridge University Press, 1959, pp.
1026–1095.
42
43
A classroom tutorial from 1983, University of Buffalo, sent to me as a PDF by Nick Bruscia,
Assistant Professor at the Department of Architecture, University of Buffalo.
44
Email from Douglas Hofstadter to Werner Van Hoeydonck, 29 April 2021. (Bill was the
nickname of William Huff)
45
A classroom tutorial from 1983, University of Buffalo.
46
A classroom tutorial from 1983, University of Buffalo.
47
48
Ibid.
49
Branko Grünbaum and G. C. Shephard, Tilings and Patterns, Dover, 2016; Heinrich Heesch
and Otto Kienzle, Flächenschluss: System der Formen lückenlos aneinanderschliessender
Flachteile, Springer, 1963; Robert Williams, The Geometrical Foundation of Natural Structure:
A Source Book of Design, Dover Editions, 1979.
50
C. Kaplan, “Islamic Patterns”, ACM SIGGRAPH Art Exhibition, 2008; C. Kaplan, “Curve
Evolution Schemes for Parquet Deformations”, Bridges Proceedings 2010, Mathematics, Music,
Art and Culture.
51
Emilio Ambasz’s Pro Memoria Garden was the winning entry in a competition for a
memorial that would remind future generations of the horrors of war. The unrealized project
consists of a series of small, irregularly shaped gardens divided by seven-foot hedgerows
and narrow paths. Children of the town of Lüdenhausen would be assigned one of the
plots at birth and assume responsibility for taking care of it at age five. This, it was hoped,
would teach them a respect for life. Over time, the hedges would be removed to make a
single large communal garden. Ambasz usually addresses the mystical and poetic side of
architecture in his work, but here he has used what he considers to be architecture’s ability
to produce myth-making acts to suggest a collective commitment to the performative
dimension of public space. His practice of giving “poetic form to the pragmatic”, as he has
described it, is in this case imbued with a specific political project. https://www.moma.org/
collection/works/648?artist_id=141&page=1&sov_referrer=artist.
52
William S. Huff Papers, University Archives, State University of New York at Buffalo,
MS 139.2. Item descriptions provided by William S. Huff, pp. 10–11. On p. 11 Huff writes:
“Aside from thickening Lane’s ink lines into hedges and using the reverse orientation, the
significant addition is two lollypop trees at entrance and end of the garden. See Ambasz’s
acknowledgments at the end of the catalog, which gives no acknowledgment to Lane or the
Huff basic design studio.” https://library.buffalo.edu/archives/pdf/ms-139-2-public-partial-
inventory.pdf.
53
Jules Moloney, Designing Kinetics for Architectural Facades—State Change, Routledge, 2011.
54
Robert Williams, The Geometrical Foundation of Natural Structure: A Source Book of Design,
Dover Editions, 1979.
Werner Van Hoeydonck received a master’s degree in Architecture in Ghent, Belgium in

1992. His final design project (Concert Hall Ghent) was graded with the highest distinction
and received the Van Hove Prize. His thesis on the work of Andrea Palladio and his Flemish
translation of Palladio’s Quattro Libri dell’ Archittetura was awarded with a scholarship at the
Academy of Fine Arts Vienna (1993–1996), where Van Hoeydonck studied under Prof. Dr. Otto
Antonia Graf and became acquainted with Graf’s analytical method of “drawing-thinking,
thinking-drawing”. Graf’s research on both Otto Wagner and Frank Lloyd Wright deeply influ-
enced Van Hoeydonck. In 1996 he organized a Graf lecture on Frank Lloyd Wright at the Henry
Van Velde Institute in Antwerp. In 1998 he wrote the main catalogue text for the Otto Wagner
exhibition in Ghent (Witte Zaal). After several working experiences in architectural offices in
Vienna and Belgium, in 2000 Van Hoeydonck founded his own practice, Tek7-architects in
Antwerp, Belgium, realizing around 100 projects, mainly private houses and renovations. In
2012 Van Hoeydonck returned to Vienna and founded Ornamental Art and Design, a research,
design, and art studio focused on geometrical patterns and ornaments. He regularly orga
nizes drawing workshops in his studio and at several art schools in Vienna. Since 2013 he is a
member of DESIGN AUSTRIA and exhibited at the Vienna Design Week and in Palais Neupauer
Breuner. In 2016 he started the research project “Space Tessellations”, which brought him to
lecture at the Technical University Vienna (Institute for Art and Design, WS 2017, 2018, 2021)
and at the Academy of Fine Arts (Descriptive Geometry, since 2018). The primary focus of
his research and teaching activities is to make young students aware of the potential of two
and three-dimensional geometry in architecture and design. In 2019 he took part in the SIS
congress and exhibition (The International Society for the Interdisciplinary Study of Symme-
try) in Kanazawa, Japan and published his first paper, “William Huff’s Parquet Deformations,
Three Viennese Experiments”. An adapted short paper with the same title was published for
the Bridges conference in Helsinki in 2020. Space Tessellations: Experimenting with Parquet
Deformations is his first book.
https://wernervanhoeydonck.blog/

Grundlehre at the HfG
— A Focus on “Visuelle Grammatik”
William S. Huff
i. An “Opinion”
We must complain about the over-schematic and

insufficient presentation of the achievements of Josef
Albers. As a clarification of the again and again debated
question of to whom the importance of the Bauhaus
didactics of that time and of today must be ascribed Hermann von Baravalle in
—we are speaking here mainly of the preparatory course the Grundlehre classroom,
(practical and formal education)—the book of Wingler1 is 1959.
of little help. Itten, as initiator of the Bauhaus didactics,
is overestimated by Wingler; Moholy-Nagy, as indefati
gable stimulator, is evaluated correctly; but Albers is
completely underestimated. […] Wingler seems to have
overlooked the fact that Albers took upon himself per- Helene Nonné-Schmidt
haps the most difficult task in the development of the lecturing, 1956/57, on the
Bauhaus didactics, a task, which he solved brilliantly, i.e., Theory of Color according to
he transformed the different and partly contradictory Philipp Otto Runge, whose
components (pedagogical activism, mystical expression color sphere shows above
ism and exaggerated constructivism) into a systematic, her head on the blackboard.
coherent and operable subject of teaching.
—Tomás Maldonado2
Josef Albers was 35 in 1923, when he began teaching

a section of the Vorlehre at the Bauhaus in Weimar; Tomás Maldonado, 1958.
Tomás Maldonado was 33 in 1955, when he began
teaching the Grundlehre (Foundation Course) at the Editorial Note: All three
Hochschule fur Gestaltung (HfG) in Ulm. These were images copyright HfG
essentially both of these men’s first teaching roles; Archiv/Ulmer Museum, Ulm.
each had come generally from similar backgrounds,
committed to art. In Albers’s case, he had to cast off
his earlier training in Germany’s prevailing Expression 1
Hans Maria Wingler, Das Bauhaus 1919–1933, Weimar,
ism, a vein of art that ran contrary to developments Dessau, Berlin und die Nachfolge in Chicago seit 1937,
in Paris and Moscow. In Maldonado’s case, the wiser Cologne, 1962.
from having the advantage of coming along three
decades later, he had taken a leading role in an 2
Tomás Maldonado, “Opinions: Is the Bauhaus Relevant
avant-garde movement, which referred anew to the Today?”, ulm 8/9, September 1963, p. 12. Liberties have been
same Paris and Moscow that had reoriented Albers. taken in amending punctuation.

ii. An Authentic Restoration of a He was the director of the school, head of Architecture,
Bauhaus Vorlehre in the HfG Grundlehre head of Product Form, and the head of the Grundlehre;
One former Bauhäusler praises the most legendary of only Otl Aicher headed Visual Communication, and
the successes of the Bauhaus: “None of us, probably Max Bense headed Information. The pity was that Bill
not even Gropius and Itten, nor even Moholy-Nagy and took on the air of the traditional European professor
Albers, who later directed the Vorkurs, could imagine who sometimes went to the lecture room, but more of-
at the time that it would be this very Vorkurs that ten delegated instruction to assistants and surrogates.8
would conquer the schools of arts and crafts through Bill was, in the main, a truant of his newly assumed
out the world.” 3 responsibilities.9 Of course, in the first academic year,
Unfortunately, the conquest was more impres- there was only the Foundation Course instruction to
sion than substance. After the abrupt closing of the deliver; and Bill had the Hochschule buildings to design.
Bauhaus in 1933, the school’s ideas were diluted and Yet, there was one redeeming glimmer: Bill’s
dispersed worldwide—the spread was broad, but shal- assistants and surrogates were an extraordinary lot.
low. This was the special fate of the Vorlehre. Whose Through the prestige that he had established in Zu-
Vorlehre, however, was being submitted for emulation? rich, Europe’s second art capital, along with his own
The problem is that there were innumerable versions overpowering, if not overbearing, personality, he was
of it and versions upon versions—the psyched-up able to attract a remarkable group of personages:
expressionistic version of Itten, the overly simplistic Bauhäusler who were still around—only 20 years
experimental version of Moholy-Nagy, the rationalistic having intervened between the closing of the Bauhaus
“open-eyed”4 version of Albers. Furthermore, the in and the opening of the HfG. He was able to assemble
struction of both Kandinsky and Klee, scheduled as a majority of the surviving Bauhaus faculty for the in-
part of the required Formlehre,5 considerably enlarged auguration of the new HfG buildings in 1955; but even
the host of permutations. before that, Bill, a former Bauhaus student himself,
Since at least the mid-twentieth century, most brought back two Bauhaus masters from his student
schools of architecture and design have claimed a days to resurrect, in effect, the Bauhaus Vorkurs:
foundation course that followed the precepts of the Josef Albers and Walter Peterhans. In addition, he
Bauhaus—though some have gone out of their way brought from earlier Bauhaus days the former mas-
to declare virtually the impossible, an unwavering ter Johannes Itten10 and a fourth Bauhaus person,
resistance to that sort of persuasion. An old issue the former student Helene Nonné-Schmidt. Even Bill
of the Yale Alumni Magazine—from around the time himself participated in the Grundlehre instruction.
that I took an introductory design course, required for The Bauhaus Vorkurs finally had a new home that was
entering Yale’s Department of Architecture—offers an genuine—and one where the Vorkurs could be re-
example from my personal experience. Under “Com- evaluated. It was up to Maldonado to do just that.
mencement Activities: Art”, the following was reported: The two most consequential of these former Bau-
häusler were Albers and Nonné-Schmidt—both giving
Other exhibits illustrate the work in some of the newer instruction in color, but to different ends. Albers pre
experimental courses conducted by Professors Gute and sented his renowned, two-decade, post-Bauhaus treat-
Switzer. These are developments from the famous Bau- ment of the perception of color. Solidly based on the
haus courses and demonstrate a philosophic approach teachings of Paul Klee and studies of her husband,
to the basic problems of form and materials which serve Joost Schmidt,11 Nonné-Schmidt’s hands-on instruc
as a common denominator for later specializations in tion thoroughly covered the theory of color (a topic
one of the arts.6 whose primacy, in the consideration of practicing de-
sign, was vehemently disputed by Albers).12
These teachers of mine do deserve my due respect; Max Bill, the original director of the Grundkurs,
their effort at reform was a resolute departure from the essentially adopted the Albers model of basic design.
Beaux Arts training for which Yale had been routinely Bill, arguably the leading exponent of Art Concret, pur-
acclaimed some few years before. That what was being sued principles in his art that were, if not coinciden-
presented had much of an affinity to the Bauhaus, tal to those of Albers, at least in rapport with them.
however, was sadly delusional—well, perhaps it did Aside from the building complex designed by Bill for
correspond to it in a very limited way.7 Ironically, only the HfG, a most agreeable masterpiece in facility and
a few years later, Yale engaged not only an authentic in temper, one of Bill’s most important acts was the
representative of the Bauhaus, but the one who most luring of Albers to teach during the school’s first two
epitomized the concept of basic design—Josef Albers. academic years—December 1953–January 1954 and
When the design school opened at Ulm in 1953, May–August 195513—if for no other reason than that
Max Bill dominated the curriculum, at least on paper. the second visit afforded Maldonado an opportunity to
66 Grundlehre at the HfG William S. Huff

3
Heinrich König, “The Bauhaus—Yesterday and Today”, in
Bauhaus and Bauhaus People, Eckhard Neumann (ed.), New
York, 1970: p. 120.
4
“When I arrived at Black Mountain College one of the boys
asked me what I planned to teach. ‘To open eyes’ was my
answer—and my first educational sentence in English.” Josef
Albers, Search Versus Re-Search, Hartford, CT, 1969: p. 11.
5
“Instruction in form”, courses in formal studies that
complemented the Vorkurs. Daniel Deboy, “Clocking the Day”.
Basic Design Studio of William S. Huff
6
“Art”, Yale Alumni Magazine 12, July 1949, p. 23. Spring 1990, SUNY at Buffalo.
7
I took the course during academic 1947/48. In a letter BDSA, Hu P 02. 050.
home, dated 28 September 1947, I wrote: “My art class is a Assignment: Programmed Design.
little puzzling though. In that class we aren’t painting, aren’t
drawing, aren’t etching, no—we’re ‘filling space’: Filling space
with planes, lines, and solids. Our media are cardboard,
9
wire, and clay. But it isn’t quite as crazy as it may seem, bauhäusler in ulm: Grundlehre an der HfG 1953–55, Ulm,
for it fits in perfectly with architecture (which in itself is 1993, p. 23.
filling space). It gives us some theories of composition.” In
10
fact, our training gave us little or no frame of reference. One It was widely rumored that, when Albers learned that ltten
assignment was stated this way: “Express Connecticut—in had been brought from Zurich (for but a mere week) in April
watercolors.” Such an assignment didn’t do anything but stir 1955, Albers declared that, after his second stint, which had
confoundment in me. Was there a future for me in design? followed ltten‘s visit, he would not return to the HfG as a
It was because of this preliminary training at Yale that I visiting teacher-and he did not. Albers feigned that Bill‘s
quested half-mindfully, so to say, and stumbled eventually gesture to ltten indicated a capricious regard for the integrity
into the Maldonado Grundlehre. of the Grundlehre.
8 11
I found the instruction at the HfG School to be more North Though Joost Schmidt, according to H. Nonné-Schmidt,
American than European in style. had been a close friend of Albers, Albers objected to Bill‘s
having hired Schmidt‘s wife to teach color—that component
of design that he thought was being thoroughly covered by
himself.
12
“Originally, we began our color course with a presentation
of various color systems, of color theories./With the
discovery that color is the most relative medium in art, and
that its greatest excitement lies beyond rules and canons,
a more sensitive discrimination was needed./The more a
creative use of color developed, the less desirable became a
merely trustful and obedient application./The seeing of color
became our first concern./As a result, we came to present
color systems not at the beginning but at the end of our
course.” Josef Albers, Interaction of Color, New Haven, 1963:
p. 66. And that presentation of color theory at the end of his
course was given short shrift.
13
bauhäusler in ulm, p. 17.
Jeffrew Orling. Basic Design Studio of
14
William S. Huff, Fall 1966, Carnegie Mellon In Maldonado’s original Kuhberg house, aside from one
University (CIT). © HfG-Archiv/Museum large painting of his own, he displayed no other paintings,
Ulm, HfG-Ar, BDSA, Hu P 05. 037. with the exception of one by student Mavignier and an Albers
Assignment: Raster—Single element. Homage to the Square.

interact with Albers.14 At the same time, Bill’s persu Maldonado, having joined his artistic inclinations with
asion of Maldonado to join the faculty must be listed a cadre of Argentinean Art Concretists in Buenos Aires,
as his third major contribution to the vital being of the came to have high regard for the work of both Max
new school—ironic as that turned out to be. Bill and Georges Vantongerloo. He went to Z urich in
194819 to get to know Bill (as well as to Paris to meet
iii. Basic Design Becomes Interdisciplinary Vantongerloo) and undertook to write a monograph20
Events that were no less imprudent in quirky human on Bill. Eventually urged by Bill to take a post at the
ways, than those that shaped the original Bauhaus HfG, Maldonado accepted with the stipulation that
Vorlehre, shaped the Grundlehre at the Hochschule fur he would go for one probationary year. Unsure in
Gestaltung. Max Bill had an incomparable model in his his German, he functioned largely as an aid to Bill
former teacher Josef Albers, who wired up the class- in the school’s second academic year; this included
room with the Gestalt reviews of his students’ work both administrative troubleshooting and assisting in
—characterized by Hannes Beckmann, another of the Grundlehre. That in time the two had the fateful
Albers’s Bauhaus students, as “a kind of group ther falling-out is a story for a different setting from this,21
apy”.15 Bill was inexplicably indifferent to this—putting but Maldonado’s involvement with writing on Bill’s
his professional commitments and artistic preoccupa- conceptualization of art and, later, his observation
tions in Zurich above his performances in the class- of Albers’s teaching, very likely helped to temper his
room at the HfG. It was a grievous miscalculation on reaffirmation of a purely formal approach to the visual
his part. Tensions over teaching practices, personality training part of the Grundlehre course.22
clashes, even private indiscretions, caused a political That being the case, Maldonado gave basic design
upheaval that split the HfG student body: two to one a vigorous infusion of Art Concret with his insightful
(pro-Bill vs. anti-Bill). Virtually by default, yet by and inclusion of two branches of geometry, namely sym
large ready, Maldonado was thrust into the looming metry theory and visual topology.23 Elements of Art
void. Maldonado rendered the all-but-indispensable Concret had already been introduced by Bill himself in
Bill dispensable. the few assignments that he had presented—and by
Maldonado’s 1963 “opinion” about Albers’s contribu- Peterhans, too. Maldonado can be said, then, to have
tion to “the preparatory course” has to be indicative of expanded that base with the lectures on symmetry24
what he had in mind when he took on the Grundlehre and topology—and by the design studio assignments
in 1955—after having assisted Bill’s Grundlehre during that addressed each of these disciplines.25
the preceding academic year. Maldonado reaffirmed The other interdisciplinary body of knowledge that
the Albers basic design model, as it was eventually Maldonado brought into play was perception, particu
articulated,16 by avoiding so-called semantic issues, the larly Gestalt psychology. Subjecting the geometric
materialization that takes place in the applied design artifact to the perceptual experience gave a validating
process, as well as pragmatic issues, the appraisal of nod to Albers’s aphoristic formulation: “The origin of
the usefulness of the product of design,17 while effec art: The discrepancy between physical fact and psy
tively dealing with only syntactic issues of design— chic effect.”26
form, color, and texture. However, he made a signifi- But Albers had not initially come to this. Since
cant contribution to that model by supplementing he had, while at the Bauhaus, rejected the neurotic
the visual training component of the Grundlehre with expressionism of Itten and, yes, the over-formalization
interdisciplinary content: that is, with abstract bodies of of van Doesburg, that which was left at his disposal
knowledge that augmented the universally recognized in the Bauhaus’s store of “new ideas”27 was the study
formal (form, color, texture) issues of design. In line of materials at a basic level.28 He quarried this vein
with this, Maldonado has stated the following: imaginatively with his students in workshop probes
that focused on the potentials of the “physical-me-
[The Bauhaus didactic, particularly its Vorkurs] was a chanical qualities”29 of various, intentionally chosen,
question of argumentative exaltation of expression, common materials: paper,30 metal wire, metal sheets,
intuition, and action, above all of “learning by doing”. wire screening, glass, and even sand. This “idea” was
But this educational philosophy is in crisis. It is incapa- embedded in the very fabric of the Bauhaus (if the
ble of assimilating the new types of relations between Bauhaus had even one thread of consistency through-
theory and practice, engendered by the most recent out most of its history)31 for the students’ options in
scientific developments. We know now that theory must design studies were discharged in the so-called work-
be impregnated with practice, practices with theory. It is shops (Werkstätte), where each student was expected
impossible today to act without knowledge, or to know to master one particular material of choice:32 stone
without doing.18 (sculpture), wood, metal, textiles (weaving), color (mu-
ral painting), glass (stained glass), clay (pottery).

15 23
Hannes Beckmann, “Formative years”, in Neumann, Though both symmetry and topology were already well de-
Bauhaus and Bauhaus People, pp. 197–198. See also: “The veloped by the nineteenth century (through Kepler, Euler, Klein,
group discussions of the results of the exercises induce Möbius, Peano, Schönflies, Federov, etc.), both are quite recent
accurate observation and new way of ‘seeing’”, Josef branches of geometry, relative to Pythagorean and E
uclidean
Albers, “Creative Education”, in Wingler, Bauhaus, p. 143. geometry; and neither had yet been specifically applied to
design. One intriguing coincidence (yet there are no known con-
16
As noted in a later passage of this section, there was nections) is the parallel set of topics that have shown up in the
one notable exception to the exclusivity of formal con- world of graphic artist M. C. Escher.
tent in Albers’s compass of basic design: his compel-
24
ling workshop directive to students to poke around the Hermann Weyl and K. L. Wolf both lectured on symmetry at
physical properties of materials. With this sort of study, a the HfG; Helmut Emde, an anthroposophist, presented geom-
foray into concrete substance, Albers’s course technically etry and mathematics in a thorough supplemental course in
ventured over the boundary line of the strictly syntactic which all students engaged.
realm of formal abstractions and into the semantic realm
25
of materiality. The inclusion of these geometries might have origins in
Maldonado’s work on the Bill monograph. In an article in the
17
Initiated in academic year 1955/56, Maldonado’s lectures monograph, “The Mathematical Approach in Contemporary Art”,
on semiotics (covered later in this paper) were primarily Max Bill stated, “I am convinced of the possibility of developing
based on Charles Morris, though C. S. Peirce was appropri- art wherein the mathematical approach is fundamental”, and he
ately cited as Morris’s forerunner. Even though there is a further stated, “The mainspring of allvisual art is Geometry, the
general disagreement today among semioticists over some correlation of elements on a surface or in space. Thus, even as
interpretations that Morris attached to Peirce, Morris’s mathematics is one of the essential forms of primary thought and
three terms syntactic, semantic, and pragmatics have been consequently one of the principal means by which we take cogni-
considered to be very effective—useful even in concep- zance of the world that surrounds us, it is also intrinsically a sci-
tual fields beyond linguistics, such as design. See Charles ence of the relationship of object to object, group to group, and
Morris, Foundation of the Theory of Signs, Chicago, 1938. movement to movement. And since this science encompasses
all these phenomena and gives them a meaningful arrangement,
18
Tomás Maldonado, “New Developments in Industry it is natural that these relationships themselves should also be
and the Training of the Designer”, in Architecture Culture captured and given form.” Maldonado, Max Bill, p. 37.
1943–1968: A Documentary Anthology, Joan Ockman (ed.),
26
New York, 1993, p. 299. Albers, Search, p. 10.
19 27
Mario H. Gradowczyk and Nelly Perazzo, Abstract Art Eugen Gomringer, Josef Albers, New York, ca. 1968, p. 27.
from the Rio de la Plata: Buenos Aires and Montevideo,
28
1933–1953, New York, 2001, p. 48. “We begin at the beginning, which is (and has been in all
essential production) the material itself.” Albers, Search, p. 33.
20
Tomas Maldonado, Max Bill, Buenos Aires, 1955.
29
Ibid.
21
See Tomás Maldonado, “From Buenos Aires to Ulm”,
30
form + zweck 20, 2003, pp. 15–21. Albers reprised his famous Bauhaus paper-folding assign
ment during his first HfG visit, 1953/54.
22
A point will be made here and kept in mind elsewhere in
this chronicle: Grundlehre is not neatly synonymous with 31
The HfG held students to a high standard in regard to craft;
basic design. In a recent email (5 June 2003), Maldona- and it backed up those expectations with well-equipped
do cautioned: “One must not confuse the Grundlehre AT workshops (plaster, wood, metal, and eventually plastics),
ULM that consisted of a set of many disciplines, with the where all students were trained by first-rate masters in all of
‘Visuelle Einführung’ (or Visuelle Grammatik) that was only the school’s selection of shops, including photography. Craft at
one of them.” To be sure, I am focusing here, though not the HfG was, perhaps, a virtuous compulsion—it stopped short
exclusively, on the history of that Visuelle Grammatik, as of being a fetish.
it was incubated by Itten, formulated by Albers, filtered
32
through Bill, and construed by Maldonado. Maldonado’s This was part of a delusory retrieval of medieval practices,
point about basic studies has been a part of my compre- symbolized by Lyonel Feininger’s woodcut of the Cathedral of
hension about basic design from the earliest occasion Socialism and by the signature name of the Bauhaus itself—a
of Maldonado’s having made clear to me the distinction belated attempt to turn the clock back on the 500-year run of
between the two. (This point is visited later in this paper.) the Renaissance.

Yuqing (Karen) Li, “Checkering of Grid
over Grid: 16 in 15”. Basic Design Studio
Assignment: Programmed Design.
(Editor's note: Original in color)
It cannot, in truth, be claimed that the Bauhaus Gestalt and depth perception, as well, and presented
defined basic design—since not even Albers’s activities, both in a comprehensive manner—far more thoroughly
while there, could have constituted that. Its definition than Albers did. Furthermore, Maldonado’s directives for
had to come out of Albers’s subsequent achievements— exercises that involved perception (as can be seen in
those that were fashioned at Black Mountain and descriptions, given below) were far more elaborate than
clinched at Yale.33 his directives that addressed symmetry and topology.
Thus, Maldonado’s earlier statement on Albers’s The case can be made that Gestalt was dealt with un-
achievement at the Bauhaus (the sounding board of this evenly and almost furtively by Albers (which did include
paper) has to be reconsidered in respect to a later state- his Gestalt teaching), but comprehensively and direct-
ment: “It must be said that the best qualified historians ly by Maldonado. In 1955/56, lecturers Hans Joachim
of the Bauhaus doubt the existence of a unified didactic Firngau and Herbert Schober were brought to the HfG to
principle in the preparatory course—as much at Weimar bolster the psychology of perception.37
as at Dessau.”34 Maldonado, who had been pondering the problem
Probably tentatively, at first, Albers came to in- of the background, otherwise ground, as a problem
fuse basic design with Gestalt. Beginnings of this can of Art Concret,38 was introduced to Gestalt theory in
very likely be accredited to Karlfried Graf von Durck- 1946 by Aldo Pellegrini, a medical doctor and Surrealist
heim-Montmartin’s lectures on Gestalt psychology poet—a major figure in the Argentine art scene of the
at the Bauhaus, which had been arranged by Hannes day. Pellegrini had suggested to Maldonado that an ap-
Meyer.35 The equivocal figure-ground phenomenon proach to his question might be gained through Gestalt
struck Albers immensely; and aside from Gestalt proper, and gave him two texts: Köhler’s Gestalt Psychology
Albers became intrigued with the contradictory read that and Koffka’s Principles of Gestalt Psychology.39 Maldo-
could be achieved through the manipulation of depth nado’s involvement with Gestalt was, however, not an
cues in linear perspectives.36 Maldonado picked up on immediate embrace; rather, it was an evolving affair.40

The radical organization of the Bauhaus around one’s in Ulm in 1953, the divisions of design were arranged and
mastery of one material was, in the long run, unsustainable. It named along practical professional lines: Architecture (later
was essentially replicated at the New Bauhaus (later, School Building); Product Form (Industrial Design), Visual Communica-
of Industrial Design) in Chicago. In this case, however, certain tion (Graphics), and Information (Verbal Text).
materials were already being bundled into six workshops as,
for instance, wood and metal; glass, stone, clay, and plastics; 33
Though I am referring here to the conceptualization of
etc., Wingler, Bauhaus, p. 194. Following the Bauhaus imper- what basic design is, the term itself has its murky history.
ative in regard to material, Bauhäusler Hin Briedendieck was Albers claimed, with justification I am sure, that he settled
responsible for Chicago’s preliminary workshop, which could on the term “basic design” (undoubtedly while at Black
put on exhibit “strikingly new and strange” (Moholy-Nagy: be- Mountain) as the least unsatisfactory term in a long list of
low) arrays of “wood carvings, mechanically produced” under unsatisfactory terms.
the brief, “Problem: To determine the specific possibilities of
34
machine-working of wood”, Wingler, Bauhaus, p. 598. Ockman, Architecture Culture, p. 298.
Briedendieck eventually acknowledged shortcomings of this
35
Bauhaus “idea”. In his article “The Legacy of the Bauhaus”, Art Individual psychology vis-à-vis social psychology. Wingler,
Journal 12, 1, 1962, pp. 16, 18, 20, he shared some of his misgiv- Bauhaus, pp. 10, 159.
ings:
36
[The Basic Workshop] was intended to “release the Albers’s Constellations came out of this.
creative power of the student”, Moholy, The New Vision, p. 20.
37
[…] The student is not required or encouraged to produce “Firngau was the last surviving Gestalt psychologist from
“premature practical results” (Moholy). Instead, he is offered the famous Berlin School. His teaching of the psychology of
an opportunity to experiment freely with various materials perception was exemplary. Schober […] related the latest
and tools. There is a strong emphasis on initiative within a in the physiology of vision—a field in which he was ‘the’
“do-it-yourself” set-up, using conventional and unconven- German authority.” Email from Dolf Zillmann, Dean Emeritus
tional means, often achieving strikingly new and strange for Graduate Studies, University of Alabama, 18 June 2003.
configurations. […] Schober went on to lecture in later years at the HfG.
A meeting arose from a general concern expressed […] by oth-
38
er faculty members. “Although the purpose of the Foundation Tomás Maldonado, “Concrete Art and the Problem of the
Course is to allow the student to develop his creative abilities Unlimited”, in Tomás Maldonado, Mario H. Gradowczyk and
freely and without restriction—in the following semesters, William S. Huff (eds.), Buenos Aires, March 2003. This paper
where the students are channeled in the direction of practical comes two years after his introduction to Gestalt.
problems, the smallest limitation becomes a new obstacle
and his creativeness has shown a tendency to ‘freeze’. In most 39
Wolfgang Köhler, Gestalt Psychology, 1929; Kurt Koffka,
cases they completely ignored all their previous training and Principles of Gestalt Psychology, New York, 1935. “Two years
fell back on the conventional.” later Pellegrini published in ‘Argonauta,’ the publishing
The following remarks by Briedendieck addressing this house he owned, the Spanish translation of Köhler’s book
concern have been rearranged: (Psicologia de la Forma).” Email from Tomás Maldonado,
[The student] may well be merely the extension of his 27 May 2003.
tools or even an unwitting victim of the numerous incidental
40
events in the process. […] Learning by doing dominates. The “In my paper of 1946 I mentioned ‘psicologia della
emphasis has been on the manipulative aspects, on train struttura’ (Gestaltpsychologie). However, I believe […]
ing rather than knowledge. […] But it is precisely the aim that there was not an explicit and continual influence of
of design education to impart to the student the means of Gestalttheorie in my way of thinking. Originally the dichotomy
achieving authority and command in order to gain ascendancy of ‘figura-fondo’ doesn’t come from Gestalttheorie, but
over the accidental. rather from the traditional art-historian terminology. Only
These “accidents” of the Chicago students were apparently not later, between 1946 and 1948, I began to be convinced of the
difficult to come by—and the students must have been duly importance of Gestalttheorie for Concrete Art.” Maldonado,
praised for them. When an even newer Bauhaus was founded 27 May 2003.

Maldonado’s subsequent, comprehensive cover- It is to be noted that Albers also used the term ba
age of Gestalt at the HfG corresponded to his own sic studies as distinguished from basic design—essen-
remonstrance to “impregnate” knowledge. Albers, tially in the same manner that Maldonado did. The only
on the other hand, would exhort his students not to difference was that, while Maldonado’s basic studies
read books: There were no solutions for their design referred to the advanced practical years in a school of
assignments in any of them, he would warn. Then, design, Albers’s basic studies referred to the advanced
suddenly in a group critique, he would talk all around practical years in a school of fine arts. In the latter part
the Gestalt principle of figure-ground in such a manner of a succinct statement on basic design for the Yale
that an innocent observer might be led to believe that Alumni Magazine Albers wrote this:
Albers was discovering that principle in a student’s Our department of design/therefore promotes
work—at that very moment. Aside from probably hear- particularly/basic studies:/
ing von Durckheim-Montmartin lecture, Albers must Basic Design and Basic Drawing/Basic Color and
have read some “forbidden” books. (He himself wrote Basic Sculpture,/also Lettering and Drafting/as re-
some significant ones.) To be sure, Albers was a bit of quired training/for specialized studies:/
a Schwindler—one of his favorite interlingual words in drawing and painting/in graphic design and pho-
whose Germanic form he preserved in his excellently tography/in typography and printmaking/elemental
articulated English. and structural sculpture.41
In his Grundkurs at the HfG, Maldonado did intro-
duce in lecture form other bodies of knowledge, fore- It is important, then, in making the argument that
most among which were ergonomics and semiotics. It basic design is a discipline unto itself, that the clear
is important to emphasize, however, that Maldonado distinction between basic design and basic studies be
never tainted his Grundlehre assignments, in which kept in mind.
visual training was involved, with a mandate to apply There were essentially two important years of the
material of either an ergonomic or a semiotic nature— Maldonado Grundkurs: Academic year 1955/56 was
while both symmetry and topology were specified in its formative year; academic year 1956/57, in which I
certain assignments; for mandating either ergonomic took part, was its consolidating year. In the years that
or semiotic material, as a part of a visual task, would immediately followed, Maldonado’s participation in the
have been to insert semantic and pragmatic issues into Grundkurs tapered off, as he shifted his focus of in-
it. Thus, the basic design component of the Grundlehre terest, first, to the Visual Communication Department,
under Maldonado, though expanded interdisciplinarily where he worked out practical studies in visual semi
in respect to Albers’s version of it, remained wholly ab- otics,42 and, later, to Product Form, where he stressed
stract, wholly nonobjective—dealing purely and solely ergonomics. This shift coincided with his subsequent
with formal, syntactic, issues. roles in the administration of the school and the
Maldonado’s justification for the inclusion of ergo- buildup of its curriculum along methodological lines.
nomics and semiotics in lecture form in the Grundlehre In fact, in the two unique years of the Grundlehre,43
was that these were issues that should be covered by during which Maldonado essentially developed his
basic studies (to be distinguished from basic design)— Visuelle Einführung (introductory visual training), he was
practical basic issues that students would encounter trying and testing many ideas that would later shape
when they advanced to the specific design departments the curriculum of the school as a whole. By 1962, he
of their choice. had revived the HfG journal, ulm,44 in which HfG ideas
Ota Ulc, “The Out and In of Yang and

Yin”. Basic Design Studio of William
S. Huff, Fall 1992, SUNY at Buffalo.
BDSA, Hu P 05. 031.
Assignment: Conflicting Depth Cues.

41
Josef Albers, “To design is”, Yale Alumni Magazine 21, April Hermann von Baravalle, Helene Nonné-Schmidt, Otl Aicher,
1958, p. 6. and Herbert Ohl (the last in 1956/57 only). Maldonado
continued to teach the visual introduction component of the
42
Reportage and corporate identity were two other feathers Grundlehre in the years of 1957/58, 1958/59, and 1959/60.
of Visual Communication that had been pursued from the
44
department’s inception. Its squarish format, which made shelving difficult, was
scrapped for a more conventional vertical format. The first
43
Maldonado was not the only instructor of what the issue in this form was ulm 6, October 1962, designed by
Grundlehre was as a whole: other instructors were Tomás Gonda.
Marsha Berger, “In-tri-cut”. Basic Design

Studio of William S. Huff, Fall 1970,
Carnegie Mellon University (CIT).
BDSA, Hu P 08. 021.
Assignment: Parquet Deformations.

could be disseminated to a larger design community the Raster: No assignment mandated a raster, yet
of academics and professionals—a goodly number of somehow the potential of rasters was discussed
whom had already shown considerable curiosity in the (halftone and three-color printing with black—and even
Hochschule by their pilgrimages to the Kuhberg. TV rasters).50 Furthermore, our class was aware that
handsome rasters had been executed in the preceding
IV. The Maldonado Grundlehre Assignments class. The vexing question of the raster was only re-
A list of the Aufgaben (assignments) from the Maldonado cently clarified by Dolf Zillmann, who had participated
section of the Grundlehre of 1955/56 is as follows:45 in the Grundlehre of 1955/56:
Visuelle Einführung
(visual introduction) Kurs Maldonado To the best of my knowledge, it all [the first appearance of
⟩ Sierpinskifläche (Sierpinski curve) the Raster] came about spontaneously in dealing with “un-
⟩ Peanofläche (Peano surface) genau durch genau”. In addressing this Aufgabe, Peter Seitz
⟩ Weierstrass-Kurve (Weierstrass curve) […] and I used Rasters in different ways.51 It was not part of
⟩ Schwarz als Farbe (black as color) the assignment. In fact, nobody thought it was a big deal
⟩ Symmetrien (symmetries) —like greatly innovative. Only after Tomás, upon comple-
⟩ Genau–Ungenau (exact through inexact) tion of the Aufgabe, singled out the Rasters as something
⟩ Ungenau–Genau (inexact through exact)46 special and as a wonderful solution to his problem did the
⟩ Räumliche Wirkung (spatial effect) Raster become the focus of attention. The Raster became
⟩ Gleichgewicht dreier Flächen popular overnight. And as you know, Almir Mavignier
(balancing of three planes) helped to make it a bit of a centerpiece of concrete art.
⟩ Störung (interruption)47 Be this as it may, it was Tomás who saw something special
in the Raster. To us, it was merely a tool to solve an as
A list of Aufgaben from the Grundlehre of 1956/57: 48 signment, and our eyes had to be opened to it.52
⟩ Symmetrie (symmetry)
⟩ Parkettierung (parquetry—now termed “tiling”) The raster, which, as texture rather than artifact, had
⟩ Näherung, Zweideutigkeit, Gleichheit–Gestalt already struck the bent of the young Art Concretist,
Übung (proximity wins; neither wins; similarity suddenly showed its “didactic importance”;53 it became
wins—Gestalt exercise) something that was wanting of further exploration.
⟩ Vordergrund–Hintergrund, Zweideutigkeit, Some of Maldonado’s assignments were, very likely,
Hinter-grund–Vordergrund–Tiefenwahrnehmung exploratory for himself as well as for his student. Zill-
Übung (foreground, central square of format ad- mann is dead right: The good pedagogue is on constant
vances; ambiguity, central square neither advanc- alert, knows when he sees it, and is ready to promul-
es nor recedes; background, central square re- gate something extraordinary that occurs under his
cedes—depth perception exercise) instruction. It is not unreasonable to conclude that this
⟩ Modifizierte Peano Kurve mit verschiedenen was a part of Maldonado’s pedagogic strategy.
Farben und Schwarz als noch eine Farbe–Schwarz Maldonado presented a group of assignments
muß nicht ein Loch sein (modified Peano curve (listed below in the order in which they were given) to
with several colors and black as one of the col- my first-year architectural class in Pittsburgh in 1963.54
ors—black must not become a hole) Many were variations of those that had been given at
⟩ Ungenau durch Genau oder Ungenauigkeit durch the HfG in the Grundlehre of either 1955/56 or 1956/57.
Genauigkeit (inexactness through exactness) In sum, they comprised a characteristic cross-section
⟩ Genau durch Ungenau oder Genauigkeit durch of the Maldonado repertoire of exercises—though due
Ungenauigkeit (exactness through inexactness) to the short duration of his visit, he did not present any
assigned, but none were executed of his more mentally challenging assignments, such as
⟩ Wahrnehmung Übung (perception exercise)—in inexactness through exactness.
two rows, a six- or seven-step deformation of a
square ring into a succession of round-cornered a. “Two Rows of 7 to 9 Transformations of a Rect-
rings: upper row of rings deformed mechanically angle”: from a central elongated rectangle, trans-
in order to actualize optical problems; lower row formations to the left, to be the inverse of those to
of rings altered visually so that all optical pro the right. Upper row to be done with straight lines,
blems appear to be corrected, especially that all lower row with curved lines.55
rings appear to belong to the same family49 b. “Series of 6 to 7 Steps of the Transformation of
a Square Ring into Successive Round- Cornered
In looking back at the 1956/57 Grundkurs in which I Rings—But not into complete Os”: upper row, in
participated, there was a question in my mind about black, to be altered optically in order to counteract

Thomas Davies. Basic Design Studio of
William S. Huff, Spring 1964, Carne-
gie Mellon University (CIT). © HfG-Ar-
chiv/Museum Ulm, HfG-Ar, BDSA,
Hu P 02. 036. Assignment: Symmetry.
Note by Huff: “Four errors in execution
of program.”
45
This list came from Hans (Nick) Roericht. Reviewed by TV tube, line by line. He speculated that perhaps the Peano
myself in the HfG-Archiv collection, the identification plates, curve could be a better path for the raster. I am advised that
documented by different students on their submissions, some computer chip circuitry has since been designed in
show a certain fancifulness in the range of names that were that manner.
individually given to the different Aufgaben.
51
See Lindinger, Hochschule für Gestaltung Ulm, p. 45.
46
Roericht does not list this Aufgabe, but Dolf Zillmann
52
responded to it in an exemplary way. Email from Dolf Zillmann, 4 February 2003.
47 54
Roericht’s index of Aufgaben lists one title as Email from Tomás Maldonado, 3 July 2003.
“Strömungen”. It may be in error for “Störung”, since the HfG-
55
Archiv has student work from 1955/56 with this description. Nine exercises assigned by Tomás Maldonado to first-year
In the 1954/55 Grundlehre, Albers had presented his classic architectural students (Basic Design Studio of William S. Huff)
assignment, titled “Gestörte Ordnung”. See Hochschule at Carnegie Institute of Technology, 25 March to 19 April 1963.
für Gestaltung Ulm: Die Moral der Gegenstände, Herbert These descriptions were carefully recorded in 1963. Exercise B
Lindinger (ed.), Berlin, 1987: p. 34. was critiqued on 3 April; A, on 9 April; D, E, F, G, and H, on 19
April. No solutions were executed for Exercises C (the faculty
48
From the memory and files of W. S. Huff. member in charge of the shop would not allow plaster to
come into it) or I (not enough time was left for it).
49
This classic exercise in this form was found among Prior to this time, this exercise was unknown to me. The
undated notes of mine. I have still not found a classmate HfG-Archiv has a variant of this exercise, executed at the HfG.
who can verify its having been assigned to the 1956/57
56
Grundlehre, yet that is most likely the time of its first In a purer version, which may have been assigned in the
appearance. A variation of this was given at Carnegie Institute 1956/57 HfG Grundlehre, only black was applied to the rings
of Technology in 1963. Maldonado connected this assignment (see above).
to the subtlety of typefaces, which are adjusted to overcome
57
optical problems. This exercise was unknown to me before 1963. The
assignment’s directive matches a sculpture by Max Bill:
50
Maldonado, who had contacts with Telefunken at the Column with Triangular and Hexagonal Section, 1966. See Max
time of my Grundlehre, had noted that the TV image was Bill (Buffalo Fine Arts Academy and Albright-Knox Art Gallery,
transmitted as a raster of dots that streamed across the 1974), p.

various perceived flaws. Lower row, the same in articulate exponent of Art Concret, he began to build
two colors—one color for the ring, one color inside his immense library of information. Certainly, however,
the ring.56 two of his most challenging assignments had to have
c. “Transformation of a Shaft with One Regular come out of his period in Zurich with Max Bill: “Un-
Polygonal Face at One End and with a Different genau durch Genau” and “Genau durch Ungenau”. In his
Regular Polygonal Face at the Other End”: to be monograph on Bill, the initial draft of which accounted
executed in plaster.57 for a large part of the time that he spent in Zurich in
d. “Black as a Color, Not a Hole”: on an 8 by 8 grid, 1948, he wrote about such conundrums:
use colors; from 6 to 10 squares are to be black.58
e. “Elementary Raster Problem”:59 using only one The new theme trend is also manifest in the pictorial
element (in size and shape), located at the inter- production of Bill. Certain notions as continuous - dis-
sections of a grid, have the field of elements pro- continuous, precise–imprecise, limited–unlimited, which
duce a Gestalt through the rotation (orientation) of constitute his favourite thematic repertoire, are, undoubt-
the element only. Repeat the square formatted field edly, filtrations coming from the most up-to-date sci-
twice on the same plate: the element of the first entific expression. […] The lines, dots and sfumatos help
field to be rendered in black, differentiated colors invalidate the traditional concept of precision as it was
to be added to the element of the second field. understood until Mondrian—to give way to a new concept:
f. “Peano Curve”: create a Gestalt with subtractive the imprecision–precision. Imprecise–precise because in
color.60 works of art of this type the purpose is constructive, even
g. “Modified Peano Curve”: apply two colors of the when certain external resources be diffuse and not very
same value.61 strict.63
h. “No Prima Donna”: within a horizontal rectangular
format, arrange ten vertical stripes: three in differ- An unpublished paper (until 2003), which he wrote in
ent textures, four in different colors, three in black. 1948, takes up what Maldonado considered at the time
No stripe is to dominate. (as mentioned above) to be the main problem of Art
I. “Sierpinski Triangle”: the smaller (upright) trian- Concret:
gles are to be colored in R, Y, and B, according to a
given diagram. Apply grays to all (inverted) trian- Figure versus ground is the fundamental issue of Con-
gles between the colored triangles; the grays must crete Art. Any figure on a ground determines a space. If
be optically adjusted for balance in both hue and this occurs within a plane, on its surface that space is
intensity, according to their surrounding colors.62 illusory. […] Concrete Art is a continuous effort to destroy
this illusory space. […] That is why recently two great
Not surprisingly, Maldonado altered his assignments
from class to class, and frequently shuffled a directive
for one format to a different format. In most of his
exercises, many of which were tests (études) in percep-
tual skills, the layouts were strictly prescribed; and the
optical irritations that were identified (often subcon-
scious and subtle, yet with the power to distract from
the harmony of the whole) were to be altered to suit
the eye. The outcomes, either more or less successful,
were subject to an intense saturation of critical aes-
thetic absorption. The collective group opinion of the
trained and in-training (instructor and students) yielded
(after interactive deliberation) the verdict on the degree
of success of such exercises in respect to its aesthet-
ic evaluation. This transpired by calling upon a faculty
that every able designer must attempt to master—thus
the utter necessity that students develop a sharpened
aesthetic response able to judge every phase of the
assignment along its development—which is to say, to Andrew Liu “Limbo”. Basic Design Stu-
carry through any act of design. dio of William S. Huff, Fall 1996, SUNY
Maldonado’s more open assignments were posed at Buffalo. © HfG-Archiv/Museum Ulm,
as challenges, which could lead to surprising out- HfG-Ar, BDSA, Hu P 04. 022.
comes. Some of the themes of his assignments, can, Assignment: Conflicting Depth Cues.
very likely, be traced to Buenos Aires, where, as the

Robert Hand, “Figure–Texture–Ground
Study–Variant 2, Five by Fifteen Check-
ers”. Basic Design Studio of William S.
Huff, Spring 1964, Carnegie Mellon
University (CIT). © HfG-Archiv/Museum
58
This directive was linked to a modified Peano curve format Ulm, HfG-Ar, BDSA, Hu P 09. 025.
in the 1956/57 Grundlehre. Assignment: Symmetry.
Note by Huff: “Fails symmetry test.”
59
I have called this type of assignment the “Single Element
Raster”.
60
A Peano curve assignment was given a different directive
in the 1955/56 Grundlehre, with handsome results. See
62
Lindinger, Hochschule fur Gestaltung Ulm, pp. 46–47. Examples from the 1955/56 Grundlehre are in the HfG-
Archiv. See Ulm Design: The Morality of Objects, Herbert
61
The directive of Exercise D “Black as a Color” and the Lindinger (ed.), Cambridge, MA, 1991, p. 47.
modified Peano curve format of Exercise G had been
63
combined in the same assignment in the 1956/57 Grundlehre. Maldonado, Max Bill, p. 18.
Maldonado’s directive at Carnegie to use two colors of the
same value enters the province of one of Albers’s most 64
Maldonado, “Concrete Art”, pp. 11, 13; see also Maldonado,
subtle interaction of color exercises (vanishing boundaries). Max Bill, p. 18.
Jorge E Calle, “Copan”. Basic Design

Studio of William S. Huff, Fall 1989,
SUNY at Buffalo. © HfG-Archiv/Museum
Ulm, HfG-Ar, BDSA, Hu P 04. 027.
Assignment: Conflicting Depth Cues.

Concrete artists, Vantongerloo and Bill, have suggested
that the route might be to overcome the limited figures.
To liquidate the figures, in a word, and to make the ground
vibrate to a maximum (by using subtle, non-figurative
elements) would be one of the ways.64
Undoubtedly, this ruminant can explain Maldonado’s

enthusiasm for Zillmann and Seitz stumbling upon
rasters as their solutions to “Ungenau durch Genau”. In
effect, Zillmann and Seitz had come up with a marvel-
ous way to “vibrate the ground”.
The companion assignment, “Genau durch Un-
genau”, baffled our Grundlehre class of 1956/57. As I
remember, no solutions were produced—at least, no
adequate ones. It was deemed a conceptually impos-
sible directive. Yet I learned in my recent contacts with
Dolf Zillmann that in the 1955/56 Grundlehre, which
preceded ours, Maldonado had judged Zillmann to Jeffrey D. Roos. Basic Design Studio
have had come up with “a perfect solution”: “One of of William S. Huff, Fall 1981, SUNY at
[his assignments] was to generate something precise Buffalo. © HfG-Archiv/Museum Ulm,
with patently imprecise means. I remember taking a HfG-Ar, BDSA, Hu P 09. 014.
piece of firewood, the end of which was shaped like a Assignment: Programmed Design.
slice from a round pie, and using it as a stamp, rotating (Editor's note: Original in color)
it around an imagined point to create a precise circle.
Whatever the value of the result, Tomás thought it to The principle calculation of the Maldonado peda
be a perfect solution to his assignment, and the piece gogy seems to have operated along two tracks: a stu-
received a disproportional amount of attention.”65 dent acquiring mastery of their visual acuity—through
Gestalt, in specific, and perception, in general; the
student’s exploration of and experimentation with for
mal syntax—augmented by symmetry theory, visual
topology, and other mathematical curiosities, such as
the Peano and Weierstrass curves. Many of the sources
of knowledge that Maldonado introduced were very
current in respect, at least, to the publication dates of
the texts at his disposal.
In 1949, while still in Argentina, he came across the
Peano and Weierstrass curves and the Sierpinski trian-
gle in a Spanish version of Hans Hahn’s “The Crisis in
Intuition”,66 later to be published in James R. Newman’s
World of Mathematics.67 He seriously entertained such
paradoxes of curve theory as potential features for his
Art Concret productions. In 1950, he was impressed by
a French pamphlet by M. A. Sainte-Laguë on topolo-
gy.68 Then, during his initial involvement with the HfG
Grundlehre, he acquired three key writings, Walther
Lietzmann’s Anschauliche Topologie, K. L. Wolf and D.
Kuhn’s Gestalt und Symmetrie, and A. S. Parchomenko’s
Was ist eine Kurve?69 Furthermore, aside from expand
ing his bookshelf with these topics, he kept current on
Gwen Herr, “Here’s Herr’s Herringbone”. most of these interests with the latest issues of Scien
Basic Design Studio of William S. Huff, tific American. By introducing these things in the con-
Fall 1969, Carnegie Mellon University text of visual training, Maldonado resolutely deviated
(CIT). © HfG-Archiv/Museum Ulm, HfG- from Albers’s injunction not to read books; rather,
Ar, BDSA, Hu P 04. 026. Maldonado suffused the Grundlehre with “operational,
Assignment: Conflicting Depth Cues. manipulable, real knowledge”.70

65
Letter from Dolf Zillmann, 19 September 2002.
66
Hans Hahn, “La crisis de la intuición”, in Crisis y
reconstrucción de las ciencias exacas, La Plata, 1936.
67
Hans Hahn, “The Crisis in Intuition”, in The World of
Mathematics, James R. Newman (ed.), vol. 3, New York, 1956,
1956–1976. An article in the same volumes, exemplary in the
eyes of Maldonado for its remarkable scientific lucidity, was
singled out as a rare reading assignment: Leonhard Euler,
“The Seven Bridges of Königsberg”, in World of Mathematics,
Vol. 1, Newman, pp. 573–580.
68
M. A. Sainte-Laguë, La topologie, Paris, 1949.
69
Walther Lietzmann, Anschauliche Topologie, Munich,
1955; K. L. Wolf and D. Kuhn, Gestalt und Symmetrie: Eine
Systematik der symmetrischen Körper, Tübingen, 1952; A.
S. Parchomenko, Was ist eine Kurve?, Berlin, 1957. A series
of lectures in Buenos Aires 1957 and 1958 on symmetry,
topology, perception theory, and semiotics prompted
Wolf and Kuhn’s text on symmetry to be translated into
Spanish: Forma y simetria: Una sistemática de los cuerpos James T. Mountain, Inversion with
simétricos, translated by Renate Leisse de Mertig and Mario Dilatative Motif. Basic Design Studio
H. Gradowczyk, Buenos Aires, 1959. of William S. Huff, Spring 1989, SUNY
70
Ockman, Architecture Culture, p. 299. HfG-Ar, BDSA, Hu P 06. 028.
Assignment: Programmed Design.
Note by Huff: “Marginal craft”.
Richard E. Stehlik, “four checkered

homeometric grids”. Basic Design Studio
of William S. Huff, Fall 1964, Carnegie
Mellon University (CIT). © HfG-Archiv/Mu-
seum Ulm, HfG-Ar, BDSA, 08. 009.
Assignment: Symmetry.
(Editor's note: Original in color)

It was implied, rather than stated, that an aesthetic Vordemberge-Gildewart had taken on the responsibility
could emerge from either track; the critical response of most of the visual training in the Grundkurs,75 the
to matters of perception and the visual concretization unified Grundkurs was abolished.76 There had already
(programming, so to say) of mathematical and geo- been experimentation in the latter years of the unified
metric abstractions.71 Indeed, that did seem to occur, Grundkurs to assign some tasks of an applied nature—
more than not—especially when the two tracks were ones differentiated according to a student’s elected
conducted in tandem. discipline.77 That was the beginning of putting into
practice one element of Maldonado’s straightforward,
v. A New Model, Partly Executed, Partly Not idealized curricular model, which mapped basic design
Maldonado was responsible for bringing a critical mass vis à vis applied design (the Design–non-applied/
of methodologists (in mathematics and in the soft sci- Design–applied model).
ences) into the HfG faculty, who eventually returned the
favor by taking over the administration of the school for
a two-and-a-half-academic-year period—June 1960 to
December 1962. While the design-oriented faculty did
finally regain the reins of the school from the method-
ologists,72 it must be concluded that this surreal inter-
lude did not have only negative consequences (one of
the negatives having been that the students analyzed
design problems with exquisite thoroughness, but did
not design any longer).73 In hindsight, the interruption Art historian Kenneth Frampton, an early HfG observer,
advanced the development of the school, not only by made the following report:
challenging the original Bill model (as Maldonado was
already doing), but by provoking Maldonado’s reevalua- It should be noted that the foundation course, or Grun
tion of his own emphasis on the role of methodology in dlehre,78 had been discontinued, after Maldonado had
design—that is, that methodology was running the risk been appointed as head of the industrial design depart-
of becoming “methodolatry”.74 ment with the reorganization of 1962. After this year, stu-
While Maldonado’s teaching in the Grundlehre was dents were channeled into one of the three departments
curtailed for a brief period, his interest in the basic from the very beginning; that is, from the first year they
design problem did not abate; there were new ideas entered directly into their chosen specialty, be it building,
to try. After academic year 1960/61, when Friedrich product design or visual communication.79
William Mc Adams, “Tri-Cellular Trade”.

Basic Design Studio of William S. Huff
Spring 1989, SUNY at Buffalo. © HfG-Archiv/
Museum Ulm, HfG-Ar, BDSA, Hu P 06. 018.
Assignment: Parquet Deformations

Thomas Breen, “Axonometry Cubed”. Basic De-
sign Studio of William S. Huff, Fall 1993, SUNY at
Buffalo. © HfG-Archiv/Museum Ulm, HfG-Ar, BDSA,
Hu P 02. 019. Assignment: Programmed Design:
Syngenometry, Note by William Huff: “Axonome-
try Cubed—one isometric view of a cube at the
center, three axes of dimetric views, three fields
of trimetric views, and three borders of orthogonal
views.” (Editor's note: Original in color)
learn to use color!”—meaning a shattering of the obsessively

71
Many geometers (H. von Baravalle was not one of them, insistent use of only black, white, and gray (plus the little red
K. L. Wolf was) are often unaware of many of the visual dot) by the students in Product Form and in Building, since
ramifications of the equations with which they work. At the Visual Communication, by its nature, could not have done
same time, these geometers make aesthetic judgments in very well without the use of color. It did not work. It took an
respect to the equations themselves. event in 1963 to do that: a lecture by the GK Industrial Design
Association of Tokyo, led by Yoshio Nishimoto.
72
“On December 15, 1962, a new constitution for the HfG
76
came into effect. On December 20, 1962, Otl Aicher was Some confusion is found in various articles about the
elected Director for the academic years 1962/63 and 1963/64. year in which this took place. Citing 1961/62 are: Karl-Achim
Tomás Maldonado was elected Vice-Director.” ulm 7, January Czemper “Die pädagogischen Ziele der Grundlehre”, output
1963, p. 2. 6/7, 1961, pp. 1–10; Hans Roericht, HfG-Synopsis, Ulm, 1982.
Tomás (Milan, [2002]); citing 1962/63 are: Kenneth Frampton,
73
Maldonado, form + zweck, p. 20. “Apropos Ulm: Curriculum and Critical Thinking”, Oppositions
3, May 1974, p. 27; Lindinger, Hochschule fur Gestaltung Ulm,
74
See Tomás Maldonado and Gui Bonsiepe, “Science and p. 33.
Design”, ulm 10/11, May 1964, p. 10. Also see an earlier
77
expression of these realizations: “There are those who Maldonado, form + zweck, p. 19.
believe that all design problems can only be solved by
78
mathematical formalization. […] The designers of [this] Vorkurs, Grundlehre, and basic design have often been
category seem to have a much more intransigent opinion used synonymously in articles by numerous writers
on methods than the scientist themselves. The scientists (including, I admit, myself). In Frampton’s article, there is a
sometimes have doubts about scientific methods, but distinct clash of these terms that allows misinterpretation
the designer of the kind referred to—never”, Tomás to gain hold.
Maldonado, “Opinions: Preliminary Note”, ulm 6, October
79
1962, p. 3. This was based on “tables that Gunter Schmitz made
available to [a] Montreal seminar” in February 1968.
75
With Vordemberge-Gildewart as his instrument, Horst Frampton, Oppositions, pp. 27–28. Actually, there were still
Rittel, leader of the insurgent group, made known to me four departments at that time; it is not known (at the time
(during my visit in June 1960) that “now the HfG students will of this article) how beginning students were handled in the
Information Department in 1961/62.

Though Frampton wrote in this same passage (but advanced years was never implemented despite his
well removed from the above) that “for a student of argument that upper-level students liked to return to
industrialized building […] some 45 percent of his time purely formal assignments from time to time: “They
would have been spent in basic design”, this falls short have nostalgie for it.”
of Frampton’s making clear that only the Grundkurs as Albers had championed this idea, as well. On the
a unified entity had been abolished, not basic design one hand, he welcomed former students who re-
studies as such.80 turned to his basic design classes for a rejuvenescent
To reiterate, what this curricular reform did, under fix;89 and in a variation of this position, “it was most
Maldonado’s guidance, was to mandate that each rewarding” to him “that many graduate students from
department take charge of its own Grundlehre cur various fields of study […] enrolled in basic art courses
riculum,81 and each could tailor its own first-year tasks officially assigned to beginners.”90 But this was catch-
(both its specific formal tasks and its specific applied as-catch-can. In fact, one should look back to the
tasks). In an oversimplification of this, it could be said Bauhaus as the design school where an intermingling
that the three-dimensional assignment types of the of the formal (basic design) and the practical (applied
originally unified introductory courses were carried design) was actually carried on throughout the whole
over into Product Form82 and the two-dimensional curriculum. Shortcomings of the workshop system
assignment types into Visual Communication.83 Under have been noted. But the continuance of formal is
“3D Non-Functional Projects”,84 among a number of sues in advanced training was assured by the work-
formal projects, one extensive study of nonorientable shop system as laid out:
surfaces (square, circular, or triangular sheets, some
with round holes, were given simple cuts, twisted, The Bauhaus […] ruled (1) that every [workshop student]91
reattached) was conducted during one first-year is taught by two masters, a craftsman and an artist, who
session in Product Form. At the same time, in accord work in close cooperation; (2) that instruction in crafts
with Maldonado’s basic design/applied design model, and in the theory of form are fundamental: no [workshop
rather elementary applied design tasks (e.g., simple student] can be excused from either.92
tools or everyday utensils, rudimentary graphic pieces,
or clever building components) were assigned in the The lines that separate the strictly formal from the
first-year courses of the respective departments. In practical might, at times, have been blurred; but the
Product, for instance, a lamp and an assortment of possibility to revisit formal issues on a continuing ba-
office implements were assigned.85 sis was built into these two rules.93
In Building, among the formal exercises that served In the 1990s at the University at Buffalo, I pressed
this department’s purposes,86 one of Josef Albers’s for my withdrawal from our department’s required
Bauhaus assignments was given new life pre-architectural design program94 and, in turn, re-
—paper-folding, which Albers had reprised for the first quested that I be permitted to offer my basic de-
Grundkurs of 1953/54. Albers had permitted cuts to be sign course on an elective basis to students in the
made in the paper in earlier days of his assignment,87 graduate program.95 This type of design instruction
but later forbade them. One instructor in Building, was entirely new to a great number of our graduate
perhaps Herbert Ohl, reintroduced the slitting of the students, many of whom came from abroad. I was
paper (usually one- or two-ply Bristol board) with small not supported by much of my faculty (though certain
repetitive cuts in the process of paper-folding. eminent, but not “in-charge” colleagues gave my mis-
I myself had a part in continuing the Maldonado sion their morale-boosting backing); and the depart-
model of visual training in the Visual Communication ment’s administration, though allowing it, nonetheless
Department from 1963 until the school’s closing in constantly thwarted it. At any rate, I can claim success
1968.88 I shared these tasks with Maldonado himself, in the trial: evidenced in small part by the splendid
Herbert Lindinger, and Tomás Gonda. (After one year of results of the students and in large part by how the
protest and noncompliance, Film—a subdepartment of students approached the opportunity with a maturity
Visual Communication, and a new one at that—opted that undergraduate students seldom brought to the
out of Visual Communication’s compartmentalized classroom.
Grundkurs.) Maldonado, Zeischegg, and Bonsiepe, and
possibly others, developed the compartmentalized vi. An Elaboration of the Maldonado Model
Grundkurs material for Product Form. I was 29 in 1956, when I found what I had been seeking
Though the breakup of the unified Grundlehre in design—perhaps, from the very start of my entry into
was carried out in 1961, Maldonado’s idealized curric- that arena. I was 33 when I dared to meet the chal
ular diagram was only partly implemented. That is to lenge of teaching my long-sought chimera—and to
say, a systematic continuance of basic design in the make it into a worthy adult occupation. Upon returning

“Office Implements”, Industrial Design Department, first
80
Had the HfG abandoned basic design under the year 1963/64, taught by Tomás Maldonado; “Microphone”,
reorganization of 1961/62, my article, “An Argument for Industrial Design Department, first year 1962/63, taught by
Basic Design”, would scarcely have been given space in Tomás Maldonado, ulm 12/13, March 1965, pp. 48–53.
the HfG journal, ulm 12/13, March 1965, pp. 25–38, nor
86
would there have been reason for the HfG to invite me “The department described the instruction for the incoming
to conduct five annual guest courses in my modified- students as ‘Einführung in die besondere Teilproblematik des
Maldonado version of basic design from 1963 to the closing industrialisierten Bauens. Aufgaben aus begrenzten wichtigen
of the school. Furthermore, the ulm journals, from 1963 on, Teilbereichen werden analytisch erfasst, verglichen und kon-
made a special point to publish teaching results of each struktiv neu gestaltet.’” Gunter Schmitz, email from 9 Sep-
department’s first year. tember 2002. See also “Basic Design for Architects”, Building
Department, first year 1966/67, taught by Gunter Schmitz,
81
Maldonado considers this to be an important ulm 19/20, August 1967), pp. 41–46.
accomplishment: “We took this approach one step further
87
and the basic curriculum was structured differently for Wingler, Bauhaus, pp. 431, 433, 435.
each department. That was the end of the idea that there
88
was one curriculum that could serve as the basis, a basic In 1963, I also taught the first year of Building.
curriculum that was the foundation for all designing
89
activities.” Maldonado, form + zweck, p. 19. My one concern Beckmann, in Neumann, Bauhaus, pp. 196–197.
is for the immature students who approach design with very
90
little formal grounding from their previous education. Under Albers, Search, p. 25.
the unified Grundlehre, these students had the advantage
91
of an eye-opening year before committing their futures to a The text here reads “apprentice and journeyman”. For an
particular design specialty. explanation of this ranking system, see Herbert Bayer, Walter
Gropius, and Ise Gropius (eds.), Bauhaus 1919–1928, New York,
82
“In connection with the preparation of an introductory 1938, p. 26.
course in three-dimensional design for the first study-year
92
of the Industrial Design Department, Walter Zeischegg made Bayer, Gropius, and Gropius, Bauhaus, p. 25.
a series of studies in the field of experimental geometry.”
93
See “Experiments with Regular Solids”, ulm 7, January 1963, That most of the “artist” masters at the Bauhaus had
pp. 11–12. been an integral part of the Modern Movement undoubtedly
put the handling of formal issues in the workshops on the
83
“Visual Communication Department: 1. Study-year”, taught highest of planes.
by Herbert Lindinger and William S. Huff, ulm 17/18, June
94
1966, pp. 40–45. My basic design course in this program was autonomous,
so it was not an issue of any need of independence that
84
“3D Non-Functional Projects”, Industrial Design spurred my request to shift gears—rather, it was my desire to
Department, first year 1965/66, taught by Gui Bonsiepe, put the Maldonado basic/applied design model to the test.
ulm 17/18, June 1966, pp. 21–34.
95
I can personally recommend this route; for, after all, this
85
“Warning and Repair Lamp for Cars”, Industrial Design is the one that I took: I attended the visual course at the HfG
Department, first year 1963/64, taught by Herbert Lindinger; after obtaining a degree in Architecture from Yale. I was well

to the States from Ulm in 1957, I had no intention to days, has also worked in this arena at Milan’s Politecni-
teach anything at all, much less to teach basic design. co under the aegis of Maldonado. Sudhakar Nadkarni
After a year’s immersion in the fundamentals of design helped to found and then headed the Industrial Design
at the Hochschule, I returned to the pursuit of an archi- Centre, IIT, in Mumbai, where he set up much of the in-
tectural career and had the good fortune to be taken troductory course, as well as other parts of the design
into the office of Louis Kahn in Philadelphia. I also program, along HfG lines.98
lobbied for the adoption of the HfG model of basic de In my teaching of basic design for most of 38 years
sign—with an emphasis, of course, on the Maldonado (1960–1998), I did, then, unapologetically adopt the
curriculum—by at least one leading American school of HfG Grundlehre model, as it had struck me: mainly
architecture. I considered Yale’s Department of Archi- the Maldonado portion, potent in its approach to
tecture to be, perhaps, the most likely of loci for this visual training, but also key segments from Helene
to come off—in part, because now I could point to an Nonné-Schmidt and Hermann von Baravalle99—all
authentic and tested alternative to the poor shadow of eye-opening experiences; a dash of Albers’s interaction
a Bauhaus-style introductory course that I had been of color was included in it too. My students were assid-
obliged to take eight years earlier; in part, because uously primed about the various origins of this course
the Bauhaus’s Albers, who had arrived at Yale after of study.
my experience with that most wanting of introductory From my experience at Ulm, I took note of a differ
courses, was still heading the Department of Art and entiation between two types of assignments—exer-
had a skilled staff under him.96 My cause was politely cises, tightly prescribed in order to develop skills and
heard, but went nowhere. to reinforce particulars of information, and projects,
At one point, Kahn suggested that I serve as his loosely stipulated along thematic lines, but formulated
assistant for a new Urban Design studio at the Univer- as challenges to ingenuity and creativity.100 Over time,
sity of Pennsylvania. I told him that I feared that I would I edited and modified various HfG exercises in minor
be more the student than the assistant in a studio of ways to suit my goals for my students. At the same
this unfamiliar territory. I did propose, however, that I time, though the projects that I developed inescapably
could confidently assist the faculty member who was had HfG genes, I elaborated from where Maldonado
in charge of Penn’s introductory course. Kahn respond- had left off—much of which was responding to signif-
ed: “No, you should not assist anyone in an introducto- icant breakthroughs of the students, who themselves
ry course. You are the one to set up a new introductory were not always cognizant of what they had unwittingly
course.” I took this remark seriously and immediately wrought.
began to outline a course along the lines of my HfG Aside from three of Maldonado’s classic exercises
Grundlehre experience. As it turned out, Penn’s dean that address perception, there was only one project
was not prepared for this proposal, even with Kahn’s that I have expressly adapted from Maldonado: Sym
weight behind it; but upon learning of my outline of a metry. His assignment statement to our class was
basic design course, Carnegie Institute of Technolo- rather loose and went something to this effect: “Make a
gy’s head of Architecture, Paul Schweikher,97 a friend design from the combination of two or more of Wolf’s101
of both the Alberses and Mies van der Rohe, lost no 13 isometric and/or homeometric operations of sym-
time in latching onto what he considered a most timely metry.”102 In time, upon my review of the ramifications
opportunity to further the reinvigoration of his depart- of Wolf’s lower levels of symmetry (syngenometry and
ment’s program. katametry), this project was retitled The Programmed
Short of any formal sanction, I deemed that I had Design. Simple programs that produced complex
Tomás Maldonado’s and other HfG faculty members’ structures were abetted: structures, whose invariant
tacit consent to carry on an HfG-type Grundlehre properties (akin to the secrets of natural law) were not
instruction in Pittsburgh—for suddenly, sheer chance obviously discernible—even though the design, from
had veered me far off the course of my original intent, beginning to end, had been wholly determined, once
merely to spread the word in the States about a con- the student had (free of choice) selected the motif
sequent course of a progressive school of design. I had element(s) and rule(s) of operation. An aesthetic (of no
stuck out my neck; my head was on the platter. particular preconception) was always the goal—but
As it turned out, in the course of time, two of my that precluded the indulgence of any arbitrary inter-
former HfG classmates from 1956/57 have also con- vention whatsoever, in order to bring about a more
ducted their versions of the Maldonado Grundlehre: desirable outcome. Only an across-the-board alteration
Shutaro Mukai, as founder and head of the Department of the program would do: that is, the uniform alteration
of Science of Design at the Musashino Art University of the element(s) or of the rule(s) or of both.
in Tokyo for 35 years, and Thomas Dawo at Krefeld. The Parquet Deformation103 grew out of Maldonado’s
Giovanni Anceschi, whom I did not know in his HfG Parkettierung Aufgabe104—which few, if any, other than

he taught in the Art Department’s basic design course.
Luis Perelman, “Étude 3 of 3”. Basic
98
Design Studio of William S. Huff, May 1961, Nadkarni is now head of the Department of De-
Carnegie Mellon University (CIT). © HfG-Ar- sign at the Indian Institute of Technology (IIT) in
chiv/Museum Ulm, HfG-Ar, BDSA, Hu P 04. 042. Guwahati, India.
Assignment: Raster—Lineal Halftone.
99
Both Nonné-Schmidt and von Baravalle were “a heritage
of Max Bill’s Grundkurs”, Maldonado, 3 July 2003.
100
prepared to receive this fundamental training. “Discovery and Invention, the Criteria of Creativeness”,
Albers, Search, p. 32.
96
Robert Engman, Erwin Hauer (who designed a whole
101
series of modular screens, based admittedly on the 1947/48 Wolf and Kuhn, Gestalt.
“lattice-oriented shell surface” by the HfG’s Walter Zeischegg,
102
whom Hauer had known at the Vienna Academy of Fine Arts. In 1965, I received a grant from the US government
See ulm 14/15/16 [December 1965], p. 41), Norman Ives, Neil for a proposal to present Wolf’s 13 symmetry operations
Welliver, and Sewell Sillman. I was too naïve at that time to and other matters of symmetry in a visual format. A
realize how threatening such a proposal might have been series of booklets, designed by Tomás Gonda, were put
considered among the Albers group. In respect to the faculty out under the title: Symmetry: An Appreciation of Its
of Architecture, many probably thought that Albers was Presence in Man’s Consciousness.
entirely enough of a Bauhaus force for any American school
103
to handle. Douglas R. Hofstadter, “Parquet Deformations: Patterns
of Tiles That Shift Gradually in One Direction”, in “Metamag-
97
Both Schweikher and Kahn were the design instructors of ical Themas”, Scientific American 249, July 1983, pp. 14–20.
my “thesis” year at Yale. Prior to Schweikher’s heading the Also see Douglas R. Hofstadter, “Parquet Deformations: A
Department of Architecture at Carnegie, he headed Archi- Subtle, Intricate Art Form”, chapter in Metamagical Themas:
tecture at Yale; and it was at his request that Albers set up Questing for the Essence of Mind and Pattern, New York,
the course “Structural Organization” for the Architectural 1985, pp. 191–212.
Department which was in name only any different from what
104
See Lietzmann, Topologie.

myself, executed in our class of 1956/57. The recognized symmetry was not possible before Fourfold Mirror-Ro
potential of innumerable variants of any given parquet tation Symmetry,114 a most uncommon structure in
(an array of congruent tiling) led to connecting vari- nature,115 had been grasped. Remarkable objects, based
ant tilings to one another serially through continuous on the properties of twofold, fourfold, and sixfold
deformation105—a device that the morphologist D’Arcy mirror-rotation, have been given shape by many hands
Thompson appropriated from Dürer, the artist, and and heads of my students; these abstract objects
developed for his analysis of evolutionarily related spe- often exhibit a monumental quality. The Trisection of
cies. As a design device, the operation of continuous the Cube (into congruent solids) emerged analogously
deformation permeates the spatiality of a planar image, from parquetry (or congruent tiling of the plane). While
unraveling it through temporality106—superbly estab- three-dimensional congruent sectioning of any com-
lishing that the dimension of time is not the exclusive plexity was simmering in my mind, I encountered a
prerogative of music and drama. trove of fine plaster models of congruently sectioned
The Color-Raster came out of Maldonado’s latent solids in the basic design studio of Andrzei Pawlowski,
interest in rasters, which was stimulated by solutions Departmental Dean of Industrial Design,116 whom I had
to an assignment in the 1955/56 Grundlehre. At first, visited at Krakow’s Akademie Sztuk Pieknych in 1970. At
my students worked with halftones and three-color first, my students continued to investigate the congru-
rasters along the lines of black-and-white and color ent sectioning of any regular or feasible semi-regular
clichés; but in succeeding years, the assignment solid; but the assignment settled into the trisectioning
eventually culminated in repeated demonstrations that of the cube, when our attention was drawn to an error
pigments, on a dark ground (or with no ground show- of omission in Martin Gardner’s regular column for Sci
ing between the raster elements), work the way that entific American in 1980117—suggesting a topic that was
spectral light works: RGB pigments in particular will wide open to exploration.
produce a whole spectrum of colors through “mixing in In my visual training courses,118 mostly for archi-
the eye”.107 (Splatters of a limited range of reds, mingled tectural design students, I rigorously addressed only
with splatters of a limited range of greens in the favor formal issues—letting those issues be conjoined
able context, will produce a rich yellow.) The brief for only to other abstract disciplines, in accordance with
Conflicting Depth Cues calls for pitting one cue against Maldonado’ insightful interdisciplinary reach. While
another to create contradictory perceptions. Since references to symbols, to associative emotions or
Albers108 and Escher had already saturated both the memories, or to cultural overtones can well be the
art market of the elite and the bazaar of pop with their context of other pedagogic endeavors, those very mat-
subtle and not-so-subtle tricks of the eye (and brain) ters are kept, as far as is possible, from intruding into
created, on the whole, through linear perspective, I had and watering down the basic design product. It is, then,
my students concentrate on the potentials of the other perhaps not surprising that many who engage in the
depth cues to elicit captivating visual contradictions. practical disciplines of design, in both academia and
The Figurative Ground109 project engaged not only the the professions, find even the best examples of the
Gestalt principle of figure-ground, but the manipula- product to be useless—but not pithily “useless” in the
tion of texture110 and sfumato,111 and the topology of the way that Anni Albers had meant it; others think them
maze. By juxtaposing gradients of texture or sfumato, exotic; but worst of all, perhaps, many of those who
figuration can be made to appear abruptly and to are uninitiated in respect to involvement in design, find
disappear just as abruptly. When carefully traced, all them, if not baffling, to be decorative.
figurative figments will prove to be integral parts of Recently, at my urging, Claudio Guerri of the Uni
their very own, singular ground. versity of Buenos Aires119 ran a semiotic analysis on
Regarding assignments of the third dimension, the question of whether basic design (as I see it) can
which I presented on an intermittent basis, I continued be regarded as a self-contained discipline.120 This
Maldonado’s examination of Nonorientable Surfaces, exercise, probing basic design’s body parts, as it were,
treated by Bill for many years as a major sculptural foil. seemed to be warranted at this juncture—more than
I posed one question: “Can symmetry be introduced eight decades after basic design’s legendary inception
into any topological variant of the Möbius band?” Yes, in Weimar. In the course of our joint venture, Guerri
the answer promptly turned up from my students’ first remarked that the product of basic design and the kind
tests; the potential of twofold rotational symmetry is of tasks that produces it are “hermetic” in the eyes of
lodged in such a configuration.112 The elusive principle many design faculties—mostly for the reason that the
of Twofold Mirror-Rotation Symmetry, whose clar- need for basic design instruction has not been made
ification I had failed to grasp from Maldonado and clear enough. Agreeing with this, my response is that
which was poorly depicted by Wolf,113 had me looking I myself have been consistent in my argument that I
into crystallography, where I learned that a theory of have dealt all along, as Albers put it, purely with “form

105 113
From “continuous transformation”, “The Theory of Trans- Wolf and Kuhn, Gestalt.
formation, or the Comparison of Related Forms”, D’Arcy
114
Wentworth Thompson, On Growth and Form, 2nd ed., Vol. 2, Though not classified as the 32nd Crystal Class, it was
Cambridge, 1959, p. 1086 (page numbers from the reprinted the 32nd of the 32 Crystal Classes to be discovered in the
edition). mid-nineteenth century.
106 115
William S. Huff, “The Landscape Handscroll and the Molecules have been found to exhibit this trait, but enti-
Parquet Deformation”, in Katachi U Symmetry, Tokyo, 1996, ties with this property have not been a part of the scenery—
pp. 307–314. at the scale of our familiar visual world.
107 116
“Optical mixing”, Josef Albers, Interaction of Color, New Pawlowski’s curriculum at the Academy in Krakow was
Haven, 1963, p. 33. influenced by that of the HfG.
108 117
“‘Optical illusion’ should be replaced with ‘optical decep- Martin Gardner, “Mathematical Games”, Scientific American
tion’”, Albers, Search, p. 21. 243, September and October 1980.
109 118
Upon the publication of Tomás Maldonado’s manuscript, There were three different one-semester formats, under
“Concrete Art and the Problem of the Unlimited” (Ramona, a variety of different names, that covered (1) syntax in 2D, (2)
2003), I learned that Maldonado had termed such an effect perception in 2D, (3) syntax and perception in 3D.
“vibrating the ground”—a ground that had no discrete figures.
119
Professor Titular of Morfologia (a version of basic design—
110
César Jannello, “Texture as a Visual Phenomenon”, Archi- i.e., the study of all formal issues) at the Faculty of
tectural Design 33, August 1963, pp. 394–396. Architecture.
111 120
Max Bill’s Unbegrenzt und begrenzt/Unlimited and Limited, The methodology of the analysis is realized by the
1947. “semiotic nonagon”, which is an operative triadic model
developed by Guerri from C. S. Peirce’s semiotic construct.
112
In many of its usual depictions, the Möbius band appears See Claudio Guerri, “Gebaute Zeichen: Die Semiotik der
to be asymmetrical. In fact, the 180° twist that is given to Architektur”, in Die Welt als Zeichen und Hypothese:
a flat strip before its two ends are joined should be a dead Perspektiven des semiotischen Pragmatismus von Charles
giveaway to the potential of capturing the twofold rotational S. Peirce, Uwe Wirth (ed.), Frankfurt am Main, 2000,
property in specific rigid pieces. pp. 375–389.
Maurizio Sabini. Basic Design Studio

Assignment: Parquet—3 Variants.

that exists for its own sake”,121 in the confidence that
immersion in formal content alone—devoid of cultural,
referential, or associative, even physically characterized,
overtones122—can unleash the sensory capacity. I speak
now for myself: My overarching objective has been to
elevate, without inordinate distraction, my students’
mastery of their own innate aesthetic acuities.
John Weiler. Basic Design Studio of William

S. Huff, Fall 1962, Carnegie Mellon Uni
versity (CIT). © HfG-Archiv/Museum Ulm,
Assignment: Symmetry (Version 3 of 3)
(Editor's note: Original orientation is 90°
counter clockwise; Original in color).

architecture to first-year architectural students. The
121
“‘Abstracting,’ he wrote, ‘is the essential function of two courses were markedly different. Basic architecture,
the Human Spirit. Abstract Art is the purest art: it strives instructed by a team, was authored by Alvaro Malo,
most intensely toward the spiritual. Abstract Art is Art in Richard Cordts, Frances Downing, and myself. Five
its beginning and is the Art of the Future.’ He saw abstract major “fragments” of architecture, five provinces of
art as a natural part of an historical development toward performance, were identified: the anthropometric and the
a ‘pure art,’ as part of a move away from ‘imitative aims’ ergonomic (the body’s relation to itself and to objects in
that pictured ‘nature, stories or sentiments’ toward ‘non- space); the constructional (the smaller and larger joinery
representative form, form that exists for its own sake, of specific materials); the formal or syntactic (emphasizing
namely for form reasons.’” Mary Emma Harris, The Arts at the void of space and involving aesthetic judgment), the
Black Mountain College, Cambridge, MA, 1987, p. 13. environmental (passive control of natural elements—light,
acoustics, climate); the contextual (urban and pastoral
122
For most of the 1980s, in addition to teaching basic landscape). Each fragment was addressed by tasks that
design to pre-architectural candidates, I taught basic strictly minimized, during targeted consideration, the
involvement of the other four fragments.
Editorial note
This text was published in Ulmer Modelle– Modelle Nach Ulm, Hochschule für Gestaltung
Ulm 1953–1968, Ulmer Museum/HfG-Archiv (ed.), Hatje Cantz, Ostfildern-Ruit, 2003.
The article is reprinted here with kind permission of William S. Huff and the HfG Ulm
Archive; accompanied by new imagery made available by Dr. Martin Mäntele and selected
by Werner Van Hoeydonck.
William S. Huff attained two degrees from Yale University: Bachelor of Arts, 1949; Master
of Architecture, 1952. He was awarded a Fulbright Fellowship (1956) to study at the
Hochschule für Gestaltung (HfG) in Ulm, Germany, where he later returned as a Guest
Teacher (1953–1968). He held academic positions at the Departments of Architecture of
Carnegie-Mellon University from 1960 to 1972 and of The State University of New York at
Buffalo from 1974 to 1998, when he was elected Professor Emeritus. In 2008, he received
a Doctor Honoris Causa from the Ministry of Science and Education of Ukraine, National
University Lvivska Polytekhnyka, Institute of Architecture. He has written on his principal
pedagogic discipline, basic design, symmetry, topology and color; with the aid of Claudio
Guerri‘s analytic Semiotic Nonagon, he resolved the age-old problem of a theory of the
logic use of color (www.academia.edu/16332326/). Huff has written on Tomás Maldonado,
under whom he studied; on Louis Kahn, under whom he also studied and in whose
architectural office he worked from 1958 to 1962; and on American artist S. H. Crone. In
1989, Huff was a founding member of the International Society for the Interdisciplinary
Study of Symmetry, ISIS-Sym, at Budapest, to which he was elected Honorary President
in 2007. He was a honorary member of a number of other international interdisciplinary
organizations; he was elected Honorary Member of SEMA (2003) and was an International
Fellow and Founding Supporter of Japan’s Society for the Science of Design Studies (1998).
His studio assignment, the parquet deformation, was recognized by Douglas Hofstadter in
Scientific American (July 1983).

Geometry of Structures
and Its Philosophical
Aesthetic Background
Cornelie Leopold
In this chapter, the design approach at the Ulm of philosophy and theory of science at the University of
School of Design, which is based on the geometry Stuttgart. Bense developed a new definition of aesthet-
of structures and transformations, will be described ics1 by starting with Hegel’s description of art. In that
and related to the background of the philosophical definition, the aesthetic state of an object is related to
aesthetic of the school, introduced by the philoso- distributions of elements or representations of order in
pher Max Bense. Max Bill’s mathematical approach the meaning of arrangements. Elisabeth Walther, pro-
to art and design coincided fruitfully with the rational fessor at the University of Stuttgart and lecturer at HfG
philosophy of Max Bense by focusing on the human Ulm, described the role of this new definition:
relationship to a technological world preconditioned
by rationality and methodological thinking. Mathe- Aesthetics, as Bense brings it into play, is the principle
matical methods were introduced into the structural of order par excellence. Aesthetics is order, and order on
analysis and creative design processes. Based on this the other hand is describable by mathematics. There-
foundation, aesthetics was developed into an infor- fore, aesthetics is important as structuring the world for
mational aesthetics derived from information theory techniques as well as architecture, literature, etc., for all
and semiotics. The redundancies arising from rules what will be created. Whenever we take something out
such as symmetries were brought into relation with of the chaos of existing and assemble it new, we need an
innovation; random and chaotic states were instigat- aesthetic foundation.2
ing by breaking the preset rules. This was the begin-
ning of the first digital design experiments. Research This background of a new understanding of aesthetics
on different levels of symmetry formed the basis for influenced the design approach at HfG Ulm. Bense’s
William S. Huff’s programmed design. The roots of his aesthetics was in close relationship to Max Bill’s math
parquet deformation assignments can be found in the ematical foundation of art and design, as Bill expressed
structural approach at the Ulm School of Design. in his essays.3
1. Design Approach Now in every work of art the basis of its composition is
at the Ulm School of Design geometry or in other words the means of determining
The Ulm School of Design (HfG–Hochschule für mutual relationship of its component parts either on
Gestaltung) existed for only short a time, between plane or in space. Thus, just as mathematics provides us
1953 and 1968, but attracted students and professors with a primary method of cognition, and can therefore
from all over the world who later, as professors and enable us to apprehend our physical surroundings, so, too,
practitioners, spread the school’s ideas and concepts some of its basic elements will furnish us with laws to
throughout many countries. One of the school’s most appraise the interactions of separate objects, or groups of
important foundations was thinking about structures objects, one to another.4
in relation to mathematical and cybernetic structures
as a fundament for design methods. Design had been Max Bill, the founding director of HfG Ulm, found in Max
based on philosophical reflections and theories, espe- Bense a guest professor for fruitful discussions and
cially through the involvement of the philosopher and interactions in the field of aesthetics between theory
science theorist Max Bense, who had been professor and practice. In 1965, at the opening of Max Bill’s

exhibition in Esslingen, Bense stated, “By coming engaged Helene Nonné-Schmidt, Walter Peterhans,
across Max Bill, I came across a kind of art that could Josef Albers, and Johannes Itten as guest teachers for
be safely regarded as an object, he provided aesthetic the first course, which started in August 1953. Geomet
objects for the theory, that could be examined.”5 rical transformations like symmetries and proportional
The students at HfG Ulm also experienced these transformations were important elements in the Draw-
interrelationships between design theory and practice ing and Paper Folding modules taught by Albers in the
as examples for developing a new information aes 1953/54 and 1955 foundation courses. Max Bill did not
thetics based on order and chaos and redundancy and teach the foundation course himself but selected the
innovation. Principles of order as an aesthetic founda- teachers and offered his critiques of some tasks.
tion can be looked for in the geometry of tessellations, The design of the City of Ulm’s pavilion for the
patterns, their spatial variants and the mathematics of Baden-Württemberg state exhibition in Stuttgart, 1955,
symmetries, developed as a theory of transformations. is an example of his architectural design being based
Bense6 noted that this corresponds to Paul Valéry’s on the idea of geometric transformation and pattern.
statement7 that compared with the role of mathemat- The pavilion was designed by Max Bill, but the graphi-
ics in the sciences, patterns and ornaments—orna- cal appearance of the pavilion was worked out by the
mental drawing—are fundamental to art. students of the visual communication department of
Together with various guest professors, Max Bill HfG Ulm under the guidance of Otl Aicher and Friedrich
and, later, Tomás Maldonado created a curriculum at Vordemberge-Gildewart. The design is based on a
the Ulm School of Design with mathematical ap fourfold rotation of the fundamental cell module, as
proaches in design and architecture.8 shown for the top view of the pavilion in Figure 1. The
design is based on the classical development of orna-
2. Geometric Structural Background ments based on the root 2 system, which can be also
A fundamental geometric background was taught in found in HfG students’ drawings.10
the foundation courses at HfG Ulm. In the first years, In our 2012 “Structure – Sculpture” summer
from 1953 to 1958, the foundation course was oriented school in Buenos Aires, we also analyzed the spatial
around the experiences of the Bauhaus. It concentrated structure that can be produced by a spatial ele-
on a visual training of eye and hand, freehand drawing, ment after a fourfold rotation. Figure 2 shows the 3D
and experiments with materials. Max Bill, the first analysis and a physical 3D model of the Ulm Pavilion,
rector of HfG, himself a former student at the Bauhaus, created by the summer school participants.
Fig. 1: Development of
ornaments based on root 2
system and top view of the
Ulm Pavilion.
Fig. 2: Spatial element of

Ulm Pavilion and result after
rotation, physical model.11
92 Geometry of Structures and Its Philosophical Aesthetic Background Cornelie Leopold

Fig. 3: Spiral arrangement
in the geometric series of
squares, regular hexagons,
and octagons.12
Fig. 4: Geometric series

of rotated squares, spiral
arrangement, and spiral
arrangement of square
rings.13
Fig. 5: Creating parabola by moving a right angle,

student Dominique Gilliard, 1955/5614 and
Limaçon of Pascal as inversion of a hyperbola,
student Hermann Edel, 1956.15
More details about the importance of structural of Waldorf education, Baravalle established a dynamic
thinking at the Ulm School of Design, especially by understanding of geometry based on movements of
Max Bill’s design approach and Max Bense’s philoso- points and lines. His concept of dynamic geometry
phy and aesthetics, have been described by Hermann complemented the Albers action-oriented way of
Edel,16 who studied at HfG Ulm in the Department teaching. Baravalle described his dynamic geome-
of Architecture between 1956 and 1960. Later, from try concept in his book Geometrie als Sprache der
1963 until 1974, Hermann Edel and Max Bill worked Formen,17 published in 1957. How fundamental the
together on modular building systems in Darmstadt, geometric series of squares are when developed as
Germany. a root 2 system (as shown in Figure 1) is illustrated in
Figure 3, where Baravalle creates spiral arrangements
3. Hermann von Baravalle’s inside the square system by black and white fillings.
Dynamic Geometry He constructed corresponding spiral arrangements
After delivering some guest lectures in the early years inside the regular hexagon and octagon.
of HfG Ulm, Hermann von Baravalle was assigned Baravalle’s dynamic understanding of geometry is
to teach constructive geometry (1955–1959) in the perfectly illustrated in Figures 4 and 5. These courses
foundation course. Having strong roots in the tradition must have made a deep impression on William S. Huff,

Fig. 6: Two examples of the Baravalle-
Kino by Hermann Edel, hyperbola (left)
and six ellipses in different sizes and
rotated (right).
since many of his students’ works take this same dlehre. In the third year, 1955/1956, he took on teaching
dynamic approach. the Grundlehre. He carried on what one might call the
Hermann Edel was especially impressed by a purification and renovation of the Bauhaus direction. He
performance by Baravalle at HfG in 1959. Lines drawn preserved what Albers had done at the Bauhaus. But at
on slides were projected on a string cylinder. A straight the same time, he introduced something more: He made
line, for example, resulted in an ellipse on the string it interdisciplinary. He brought in some other subjects,
cylinder. Then the slide projector was moved, which such as symmetry and topology and a good dose of Ge-
produced changes in the curves on the cylinder. stalt theory. He introduced information about these sub-
Hermann Edel recreated this performance in 1959, jects into the course, which Albers had not done. […] But
with Baravalle’s consent. In 2013, Edel presented a Maldonado was a person who wanted you to know where
redesigned “Baravalle-Kino” (Figure 6).18 This dynam- the different parts of these disciplines came from.19
ic approach to geometry and experimentation using
transformational concepts of geometry was funda- The most important aim was to mediate a way of
mental at the Ulm School of Design. thinking that could be later used for the applied design
tasks. The tasks Maldonado assigned in the foundation
4. Mathematical and Theoretical Background course did not have a reference to practical design,
by Tomás Maldonado though. Instead, the focus was on a methodical ap-
A new type of foundation course that took a more proach to connect science and design. Drawing was
scientific approach was conceived by Tomás Maldo- systematized, and the guiding principle was reflective
nado. In his early years in Argentina, he was one of the visualization.20 In their paper “Wissenschaft und Ge-
leading artists and founders of Arte Concreto-Invención. staltung”,21 Maldonado and Gui Bonsiepe listed the
In Europe, he contacted the European Avant-Garde, following mathematical disciplines as operable for the
met Max Bill in 1948, and decided to write a monograph product designer in the design practice:
on Max Bill’s work, which appeared in 1955. Max Bill of-
fered him a position at the newly founded Ulm School 1. Combinatorics (for modular construction systems
of Design, first as his assistant, and he was later given and problems of measure coordination)
responsibility for the foundation course. Finally, he had 2. Group theory (as symmetry theory for
been rector and member of the rectorate of HfG for construction of patterns and grids)
several years. He introduced working and designing on 3. Theory of curves (for the mathematical treatment
the basis of theoretical knowledge in perception theory of transitions and transformations)
and mathematics. “Visual Methodology” played a lead 4. Polyhedral geometry (for the construction of
ing role in the foundation course in this second phase. regular, demiregular, and irregular solids)
William S. Huff, an American student and later teacher 5. Topology
at HfG Ulm, described the concept of Maldonado’s
foundation course in an interview: Figure 7 shows examples from Maldonado’s foundation
course related to these topics.
Maldonado came to Ulm in the second year. […] So, he In the section on symmetry, Maldonado integrated
started thinking about better ways of doing the Grun different levels of symmetry. In this context, Maldonado

Fig. 7: Peano curve example, student
Klaus Frank; inaccuracy with precise
means, student Adolf Zillmann; pattern
arrangement by student Bertus Mulder;
course taught by Tomás Maldonado,
1955/56, 1956/57. © HfG-Archiv, Ulmer
Museum, Ulm
Fig. 8: Form-fitting net of katametric

elements by student Klaus Schmitt;
grid of katametric elements, student
Jan Thylén, 1961/62, teacher Tomás
Maldonado. © HfG-Archiv, Ulmer Muse-
um Ulm.
picked up the notion of katametry, likely from the 5. Visual Methodology by Anthony Froeshaug
publication on symmetries by German chemists Karl These topics remained important in the following years.
Lothar Wolf and Robert Wolff, who were working on A visual methodology was devised in more detail as a
molecular structures. Their book is subtitled “Versuch key part of the foundation courses. Anthony Froeshaug
einer Anweisung zu gestalthaftem Sehen und sinnvol- came from London and taught at the Ulm School from
lem Gestalten”22 and suggests different levels of sym- 1957 until 1960 in the Department of Visual Communi-
metric structures (see Chapter 6). Katametry is a low cation. In 1958/59, he took over the foundation course
level of symmetric structure, not clearly geometrically for all departments with the support of Maldonado.
defined. Two examples (Figure 8) from Maldonado’s Froeshaug further developed Maldonado’s concept of
course show how he used the notion of katametry as visual methodology into the main focus of the general
a design method. Kurd Alsleben23 and William S. Huff24 foundation course, with the aim of introducing patterns
analyzed, in detail, the symmetric concepts of Wolf and and grids systematically, first in two dimensions and
Wolff (see Chapter 6). then in three.

Fig. 9: Regular and semi-
regular 2D lattice structures
by Anthony Froeshaug, 1959.
© HfG-Archiv, Ulmer
Museum, Ulm.
Tetrahedron+octahedron
Triangular prism—8 edges
12 edges
Tetrahedron+cuboctahedron
8 edges
Rhombic dodecahedron—8 edges dual solid Tetrahedron+truncated
of Archimedean solid Cuboctahedron tetrahedron—6 edges
Tetrahedron+cube+rhombic
cuboctahedron—6 edges
Cube—6 edges Octahedron+truncated

cube—5 edges
Hexagonal prism—5 edges Truncated cuboctahedron

+cube+truncated octahe-
dron—4 edges
Truncated octahedron—4 edges
Fig. 10: Regular and semi-

regular 3D lattice structures
by Anthony Froeshaug, 1959.
© HfG-Archiv, Ulmer
Museum, Ulm.
Fig. 11: Space units for resi-

dential buildings 1961,
design: Bernd Meurer,
Herbert Ohl.25 © HfG-Archiv,
Ulmer Museum, Ulm.

Fig. 12: Use of space without gap, positive-locking stacking solids by student
Herbert Falk, teacher Walter Zeischegg, 1959/60;26 integral building construction,
teachers Herbert Ohl, Günter Schmitz, Rupert Urban, 1957–1963; node example,
teacher Walter Zeischegg, 1962/63.27 Photo left: Christian Staub. © HfG-Archiv,
Ulmer Museum, Ulm.
These grids were the basis for general ideas about structural backgrounds in patterns and spatial lattice
graphs and therefore applicable to many questions. structures found their usage (Figure 11). The studies of
Froeshaug provided an example of a floor plan of Le 3D lattice structures are motivated by designing struc-
Corbusier’s house in La Plata in 1954, using the graphs tures, as it had been an important and innovative field
as an analysis of circulation paths inside the house. The of research for Konrad Wachsmann. He was a guest
graphs were understood as a topology—as connec- lecturer for industrial building at HfG Ulm between
tions, not forms—but were deduced from possible 1954 and 1957. Exercises in spatial tessellations and
regular and semi-regular tessellations. In the 1959 designing nodes were developed in various courses
visual methodology course taught by Froeshaug,28 the (Figure 12).
plane grid versions of regular and semi-regular lattice Studies on space fillings with minimal and maximal
structures (Figure 9) were studied first. spatial packings were integrated into the architecture
The 3D versions of regular and semi-regular lattice department course by Herbert Ohl in 1957. An example
structures (Figure 10) then followed as a basis for by student Hermann Edel, titled “Kristallographie”, is
spatial configurations. shown in Figure 13.
Working with patterns in two and three dimensions
had also been important later in the application fields, 6. Symmetry Concepts by William S. Huff
especially in the architecture and design departments. In the HfG’s later years, the foundation courses were
Modular building methods and design nodes followed taught separately by each department. William S. Huff
in the respective specialized departments, where these taught the basic course in the visual communication
maximum
space filling
medium
space filling
minimal space Fig. 13: Studies on minimal, medium,

filling and maximum space fillings by student
Hermann Edel, teacher Herbert Ohl,
1957.29

department from 1963 to 1966. He referred explicitly The examples in Figure 14 illustrate his structural ap-
to Maldonado’s foundation course, as he had taken it proach, which he described as follows:
as a student, with a Fulbright Scholarship, in 1956/57.
Huff further developed the course to feature a con- By manipulating an element or group of elements with
centration on the notion of structure and manipulating one or more of the symmetry coverage operations,
structures using transformations and deformations. He structures (or systems) are produced. Simple applica
described his view in 1965: tions result in the higher degrees of symmetry (isometry
and homoeometry). More complex combinations of rules,
By structure I strictly mean: the relationship or arrange- though still rigorously applied, lead to the lower orders of
ment of parts of elements. To design, then, is first of all symmetry (syngenometry and katametry). […] In our stu-
to structure; and for me the study of structure (in the ab- dio project, symmetric structures are produced through
stract) is the equal of that which has been known as basic such a programming process.31
design or foundation studies. […] In the physical manipu-
lation of structure, we are interested in those things that The reflections on symmetries and transformations
are invariant and those that are variant, with the study of laid the foundations for his unique tasks on net trans-
deformations and transformations constituting perhaps formations, later termed parquet deformations by
the most crucial underlying motif. […] In our basic design Huff.
course, then (as it comes from Ulm’s Maldonado), my The parquet deformation exercises became Huff’s
students and I explore, of the physical nature of structure, main student assignment for almost four decades.
such groups as can be analysed by symmetry, topology, Figure 16 shows an example of his later teaching in the
combinatorial analysis, theories of color and texture.30 United States.
Fig. 14: Programmed structures by symmetry operations, student Albrecht Hufnagel 1966/67;32
symmetry exercises:33 elements rotate in a combinatorial system involving groups of three
and five, student Dennis Becker 1964; groups of rotating and expanding ellipses are arranged
on a concentric square module in a complexity of translation, rotation, and mirror operations,
student Michael Pollak, 1964, teacher William S. Huff. © HfG-Archiv, Ulmer Museum, Ulm.
Fig. 15: Net transformation or parquet deformation, student Arno Caprez, teacher William S.
Huff, 1965/66.34 © HfG-Archiv, Ulmer Museum, Ulm

Fig. 16: Parquet Deformation “Wreathes of Holly”, by Tina Macica, 19 × 27.88 in, India ink. Fall 1991,
State University of New York at Buffalo. © HfG-Archiv/Museum Ulm, HfG-Ar, BDSA, Hu P 01. 042.
Huff’s background included intensive studies on after K. L. Wolf”35 in 2000. There, he compared
the theory of symmetrical structures, as initiated Wolf’s system with the typology of mapping after
by the input of Maldonado during the foundation March and Steadman.36 Katametry, as introduced
course. This topic became Huff’s main research topic by Maldonado in designing methods, is integrated
and resulted in an essay called “Ordering Disorder into Wolf’s system.
position size angle shape certainty rule
autometry I I I I I I
isometry V I I I I I
homoeometry V I I I I
syngenometry V I I I
katametry V I I
hypometry V I
heterometry V Table 1: Levels of structures after

Wolf and Wolff (1956).
ametry I = Invariant; V = Variant.37
invariant
mapping position length angle and ratio parallelism cross-ratio neighborliness
identity ● ● ● ● ● ●
isometry ● ● ● ● ●
similarity ● ● ● ●
affinity ● ● ●
perspectivity ● ●
topology ●
Table 2: Typologies of mapping after

March and Steadman.38

These typologies of symmetries, which March and by Kurd Alsleben,41 published in 1962. Maldonado
Steadman called mappings in a more mathematical brought Alsleben to HfG Ulm as a lecturer, where he
concept, correspond with the understanding of sym- taught structure theory and Boolean algebra between
metry as transformations that had existed since Felix 1965 and 1968. Together with physicist Cord Passow,
Klein’s Erlangen program, in which the invariants of he produced and published early programmed com-
figures characterize the respective transformation. The puter drawings. The book refers to semiotics, Gestalt
transformation or mapping, according to March and psychology, information theory, and aesthetics, all
Steadman, became the fundamental element form- disciplines which formed the fundament of teaching
ing the respective geometry.39 The Euclidean plane by Max Bense, Abraham Moles, and Tomás Maldonado
or spatial geometry or congruence geometry can be at HfG Ulm. Alsleben described the symmetry theory
characterized as a transformation group of isometric of Wolf and Wolff as important for providing rules for
and conformal transformations. Similarity transfor- image creation, even if the particularly productive low
mations are not isometric and allow uniform scaling. symmetry levels still elude systematics. The symmetry
Non-conformal transformations that maintain paral- theory shows how the geometric position and shape
lelism, but not lengths and angles, are affine trans- of signs can be changed according to rules. Then he
formations. The perspective transformation or pro- described the different levels of the used symmetry
jective transformation only maintains collinearity and notions:
cross-ratio. Ultimately, only neighborhoods remain in ⟩ Autometry: Identical
the topological transformation. These typologies corre- ⟩ Isometry: Equality of elements and their
spond with strong geometric concepts, whereas Wolf’s uniform repetition
levels of structure, shown with examples in Figures ⟩ Homoeometry: Similar, repetition in the same
17 and 18, meet more open characteristics relevant change in size, position, or behavior
for Gestalt-like seeing and reasonable designing, as ⟩ Syngenometry: Shape related
expressed in the subtitle of Wolf’s book.40 ⟩ Katametry: Design related
A profound analysis of Wolf’s symmetry classifica- ⟩ Heterometry: Shape different
tions can be found in the book Ästhetische Redundanz ⟩ Ametry: Unformed, given only in the idea
Fig. 17: Isometric surface net

and homoeometric strip.42
Fig. 18: Syngenometric strip

and katametric strip.43
Fig. 19: Examples of degrees of similarity
by Kurd Alsleben.44
In the examples created by Alsleben, the meaning book, Moles classifies this as part of information aes
of the low-level symmetries remains vague. Alsleben thetics, which was one of the most significant develop-
characterized symmetry as corresponding to repetitive ments in aesthetics since Hegel. The interplay between
redundancy. As an artistic device, symmetry belongs to redundancy and information is the basis of this theory
syntactically effective tools.45 His reflections are embed- or, in the words of Alsleben, “Redundancy makes the
ded in information aesthetics, developed by Max Bense style of a work of art, information its originality.”47 The
and Abraham Moles46 in the 1960s and introduced as artist or creator moves between originality with perfect
a relevant theoretical philosophical background at HfG irregularity and banality with perfect regularity, as
Ulm (see Chapter 7). In his introduction to Alsleben’s shown in Figure 20.
unpleasant
unpleasant
fascinating
interesting
harmonic
neutral
neutral
originality banality
perfect perfect
irregularity regularity
structure dispersion
Fig. 20: Originality—banality—scale,

according to Alsleben48 and Moles.

Huff did not refer to information aesthetics in his Bense supplemented Birkhoff’s numeric aesthetics
papers, but his evaluation of the lower levels of sym- with information aesthetics. The aesthetic state had
metric structures in the creation process of design, been defined by Bense as the relation of an ordered
what he called programmed design, coincides with the to a not-ordered state. Birkhoff interprets his aesthetic
role of redundancy and innovation in the concept of measure M as aesthetic information. The order relation O
information aesthetics. corresponds to redundancy in finding order relationships
and symmetries. Redundant features are necessary for
The programmed design is a staple assignment of my innovation to become recognizable:
formative design studio in architectural studies. While
the succinct brief for the assignment permits anything A perfect innovation in which there were only new states
from isometry on down Wolf’s scale of symmetric struc as in chaos would not be recognizable. A chaos is finally
tures to katametry, the exploration of the lowest level has unidentifiable. The recognizability of an aesthetic state re-
led to some of the more striking results: designs that, quires not only the recognizability of its singular innovation,
determined by indiscernible, though not overly complex but also their identifiability based on their redundant order
rules—indeed, the leaner the rules, the more fulfilling the characteristics.55
provocative outcome—appear to be casual or capricious.
Programmed randomness? A contradiction, indeed! That Therefore, the interplay of redundancy and innovation—
is to say, what is often perceived as random may not be order and chaos—must be in an optimal relation to one
random at all, but wholly programmed.49 another to achieve an aesthetic state. The artwork gives
aesthetic information as a material carrier of the aes-
7. Information Aesthetics and Geometry thetic state. Information is always transmitted by signs.
Geometric structures and transformations are applied Here, semiotics as the science of signs comes into play
in design processes, as illustrated in the shown exam- for the transmission of information. The interrelationships
ples. The question is, though, how aesthetic results can between information aesthetics and semiotics became
be achieved and how aesthetics can be substantiated obvious. Semiotics played an important role in research
in relation to order structures, rhythms, and—as Bill and teaching at HfG Ulm beyond Bense and Moles; it
characterized it—in the individual creative decisions also played a role in the courses of Maldonado.
that constitute the difference between art and mathe- The structural approach in designing must be seen
matics.50 Information Aesthetics developed criteria for in light of the described aesthetic and philosophical
aesthetic measures and evaluations. It was initiated by background. Huff saw the notion of structure and then
Max Bense and Abraham Moles in the 1960s, mainly working on those structures as his fundament. Abraham
in Germany and France, as an aesthetic theory based Moles formulated this comprehension as valid for our
on a rational mathematical fundament. Frieder Nake, a relationship to the world, science, and design:
protagonist of information aesthetics and a student of
Max Bense who later became a professor for computer To grasp the world, we must grasp it. In order to grasp
graphics in Bremen, summarized the theory, its applica- it, we must first structure it. However, there are not
tions, and its critics in an article.51 structures in themselves, but only perceived structures.
There are two roots of this new aesthetic theory: Science as an essential form of understanding the world
information and aesthetic measures. Information as provides us at the same time with measures and forms
a root was introduced by Claude E. Shannon52 during for this structuring.56
the rise of communication theory and communica-
tion technology. His mathematical information model Huff’s work refers to this general background of perceiv
integrated the stochastic nature of news.53 The possible ing structures. The rules and regularities of the designs of
states of a system can be described in combination patterns are often not caught directly but are only uncov-
with a set of transition probabilities from one state to ered by analyses. He describes parquet deformations
the next. Bense applied Shannon’s information the- as being in the “territory between monotony and bewil-
ory to aesthetics. Aesthetic realizations are seen as derment”, in other words for between redundancy and
part of a communication process. The second root is innovation and between order and chaos in the concept
Georg David Birkhoff’s aesthetic measure.54 The aes- of information aesthetics:
thetic measure is defined as the function of the order
and complexity grade of the viewed configuration: M = We have especially addressed structures, i.e., programmed
O/C, where O indicated the number of order relations, designs, that are regulated […] by relatively few elements
symmetries, and harmonies, and C is complexity. The and relatively few rules and that do not exhibit readily per-
aesthetic measure of an artwork could be calculated as ceived regularities—though the regularities can be uncov
numeric quantities based upon order and complexity. ered by right analyses. These patterns can engage the visual
sense in territory between monotony and bewilderment, first instance to be random, yet is distinctly regulated and,
a complexity of experience, achieved through minimal thus, sensed to be coherent. Can this be called “hidden
regularization—pattern-distribution that appears in the harmony”?57
References
1
Max Bense, Aesthetica, Agis, Baden-Baden, 1965, 2nd edition 1982.
2
Elisabeth Walther, “Philosoph in technischer Zeit – Stuttgarter Engagement. Interview mit
Elisabeth Walther, Teil 2”, in: B. Büscher, von H.-G. Herrmann, C. Hoffmann (eds), Ästhetik als
Programm. Max Bense/Daten und Streuungen, Diaphanes, Berlin, 2004, p. 72, translated by C. L.
3
Max Bill, “Die mathematische Denkweise in der Kunst unserer Zeit”, Werk 36, 3, Winterthur
1949, English version: “The Mathematical Way of Thinking in the Visual Art of Our Time”, in:
Michele Emmer (ed.), The Visual Mind: Art and Mathematics, MIT Press, Cambridge, 1993, pp.
5–9; Max Bill, “Structure as art? Art as structure?”, in: György Kepes (ed.), Structure in Art
and in Science, Braziller, New York, 1965, pp. 150–151.
4
Bill, “Die mathematische Denkweise”, pp. 7–8.
5
Max Bense, Artistik und Engagement. Präsentation ästhetischer Objekte, Kiepenheuer &
Witsch, Cologne/Berlin, 1970, p. 92, translated by C. L.
6
Max Bense, Konturen einer Geistesgeschichte der Mathematik II. Die Mathematik in der
Kunst, Claassen & Goverts, Hamburg 1949, p. 59.
7
Paul Valéry, “Introduction à la méthode de Léonard de Vinci”, La Nouvelle Revue Française,
Paris, 1895.
8
The mathematical approach combined with the philosophical background, especially by Max
Bense has been analyzed in two papers: Cornelie Leopold, “Precise Experiments: Relations be-
tween Mathematics, Philosophy and Design at Ulm School of Design”, Nexus Network Journal 15,
2013: pp. 363–380; Cornelie Leopold, “The Mathematical Approach at Ulm School of Design”, in:
Emmer Michele, Abate Marco, Villarreal Marcela (eds.), Imagine Maths 4. Between Culture and
Mathematics, Unione Matematica Italiana, Bologna, 2015, pp. 15–28.
9
Ulmer Museum/HfG-Archiv (ed.), ulmer modelle – modelle nach ulm. Hochschule für
Gestaltung Ulm 1953–1968, Hatje Cantz, Ulm, 2003, pp. 6–7.
10
Sketches by Hans G. Conrad from 1954 show examples of such an analysis and creation of
ornamental drawings.
11
Model by the participants of the summer school in Buenos Aires, photo by Willem Roelof
Balk, cf. Fachbereich Architektur, Technische Universität Kaiserslautern (ed), rup’, Technische
Universität Kaiserslautern, Kaiserslautern, 2012.
12
Ibid., pp. 27–28, Figs. 59, 61, 62.
13
Ibid., p. 28, Figs. 63, 64, 67.
14
Exercise in the course by Hermann von Baravalle, photo: Oleg Kuchar. © HfG-Archiv, Ulmer
Museum Ulm.

15
Exercise in the course by Hermann von Baravalle, reproduced by courteous permission
of Hermann Edel.
16
Hermann Edel, “Strukturelles Denken an der Hochschule für Gestaltung Ulm”, in:
Joaquín Medina Warmburg, Cornelie Leopold (eds.), Strukturelle Architektur. Zur
Aktualität eines Denkens zwischen Technik und Ästhetik, Transcript, Bielefeld, 2012,
pp. 55–73.
17
Hermann von Baravalle, Geometrie als Sprache der Formen, Verlag Freies Geistesleben,
Stuttgart, 1957, 3rd edition 1980.
18
Hermann Edel, “Geometrische Lichtprojektionen auf einen Fadenzylinder – Baravalle-
Kino”, in: Cornelie Leopold (ed.), Über Form und Struktur – Geometrie in Gestaltungs-
prozessen, Springer, Wiesbaden, 2014, pp. 99–102. Video: https://vimeo.com/656511278
(accessed on 12/2021).
19
Martin Krampen and Günther Hörmann, Die Hochschule für Gestaltung Ulm/The School
of Design. Anfänge eines Projektes der radikalen Moderne/Beginnings of a Project of
Radical Modernism, Ernst & Sohn, Berlin, 2003, pp. 101–103.
20
Ibid., p. 101.
21
Tomás Maldonado and Gui Bonsiepe, “Wissenschaft und Gestaltung”, ulm 10/11,
Hochschule für Gestaltung, Ulm, 1964.
22
Karl Lothar Wolf and Robert Wolff, Symmetrie. Versuch einer Anweisung zu
gestalthaftem Sehen und sinnvollem Gestalten, systematisch dargestellt und an
zahlreichen Beispielen erläutert, Böhlau-Verlag, Münster/Cologne, 1956.
23
Kurd Alsleben, Ästhetische Redundanz. Abhandlungen über die artistischen Mittel der
bildenden Kunst, Verlag Schnelle, Quickborn, 1962.
24
William S. Huff, “Ordering Disorder after K. L. Wolf”, Forma 15, 2000, pp. 41–47.
25
Ulmer Museum/HfG-Archiv (ed.), ulmer modelle – modelle nach ulm. Hochschule für
Gestaltung Ulm 1953–1968, Hatje Cantz, Ulm, 2003, pp. 24–25.
26
Martin Krampen, Günther Hörmann, Die Hochschule für Gestaltung Ulm/The School of
Design. Anfänge eines Projektes der radikalen Moderne/Beginnings of a Project of Radical
Modernism, Ernst & Sohn, Berlin, 2003, p. 120.
27
Herbert Lindinger (ed.), Hochschule für Gestaltung Ulm ... Die Moral der Gegenstände.
Ernst & Sohn, Berlin, 1987, p. 60, 202.
28
Anthony Froeshaug, “Visuelle Methodik”, ulm 4, Hochschule für Gestaltung, Ulm, 1959.
29
Private archive of Hermann Edel, reproduced by courteous permission of Hermann
Edel.
30
William S. Huff, “An Argument for Basic Design”, ulm 12/13, Hochschule für Gestaltung,
Ulm, 1965, p. 26.
31
Note 1 by William S. Huff, “Symmetry or Programmed Design”, 1960—after Tomás
Maldonado.
32
Lindinger, Hochschule für Gestaltung, p. 57.
33
Huff, “An Argument”, pp. 30, 36.
34
Ibid., p. 64.
35
36
Lionel March and Philip Steadman, The Geometry of Environment, MIT Press, Cambridge,
MA 1974), pp. 24ff.
37
Huff, “Ordering Disorder”, p. 43.
38
March and Steadman, The Geometry of Environment, p. 25.
39
More on this concept can be found in Cornelie Leopold, “GeometrischeTransforma-
tionen als Entwurfsmethodik/Geometric Transformations as Design Methodology”, in: Arena
A., et al. (a cura di), Connettere. Un disegno per annodare e tessere. Atti del 42° Convegno
Internazionale dei Docenti delle Discipline della Rappresentazione/Connecting. Drawing for
Weaving Relationships. Proceedings of the 42th International Conference of Representation
Disciplines Teachers, FrancoAngeli, Milano, 2020, pp. 1221–1240.
40
Wolf and Wolff, Symmetrie. Versuch einer Anweisung zu gestalthaftem Sehen und
sinnvollem Gestalten. The subtitle could be translated to “Attempt at instructions for
gestalt-like seeing and reasonable designing”.
41
Alsleben, Ästhetische Redundanz.
42
Wolf and Wolff, Symmetrie, p. 4.
43
Ibid., p. 5.
44
Alsleben, Ästhetische Redundanz, pp. 60–61.
45
Ibid., p. 55.
46
Abraham André Moles, Information Theory and Esthetic Perception, University of Illinois
Press, Urbana, 1966. French original 1958.
47
Alsleben, Ästhetische Redundanz, p. 22.
48
Ibid., p. 23.
49
Huff, “Ordering Disorder”, p. 46.
50
Max Bill, “Structure as art? Art as structure?”, in: György Kepes (ed), Structure in Art and
in Science, Braziller, New York, 1965, pp. 150–151.
51
Frieder Nake, “Information Aesthetics: An heroic experiment”, J Math Arts 6(2–3), 2012,
pp. 65–75.
52
Claude E. Shannon, “A Mathematical Theory of Communications”, Bell Tech J 27, 1948, pp.
379–423, 623–656.

53
The development of information aesthetic in relation to the art works of Gerard Caris is
explained by the author in: Cornelie Leopold, “Geometric and Aesthetic Concepts Based on
Pentagonal Structures”, in: B. Sriraman (ed), Handbook of the Mathematics of the Arts and
Sciences, Springer, Cham, 2019.
54
Georg David Birkhoff, Aesthetic Measure, Harvard University Press, Cambridge, 1933.
55
Bense, Aesthetica, p. 356.
56
Abraham André Moles, “Produkte: ihre Funktionelle und strukturelle Komplexität”, ulm 6,
Hochschule für Gestaltung, Ulm, 1962, p. 4.
57
William S. Huff, “On Regulation and Hidden Harmony”, in: O. Bodnar (ed.), Harmony
of Forms and Processes: Nature, Art, Science, Society, International Society for the
Interdisciplinary Study of Symmetry, Lviv, Ukraine, 2008.
Cornelie Leopold teaches and researches in the field of architectural geometry at fatuk,
Faculty of Architecture, Technische Universität Kaiserslautern, Germany, in the position
of academic director and head of the Descriptive Geometry and Perspective section.
She received her degree in Mathematics, Philosophy, and German Philology at University
of Stuttgart, Germany, with specializations in Geometry and Philosophy (Semiotics,
Aesthetics, Logic, and Philosophy of Science). She is a member of the Editorial Board of
the Journal for Geometry and Graphics and of the Scientific Committee of the Journal
Disegno of UID—Unione Italiano Disegno. Since 2019 she is corresponding editor of
Nexus Network Journal: Architecture and Mathematics. She was founding president of the
Deutsche Gesellschaft für Geometrie und Grafik (DGfGG) and is a member of the board for
the Committee of the International Society for Geometry and Graphics (ISGG) as director
for Europe/Near East/Africa. She has contributed lectures, papers, and reviews to many
international conferences and journals. In the past, she was a guest lecturer in Krakow,
Istanbul, Milan, Porto, Venice, and Buenos Aires. In 2017, she was visiting professor at
Università Iuav di Venezia, Italy with a research focus on perspective transformations.
Her research interests include the development of spatial visualization abilities, geom-
etry and architectural design methods, structural thinking, the philosophical background
of architecture, visualization of architecture, geometry, and representation. Results of her
research have been published in conference papers, books, and articles. Her book Geo-
metrische Grundlagen der Architekturdarstellung, first published in 1999, was re-published
in 2019 in the 6th extended edition by Springer. Research on the interrelations between
geometry and her background in philosophy was published in the co-edited book Struk-
turelle Architektur. Zur Aktualität eines Denkens zwischen Technik und Ästhetik at Transcript
and in articles focusing on the philosophy of Max Bense and the role of mathematics at
Ulm School of Design.
In 2018 she took part in the Conference Chairs Team of RCA, Research Culture in
Architecture - International Conference on Cross-Disciplinary Collaboration, organized and
hosted by fatuk and co-editor of the connected book, published in 2020 at Birkhäuser.
The international conference NEXUS 20/21: Relationships of Architecture and Mathematics
organized at TU Kaiserslautern and coordinated by her together with Kim Williams in
summer 2021.
https://geometrie.architektur.uni-kl.de
https://www.researchgate.net/profile/Cornelie_Leopold
The Tiles, They Are a-Changin’
Craig S. Kaplan
School of Computer Science,
University of Waterloo, Ontario, Canada;
[email protected]
Like many people, I first encountered parquet defor- much artistic satisfaction from writing code to produce
mations through Douglas Hofstadter’s essay, which I drawings as I might from executing them by hand. I
read in Metamagical Themas1 in my late teens or early hope that we live in an age where it is uncontroversial
twenties. I was primed to fall in love with them. I have to speak of code as a creative medium.2
admired the art of M. C. Escher from the earliest days It turns out that a family of tilings called isohedral
of my childhood; I begged my parents to get me a tilings are particularly well suited to the creation of tes-
reprint of his Metamorphosis II, which still hangs in my sellation-based geometric designs like parquet defor-
office today. As a PhD student in computer science, mations. In this chapter, I will first give a brief overview
I then had the good fortune to be drawn back into of the isohedral tilings, focusing on their mathematical
this topic from a more informed perspective. In my and computational properties that are relevant in a de-
research, I adapted ideas from the branch of mathe- sign context. I will then present the results of a number
matics called tiling theory to generate, manipulate, and of experiments that use the isohedral tilings as a basis
render tessellations inspired by Escher and by Islamic for drawing parquet deformations. I will conclude with
art. As part of that research, I produced my first par- a few remarks about the constraints Huff imposed
quet deformations, and have continued to experiment upon this art form, and when it may be appropriate to
with them in the almost 20 years since then. My latest break them.
phase of experimentation is a direct result of Werner
Van Hoeydonck’s invitation to contribute a chapter to 1. The Isohedral Tilings
this book. I am grateful for the motivation to return to Tilings have been part of art and ornamentation for
this topic and resurrect my old software. thousands of years, but only more recently did a formal
Every parquet deformation is a tiling—or a finite branch of mathematics develop around their study. To-
excerpt of a conceptually infinite tiling, at any rate—and day tiling theory is a beautiful, deep topic that overlaps
so it is natural to expect that tiling theory would be an with many other parts of mathematics. Happily, the
ideal tool for studying them and for generating new sorts of tilings that interest us here tend to be relatively
designs. Of course, one need not be a mathematician tame in comparison to the frontier of research, and
to draw beautiful tilings! Escher demonstrated this fact well-established ideas can be used to great effect.
beyond any doubt, and we can assume that most of While I wish to provide some intuition for the mathe-
Huff’s students were guided more by intuition and visu- matical machinery that I use to construct parquet defor-
al rhythm than by mathematical rules. Mathematics mations, it would be counterproductive to give a full
is but one possible route to geometric design, albeit a account of the background. Interested readers should
powerful one. In this context, the power of mathemat- consult Grünbaum and Shephard’s masterful Tilings
ics lies in its ability to map out a design space system- and Patterns3; artists and designers will also find
atically, categorizing a body of existing work and making endless inspiration in its pages. My own book offers a
sure we have not overlooked any opportunities. A com- condensed introduction more focused on algorithms
puter scientist can then reduce these mathematical and data structures for writing software to draw tilings.
ideas to a practical piece of software, one that allows A tiling is a collection of shapes that cover the
an artist to focus on creative exploration rather than plane with no gaps and no overlaps. We are often inter
the execution of the drawing. For my part, I derive as ested in tilings formed from copies of a single shape,

3.3.3.3.3.3 3.3.3.3.6 3.3.3.4.4 3.3.4.3.4
3.4.6.4 3.6.3.6 3.12.12 4.4.4.4
Fig. 1: The eleven Laves tilings, which

serve as a basis for constructing iso
hedral tilings. The symbolic name for
each tiling describes the number of
3.6.12 4.8.8 6.6.6 edges meeting at every vertex around
the boundary of a single tile.
where each copy is transformed by some combination are denoted IH1, IH2 … all the way to IH93. Each type
of translation, rotation, and reflection. Such tilings encompasses a broad range of tilings that are arranged
are called monohedral, and the single shape is the according to a shared set of neighbor relationships.
tiling’s prototile. Twelve of these types are less useful for our purposes,
In art and design, we are typically interested in because they can only be realized visually with the
tilings that have a degree of what we can informally addition of extra markings to the interiors of tiles. We
call “regularity”. By virtue of its single prototile shape, a typically skip over those types and use the other 81 in
monohedral tiling has some regularity, but the arrange- decorative applications of tilings. Figure 1 shows small
ment of tiles can still be uncomfortably complex, both sample tilings for each of these 81 isohedral types.
mathematically and aesthetically. Escher developed
a “layman’s theory”5, codifying his intuition for regular 2. Drawing Isohedral Tilings
division of the plane, after encountering a research arti The isohedral tilings occupy a sweet spot at the inter
cle on the subject by Pólya. Later, Heesch and Kienzle section of mathematics, computation, and design.
formalized regularity in a tiling by requiring that every They are a good fit for the intuitive notion of regularity
tile be surrounded by its neighbors in the same way.6 adopted by Escher. They are fully and systematically
This local constraint imposes global order, allowing described using simple notation. And this notation can
Heesch and Kienzle to divide all such tilings into 28 be converted into data structures and algorithms for
distinct families. representing, manipulating, and drawing isohedral tilings
Grünbaum and Shephard offered the definitive efficiently. I created a software library called Tactile for
treatment of this notion of regularity in their analysis this purpose early in my research on this subject, and
of the isohedral tilings.7 Viewed informally, they define more recently published an updated version.8
an isohedral tiling as one in which the individual tiles To describe an isohedral tiling (or rather, its single
behave as simply as possible with respect to the tiling’s prototile), we first select one of the eleven Laves tilings,
symmetries. As in Heesch and Kienzle’s work, this shown in Figure 2. A Laves tiling is the scaffolding upon
global constraint is equivalent to requiring that every which we will affix the details of the tile shape: It re-
tile be surrounded by its neighbors in the same way. cords the connectivity between tiles and their neighbors,
Grünbaum and Shephard develop a compact symbolic without regard for precise tile shape. The Laves tilings
notation for describing a tile’s internal symmetries and are canonical representatives of all different possible
its relationships to its neighbors, and list all possible connectivities for isohedral tilings. Each one has a sym-
symbols that lead to valid tilings. The result is a set of 93 bolic name that lists the number of edges that meet at
families of tilings called the isohedral tiling types, which the vertices around the boundary of a single tile.
110 The Tiles, They Are a-Changin’ Craig S. Kaplan

IH1 IH2 IH3 IH4 IH5 IH6

Fig. 2: Examples of tilings for each

of the isohedral types. There are 81
swatches: We skip over the twelve types
like IH19 that require interior markings,
and that are therefore not useful in the
IH90 IH91 IH93 context of parquet deformations.
Depending on the isohedral tiling type, we might have Of course, we often wish to draw the edges of tiles
the opportunity to alter the shape of the Laves tile. For not as straight lines, but as more expressive curves.
example, 36 different isohedral types are all based on We complete the description of a prototile by supply
4.4.4.4, the regular tiling by squares. In different types, ing a set of distinct paths that will be used to join
those squares can be deformed into rectangles, or paral- the vertices of the base tile. A given prototile tile may
lelograms, or other kinds of quadrilaterals. Every isohedral re-use the same path multiple times, whether because
type is equipped with a tiling vertex parameterization, a of internal symmetries or to allow it to interlock with
set of numbers that can be chosen freely to determine an a translated neighbor. Any path that can be expressed
alteration to the initial Laves tile. Different types have dif- computationally may be used here; the most common
ferent numbers of parameters, from zero (a type with no choices in digital graphic design are piecewise polyg-
degrees of freedom in its vertex locations) up to six. I refer onal paths (made up of straight line segments) and
to this modified Laves prototile shape as the base tile. smooth cubic curves.

Fig. 3: The construction of a prototile for an isohedral tiling. On the left, we select the Laves
tiling 4.4.4.4. The square prototile is then modified into a parallelogram based on the values of
two tiling vertex parameters. The parameters p1 and p2 can be thought of as controlling the x
and y coordinates of the upper-right corner of the prototile, thereby allowing us to choose any
parallelogram as the base tile. We then apply two distinct paths to define the shapes of the
edges. The resulting shape produces a tiling of isohedral type IH41.
My Tactile library serves as a simple means of by specifying a sparse set of keyframes, each of
collecting all the information above and combining it to which associates an isohedral prototile with a position
determine the shape of the prototile and the arrange- along one axis of the canvas. The computer can then
ment of tiles in an isohedral tiling. Figure 3 illustrates smoothly interpolate the numbers controlling the tile’s
the construction of one particular isohedral tiling, of shape, producing a patch of tiles with shapes that
type IH41, based on a choice of Laves tiling, values for a evolve gradually in response to the keyframes. Here the
tiling vertex parameterization, and paths for the edges. rigid nature of mathematical interpolation is perhaps a
virtue, as it plays into the already geometric aesthetic
3. Parquet Deformations of this art form.
as Spatial Animations As a simple example, consider the diagram in Figure
In traditional cell animation,9 a lead animator would 4. The squares represent a fragment of a tiling that we
draw a relatively sparse set of keyframes, showing the wish to elaborate into a parquet deformation. For each
most important poses taken by a character in motion. square edge, we use the x coordinate of the midpoint of
Other animators called in-betweeners would then fill that edge (marked with a dot) to look up an interpolation
in the intermediate frames to depict the full motion of amount between 0 and 1. A value of 0 tells us to draw
the character, usually at 24 frames per second. When tiles resembling a keyframe at the left side of the can-
the frames are projected back in sequence, the eye vas, and 1 tells us to draw the right keyframe. Interme-
fuses these discrete snapshots into smooth motion. diate values should interpolate smoothly between the
Computer animation offers the promise of alleviat- keyframe shapes. We can visualize this lookup process
ing the burden of in-betweening by replacing it with by drawing a vertical line downward from the midpoint
interpolation, the direct calculation of intermediate and finding its intersection with the linear ramp D(x).
values for pose parameters like joint angles. In prac Figure 5 shows the parquet deformation that results
tice, this promise has its limits: Automated interpolation from using these values of D(x), where the left keyframe
looks too rigid, and the most expressive animated char is a square and the right keyframe has zigzag edges. A
acters are still meticulously posed frame by frame. single evolving edge is shown underneath for reference.
Previously, I demonstrated how to create temporal In practice, D(x) can be a more general function than
animations of isohedral tilings by analogy with traditional a linear ramp, allowing us to control the speed at which
animation.10 We can regard a parquet deformation as a tile shapes evolve, or even to move back and forth be
kind of spatial animation, depicting the evolution of our tween two keyframes multiple times. We can also chain
“character” from one side of the canvas to the other together more than two keyframes, producing longer
rather than through time. We describe the a nimation animations with more intermediate steps.

Fig. 4: A visualization of how
interpolation values are calculated
in a linear parquet deformation.
Each edge midpoint (marked with a
dot) is projected downward onto an
interpolation function D(x), which yields
a value between 0 and 1.
Fig. 5: A parquet deformation

constructed using the tilted squares
of Figure 4. Keyframes are defined at
the left and right edges of the canvas,
both of isohedral type IH62. The left
keyframe has straight edges, and the
right keyframe has zigzag edges.
4. Evolving Edge Shapes

In many of the parquet deformations produced in the squares of the tiling are each subdivided into finer
Huff’s studio, a single base tile is used for the duration grids, and at each evolutionary step the edge shape
of the drawing, and only the edge shapes are permit- shifts to enclose or liberate one grid cell. Hofstadter
ted to evolve. This constraint is natural to adopt in the refers to deforming a path by “introducing a ‘bump’ or
context of hand drawing—it becomes easier to con- ‘pimple’ or ‘tooth’”; I will refer to this device less viscer
ceive of the layout of tiles, while still permitting a wide ally as “grid-based” evolution.
space of creative exploration. It also fits with Huff’s use A grid-based system is ideal for interactive design
of square and isometric lattices as a basis for geomet in software. I developed an interface in which the artist
ric design. Over the years I have conducted a number clicks on grid squares, defining a sequence of discrete
of experiments based on this simplified framework. shifts to be taken by the edge paths. The software
In several of them I restrict my attention purely to the renders the final parquet deformation by constructing
unmodified Laves tiling of squares (4.4.4.4) and consid- each edge in the tiling with a number of shifts based
er different strategies for evolving the paths that define on the edge midpoint’s horizontal position in the draw-
the edges.11 I will present three such strategies here, ing. The interpolation is “snapped” to the nearest whole
and then discuss a more general system for interpolat- number of shifts—edges do not evolve continuously,
ing paths in all isohedral tiling types. even in principle. Figure 7 illustrates the definition of a
single edge evolving according to a grid-based meth-
4.1 Grid-Based Evolution od, and a parquet deformation constructed from that
The first evolution strategy is typified by a drawing like edge. Figure 8 shows two additional examples that use
that shown in Figure 6, produced in Huff’s studio. Here, grid-based evolution.
Fig. 6: “Memory Chip Meander”, by Frank W. Dunn, 19 × 27.88 in, India ink. Basic Design Studio
of William S. Huff, State University of New York at Buffalo, 1983. © HfG-Archiv/Museum Ulm,
HfG-Ar, BDSA, Hu P 06. 015. An example of grid-based curve evolution.

Fig. 7: A demonstration of grid-based path evolution. The top-left grid defines a sequence of
shifts for an evolving path, which grows to wrap around each numbered square in turn. The paths
can then be distributed across a patch of tiles to define a parquet deformation, here of type IH61.
Fig. 8: Two examples of grid-based evolution, of types IH62 (top) and IH73 (bottom). The top
example uses a mountain-shaped deformation function D(x), which interpolates from the
first tile shape to the second and then back again.
4.2 Organic Evolution temporal animation of an evolving shape, which always

Computer graphics researchers and artists have experi tiles the plane. I harness this animation to draw a par-
mented with approaches for evolving organic forms quet deformation by replacing the edges of the base
using physical or biological simulations. In my work I tiling with snapshots of edges from different points in
adapted an algorithm by Pedersen and Singh,12 in which time. The simulation deforms the tile gradually enough
a curve undergoes random perturbations while at- that these snapshots produce a continuous parquet
tempting to remain smooth and holding other parts of deformation. Figure 9 shows two examples of parquet
the curve at a fixed distance. I adapt the algorithm so deformations based on organic evolution. I used this
that multiple copies of the curve may attract and repel technique to design low-relief parquet deformations
each other around the boundary of a prototile, while for the entrance of the National Museum of Mathemat
the vertices of the underlying Laves tiling remain fixed. ics in New York City, two small sections of which are
The sequence of time steps in this simulation forms a shown in Figure 10.

Fig. 9: Two examples of organic evo-
lution, of types IH61 (top) and IH74
(bottom).
4.3 Fractal Evolution

Fractals have been a mainstay of mathematical art for as-is. In the first few steps, the difference between
decades, and so it is natural to investigate their use in consecutive paths is too large and abrupt; later,
parquet deformations. I have conducted a few visual changes are invisibly small. I mitigate this problem by
experiments using simple fractal curves based on iter interpolating smoothly between consecutive refine-
ated function systems. ment steps using a linear blend, producing a more
Suppose that P is a polygonal path. We can con- continuous evolution of form. Here, blending refers to
struct a slightly more elaborate path by replacing every a class of algorithms that interpolate smoothly be
line segment in P by a scaled, translated, and rotated tween paths. A blend algorithm might take two paths P1
copy of P itself. We can iterate this process, yielding a and P2, together with a number t between 0 and 1, and
sequence of ever more detailed paths. The hypothetical produce a new path. We expect that t = 0 produces a
limit of repeating this process infinitely many times is copy of P1, t = 1 produces a copy of P2, and other values
a fractal curve, assuming that we started with a path P of t will carry out a continuous progression between
that is reasonably well-behaved.13 these two limiting shapes. Software for illustration and
We might try to take the first few curves produced animation often includes tools for constructing blends.
in this sequence and distribute them across a design, Figure 11 shows two examples of parquet deformations
but in practice they exhibit poor behavior when used based on fractal evolution.
Figure 10: Two fragments of a long organic Parquet Deformation of type IH73, executed in
bas-relief bronze at the entrance of the National Museum of Mathematics in New York City.
Photographs courtesy of of National Museum of Mathematics (momath.org).

Fig. 11: Two examples of fractal evolution,
of types IH62 (top) and IH71 (bottom).
Fig. 12: Two examples of generic path blending, of types IH29 (top) and IH18 (bottom). Both
examples use edges that start straight and grow in complexity. The bottom example passes
through multiple intermediate keyframes to produce a smoother progression of form.

Fig. 13: Two designs based on keyframes, of types IH61 (left) and IH66 (right). If the tiles are
left to interact directly, their different patterns of orientations will produce multiple distinct
tile shapes. If desired, we can avoid this interaction by passing through the square tiling
shared by these two isohedral types.
4.4 General Interpolation of Edges

The blending process described in the previous sec- need not all have the same orientation. Furthermore,
tion suggests a general scheme for producing parquet different isohedral types can place tiles in different
deformation based on keyframes. Recall from Sec- combinations of orientations. For example, consider the
tion 2 that an isohedral prototile can be described by tilings at the left and right sides of the design in Figure
choosing a Laves tiling, a set of tiling vertex parameters 13. On the left, the tiles are arranged in a checkerboard
that modify the initial shape of the Laves tile, and edge of two distinct orientations; on the right, the orienta
shapes to join its vertices. Suppose that two keyframes tions form alternating rows. When those two tiling
are defined at the ends of a drawing, using the same types collide, different combinations of paths will meet
base tile (that is, the same combination of Laves tiling at different edges of the shared Laves tiling, leading
and tiling vertex parameters). A tiling by the base tile here to intermediate tilings made from two distinct tile
provides a common ground for interpolating between shapes. Huff’s rules for parquet deformations required
the two keyframes. Each base edge carries two paths, that each position in the drawing have an implied
one from each of the keyframe tilings. As in Figure 4, we monohedral tiling, a requirement that is violated here.
can use the midpoints of an edge to look up an interpo- We could ignore this rule and accept the drawing. But
lation amount D(x) between 0 and 1. These values can if we wish to adhere to Huff’s constraints, one option
then be plugged into a blending algorithm, together with is to introduce a third keyframe at the center of the
the two paths that share that edge, to yield an interme- drawing in which all edges are straight, as shown at the
diate path. As we move along the length of the canvas, bottom of Figure 13. This straight-edged tiling is repre-
the values of D(x) change gradually, producing tiles that sentable within the space of every isohedral type based
morph smoothly from the first keyframe to the second.14 on it, and so we can always pass through it mono
Figure 12 shows two examples that interpolate between hedrally on the way from one isohedral type to another.
general paths.
A problem can arise with this approach when two 5. Evolving the Laves Tiling
keyframes share a Laves tiling but do not use the same In the techniques of the preceding section, we as
isohedral type. Although the tiles in a given isohedral sumed that keyframes share the same base tile. The
tiling must all be copies of the same shape, they tilings associated with two keyframes can therefore

be brought into perfect alignment, and interpolation is The most extreme type of change is what I call
purely a matter of deforming edges relative to a fixed a “topological transition”, in which the keyframes are
lattice of vertices. When using isohedral tilings as a based on different Laves tilings altogether.15 How might
basis for constructing parquet deformations, allowing we build a parquet deformation from, say, 4.4.4.4
two keyframes to have different Laves tilings, or dif- (squares) to 3.3.3.3.3.3 (hexagons), or 3.4.6.4 (kites) to
ferent tiling vertex parameters, will necessarily require 3.6.3.6 (rhombs)? In my research on this subject I have
more sophistication. begun to articulate principles that make such trans
If two keyframes have the same Laves tiling but itions possible. Often, a transition can be achieved
different tiling vertex parameters, then interpolation by having an isohedral tile based on one Laves tiling
becomes less reliable. In principle we can still establish masquerade as another Laves tiling. We might adjust
a one-to-one correspondence between the vertices the tiling vertex parameters so that some edges shrink
of the two tilings and use that to smoothly interpo down to points, or deform edge paths until they meet
late vertex positions. Unfortunately, such compositions each other, subdividing tiles. I offer two specific exam-
are often stretched or skewed in unappealing ways. In ples of topological transitions in Figure 15; many others
some cases, however, this approach can succeed. In are possible. Note that by concatenating multiple
the parquet deformation of Figure 14, for example, half topological transitions and edge deformations, it may
of the tiling vertices remain fixed, while the other half be possible to build a parquet deformation whose end
flex around them. The fixed vertices lock the pattern in points are any two isohedral tilings.
place, permitting a stable composition to emerge.
Fig. 14: A computer-generated reproduction of “Dizzy Bee” by Richard Mesnik, 1964.

© HfG-Archiv/Museum Ulm, HfG-Ar, BDSA, Hu P 01. 023. This design demonstrates a case
where tiling vertex parameters can vary between two keyframes, but nevertheless produce a
stable composition.
Fig. 15: Two examples of topological transitions. The top transition moves from 3.3.3.3.3.3
(hexagons) to 4.4.4.4 (squares, here oriented diagonally). The bottom transition moves from
3.6.3.6 (rhombs) to 3.4.6.4 (kites).

6. Limitations and Opportunities in a basic design studio, color might best be studied in
Huff imposed narrow constraints on the structure of isolation.
parquet deformations. These constraints are entirely When drawing parquet deformations in software,
understandable given his interest in basic design and we can fill tiles with color with almost no extra work.
the need to offer a feasible exercise to students. Of Tactile’s representation of an isohedral prototile already
course, when freed from the classroom, and acceler includes an assignment of colors to tiles. Given two
ated by software, we can re-evaluate these constraints keyframes with different color assignments, interme-
and choose which ones to retain. diate tile colors can be interpolated even more easily
than edge shapes. Figure 16 gives an example of a
6.1 Color parquet deformation with colored tiles. I find that color
Huff intended parquet deformations to be an exercise adds a rich aesthetic dimension that could allow these
in the manipulation of shape and line. Filling tiles with drawings to be applied in a broader range of art and
color would have been a time-consuming distraction; design contexts.
Fig. 16: An example of a colored parquet

deformation, based on isohedral type
IH21.
6.2 Two-dimensional Deformations

Most of the designs produced in Huff’s studio showed general view that at every point in the plane, we can
a progression along one dimension of space. But some compute a set of “weights”, one for each keyframe.
students inevitably experimented with two-dimensional The weights must be nonnegative and sum to 1. A
developments, in which tile shapes evolved along given keyframe’s weight should be close to 1 in its
multiple axes. immediate neighborhood and fall off to 0 as we move
The keyframe-based approach I have developed away from it. The problem of computing continu
here can be extended to two dimensions in a natural ously varying weights from keyframe locations can be
way. A keyframe still consists of an isohedral prototile seen as an instance of scattered data interpolation in
assigned to a fixed position on the canvas. But rather computer science.16
than restricting keyframes to a single axis, we allow The second challenge is that in general, any
them to be placed anywhere on the canvas. In the final number of the weights computed above may be posi
drawing, we expect tiles near a keyframe to closely tive, potentially requiring us to blend between more
resemble that keyframe’s prototile, and to evolve than two paths for a given tiling edge. Most blending
smoothly between prototiles everywhere else. algorithms operate on exactly two paths, and must be
This approach entails two challenges. First, in generalized to handle this case. I extended the linear
the one-dimensional case, every edge midpoint falls blending I use in one dimension to handle arbitrary
somewhere between two keyframes, allowing us to numbers of paths and their weights.
compute the proportions of those prototiles to use Figure 17 shows two examples of two-dimensional
when blending. In two dimensions, there is no obvi- parquet deformations, one monochromatic and one
ous notion of betweenness. Instead we adopt a more colored.

Fig. 17: Two examples of
two-dimensional parquet
deformations.

6.3 Monohedrality and Handedness onohedral constraint: It produces a greater sense of
m
Huff required that a parquet deformation never contain visual unity and inevitability across a design.
“A and B elements”, two different interlocking tiles. Huff also prohibited “A and A′ elements”, by which
Expressing this constraint in a mathematically rigorous he meant tilings containing both a shape and its mir-
way is surprisingly difficult—every parquet deformation ror reflection, or equivalently, left- and right-handed
will necessarily include tiles of a wide variety of distinct versions of the same shape (he did allow tiles with
shapes! I interpret this rule to mean that a given design internal bilateral symmetry, which do not change un-
is like a coarse sampling of a space of monohedral til- der mirror reflection). I consider this restriction to be
ings, with prototile shapes that vary continuously too heavy-handed. Many designs produced in Huff’s
across the canvas. In principle we could query any studio contain lines of mirror symmetry that pass
point on the canvas and be shown the associated through bilaterally symmetric tiles, with an aesthetic
tiling. outcome comparable to that produced by reflections
Figure 13 has already illustrated that there is no between adjacent tiles. It is true that if a left-handed
mathematical impediment to drawing parquet defor- tile reflects directly across an edge to its right-handed
mations with two (or more) distinct shapes, arising counterpart, then that edge must be a straight line
from interpolation between inconsistent tile orien and hence a possible distraction. But we can con-
tations in keyframes. Of course, many of Escher’s struct many parquet deformations in which left- and
metamorphoses use multiple tile shapes. I have right-handed tiles are related by glide reflections
also previously experimented with Islamic geometric (combinations of reflections and translations), which I
patterns that evolve in the manner of parquet defor- find aesthetically pleasing. Figure 18 gives an example
mations,17 which almost inevitably involve more than of a design that uses glide reflections to good effect.
one shape. Jay Bonner’s chapter in this book discusses Indeed, a few drawings produced in Huff’s studio also
the connections to Islamic patterns in detail. When contain glide reflections, whether in conscious or un-
aiming to emulate Huff’s style, I tend to respect the conscious violation of this constraint.
Fig. 18: A topological transition from

3.3.3.3.3.3 (hexagons) to 3.3.4.3.4 (Cairo
tiles). The Laves tiles on the left and
right edges of the design are mirror
symmetric, but all other tiles come in
left- and right-handed pairs related by
vertical glide reflections.
6.4 Other Opportunities new realms. Some, such as three-dimensional designs,

The experiments presented here represent just a designs based on aperiodic tilings, or the fusion of par-
small sampling of the application of mathematics and quet deformations with Islamic geometric art, you will
compute science to the creation of parquet deforma- find elsewhere in this book. Others will inevitably arise
tions. Many opportunities await for incorporating new over time, by those seeking inspiration at the intersec-
mathematical ideas, or for translating these designs to tion of mathematics and art.

References
1
Douglas Hofstadter, Metamagical Themas: Questing for the Essence of Mind and Pattern,
Bantam Books, 1986.
2
Golan Levin and Tega Brain, Code as Creative Medium: A Handbook for Computational Art
and Design, The MIT Press, 2021.
3
Branko Grünbaum and G.C. Shephard, Tilings and Patterns, Dover, 2nd edition, 2016.
4
Craig S. Kaplan, Introductory Tiling Theory for Computer Graphics, Morgan & Claypool, 2009.
5
Doris Schattschneider, M.C. Escher: Visions of Symmetry, Harry N. Abrams, second edition,
2004.
6
Heinrich Heesch and Otto Kienzle, Flächenschluss: System der Formen lückenlos
aneinanderschliessender Flachteile, Springer-Verlag, 1963.
7
Grünbaum and Shephard, Tilings, chapter 6.
8
Available at: github.com/isohedral/tactile.
9
Frank Thomas and Ollie Johnston, The Illusion of Life: Disney Animation, Disney Editions,
1995.
10
Craig S. Kaplan, “Animated Isohedral Tilings", in: Susan Goldstine, Douglas McKenna, and
Kristóf Fenyvesi (eds.), Proceedings of Bridges 2019: Mathematics, Art, Music, Architecture,
Education, Culture, Tessellations Publishing, Phoenix, AZ 2019, pp. 99–106. Available online at
http://archive.bridgesmathart.org/2019/bridges2019-99.pdf.
11
Craig S. Kaplan, “Curve Evolution Schemes for Parquet Deformations”, in: George W. Hart
and Reza Sarhangi (eds.), Proceedings of Bridges 2010: Mathematics, Music, Art, Architecture,
Culture, Tessellations Publishing, Phoenix, AZ 2010, pp. 95–102. Available online at http://
archive.bridgesmathart.org/2010/bridges2010-95.html.
12
Hans Pedersen and Karan Singh, “Organic Labyrinths and Mazes, in: NPAR ‘06: Proceedings
of the 4th International Symposium on Non-photorealistic Animation and Rendering, ACM
Press, 2006: pp. 79–86.
13
Jeffrey J. Ventrella, Brain-Filling Curves—A Fractal Bestiary, Eyebrain Books, second edition,
2012. Available online at fractalcurves.com.
14
Craig S. Kaplan, “Metamorphosis in Escher’s Art”, in: Reza Sarhangi and Carlo H. Séquin (eds.),
Bridges Leeuwarden: Mathematics, Music, Art, Architecture, Culture, Tarquin Publications, London,
2008, pp. 39–46. Available online at http://archive.bridgesmathart.org/2008/bridges2008-39.html.
15
Ibid.
16
Richard Franke and Gregory M. Nielson, “Scattered Data Interpolation and Applications: A
Tutorial and Survey”, in: Hans Hagen and Dieter Roller (eds.), Geometric Modeling, Springer,
Berlin/Heidelberg, 1991, pp. 131–160.
17
Craig S. Kaplan, “Islamic Star Patterns from Polygons in Contact”, in: GI ‘05: Proceedings
of the 2005 Conference on Graphics Interface, Canadian Human-Computer Communications
Society, 2005: pp. 177–185.

Craig S. Kaplan is an associate professor in the David R. Cheriton School of Computer

Science at the University of Waterloo in Waterloo, Ontario, Canada. Originally from Montreal,
he earned a bachelor’s degree in Pure Mathematics and Computer Science from Waterloo
in 1996 and a PhD in Computer Science from the University of Washington 2002. Kaplan’s
research focuses on relationships between computer graphics, art, and design, with an
emphasis on applications to graphic design and illustration. He explores the mathematical
theories of symmetry and tilings as a means of creating patterns, as well as the perceptual
basis for our aesthetic appreciation of those patterns. This work frequently leads him into
the broader world of computer graphics, particularly non-photorealistic rendering. He also
occasionally ventures into related areas such as computational geometry and human-
computer interaction. Kaplan is the author of the short textbook Introductory Tiling Theory
for Computer Graphics, published by Morgan & Claypool, as well as numerous papers in the
Computer Graphics and Mathematical Art literature. He helps organize the annual Bridges
conference on art and mathematics, and serves as an Associate Editor of the Journal of
Mathematics and the Arts.

Parametric Modeling of Parquet
Deformations: A Novel Method
for Design and Analysis
Tuğrul Yazar
Introduction The Parquet Deformation exercise is generally

Introducing students of architecture to systems think attributed to William S. Huff, who introduced and con-
ing in the form of patterns and their deformations ducted it at several schools of architecture starting in
is becoming a significant research topic.1 Alejandro the 1960s. Huff defines the exercise as rooted in two
Zaera-Polo depicts a history of pattern studies in analytical disciplines; monohedral tilings in geometry,
architecture, emphasizing a conceptional shift in the and the continuous deformations in biological mor-
“Computational Design Era”.2 This conceptional shift phology, generally exemplified by D’Arcy Thompson’s
can be explained with the development of digital de- and Albrecht Dürer’s studies.7 Huff emphasizes its
sign technology, which makes it “possible to process temporal quality by describing the exercise as a form
multiple layers of patterns such as social, economical, of visual music with themes, events, intervals, rhythm,
cultural, formal, etc., and their deformations simul- and repetitions,8 and harmony of figuration,9 defining
taneously with desired precision”.3 In contemporary the designer’s role as being similar to that of a com-
architectural geometry, deformation is defined as “an poser. In his later writings, Huff mentions the tempo-
alteration of shape which is based on an underlying ral and spatial variation of Sino-Japanese landscape
mathematical principle”.4 Today, the static pattern of handscrolls10 and symmetry studies in chemistry11
configurations, tessellations, or any form of structural concerning the exercise. These relationships and defi-
order can be mediated into a system of both genera- nitions verify the artistic and scientific backgrounds of
tive and differentiated potential.5 This paper is about the exercise. To better investigate these backgrounds,
a special studio exercise of architectural education we can examine the academic environment in which
named “Parquet Deformation”. The exercise was the Parquet Deformation exercise has emerged.
derived from a special educational discourse in the William Huff’s educational discourse began with
mid-twentieth century and reached today with the his visits to the Ulm School of Design (The Hoch-
contributions of various researchers and instructors. schule für Gestaltung Ulm, HfG) as a graduate student
This paper aims to explain an alternative paramet- between 1956 and 1957, and as a visiting instructor for
ric modeling workflow and test it via re-constructing the Basic Course (Grundkurs) between 1963 and 196812.
typical parquet deformations. The Parquet Deforma- The educational doctrine of the HfG had an import-
tion exercise is about the tessellations of the plane ant role in the development of this exercise. HfG was
that gradually shapeshift in two dimensions without founded in 1946 by educationalist Inge Scholl and
gaps or overlaps (Figure 1, Figure 2, and Figure 3). The graphic designer Otl Aicher, with the aim of rebuild-
exercise encourages students of architecture to think ing the social structure of post-war Germany. The
about the geometric relationships between sequences school went through different periods in its short life.
of continual shape-shifting while developing reasoning The timeline at the HfG Ulm Archive website (www.
about a pattern as a structural whole. This is expect- hfg-archiv.ulm.de), produced from a review with Otl
ed to be explored by morphing the cells of a pattern Aicher in 1975, summarizes the periods HfG has gone
while sustaining its visual continuity and rhythm.6 through. In that diagram, William Huff’s name appears
This study is a continuation of a previous study on the under the course names “Value-based Design” and
re-construction of parquet deformations via contem- “Programmed Design” (Wertbestimmtes Design and
porary parametric modeling tools. Programmdesign). The first rector of the school was

Fig. 1: “Trifoliolate”, designed
by Glenn Paris in 1966.13
Redrawn by the author.
Fig. 2: “I at the Center”,

designed by David Oleson
in 1964.14 Redrawn by the
author.
Max Bill, who led the school’s claim to be the contin of the school after the resignation of Bill.16 However, the
uation of the Bauhaus in Weimar. According to Bill, timeline of the HfG Archive indicates that a Rector’s
design education should be a combination of science, College (Rektoratskollegium) was in charge; including
technology, and fine arts. As quoted by Peter Kapos; for Aicher, Gugelot, Vordemberge-Gildewart, and Maldo-
Bill, design work should proceed following the “spiritual nado. However, it is possible to indicate that Tomás
substance” of modern art.15 Maldonado became an influential figure in the new
William Huff’s first visit to HfG was during Bill’s period of the school after Max Bill. Unlike Bill, Maldo-
administration. However, it is possible to claim that the nado was more interested in production, consumption,
administration that Huff was more influenced and ben- exchange systems, and multi-disciplinary approach-
efited from intellectually was the period of “scientific es. He criticized Bill’s thoughts on aesthetics as the
operationalism” introduced by Tomás Maldonado. In natural basis of function. According to Maldonado,
some sources, Maldonado is introduced as the director function includes historicity, and it is possible to see it
128 Parametric Modeling of Parquet Deformations Tuğrul Yazar

Fig. 3: “Crossover”, designed
by Richard Long in 1963.17
Redrawn by the author.
in production-consumption systems. Peter Kapos sum- “Defining Basic Design as a Discipline”, Huff describes
marizes this idea as: “Building form, for example, would the influences of Maldonado, Albers, and Chernikov on
be principally determined by methods of prefabrication his “Formative Design Studio”.25 These confirm that the
and techniques of systematic construction. […] No lon- initial development of the Parquet Deformation exercise
ger directing production by decree according to artistic was one of the collective results of a special academ-
principles from an external position, it had become ical atmosphere. However, it was William Huff who
necessary for the designer to become fully integrated described the exercise clearly, conducted it continu-
within the production process.”18 Maldonado claimed ously at his design studios, and continued to publish his
that a designer’s success is related, among other opinions and studies long after the closure of the HfG.26
things, to the precision of thinking and doing methods, In 1983, Douglas Hofstadter published and com-
and the adequacy of scientific and technical knowl- mented on some of the parquet deformations in one
edge. During Maldonado’s influence, the educational of his journal papers and a chapter of the book titled
program of HfG has changed, to a new method called Metamagical Themas: Questing for the Essence of Mind
“scientific operationalism”. This method is also known and Pattern.27, 28 Hofstadter describes the temporal
as the “Ulm Model” at the international level. qualities of the student works, created in William Huff’s
The first-hand resource of the Ulm Model is the studios. Shortly after these publications, the exercise
Ulm Journal (Journal of the Hochschule für Gestal started to capture a wider multi-disciplinary interest.
tung), published between 1958–1968. The philosophical Another key personality in the development of the ex-
and pedagogical background of the Ulm Model can be ercise was Louis Kahn. Huff worked at Kahn’s office be-
studied from Maldonado and Bonsiepe’s paper in the tween 1958 and 1960. This coincides with his two visits
Journal.19 According to Peter Kapos; in this period, the to HfG, as a student and as a teacher. Hofstadter ad-
school opened new positions for cybernetics, game dresses their connection by quoting Kahn’s admiration
theory, and mathematics.20 Today, this interdisciplinary of Huff’s Basic Design discourse.29 Besides, Huff often
approach is still regarded as an original example for the mentions Kahn’s ideas on “order, disorder, change, and
discussion on the education of science and mathemat chaos” while explaining his point of view on the rela-
ics in the schools of architecture.21 tionships between geometry and design.30 Although any
There were other exercises at HfG that were de direct effect of Kahn on the particular studio exercise
signed for similar purposes with parquet deformation. is questionable, this connection indicates the theory of
For example, Herbert Kapitzki’s “Spatial Operations structuralism that influenced a generation of architects
in the Plane” (Räumliche Operation in der Ebene), and educators when studio exercises such as Parquet
“The Symmetry Exercise”, and Maldonado’s “Raster” Deformation were developed.31
approach similar educational purposes from differ It is possible to find pattern studies similar to
ent angles.22 Several design exercises including “The parquet deformations in various fields of art since the
Programmed Design”, “The Conflicting Depth Cues”, beginning of the twentieth century. One of the earliest
“The Figure-Ground Figure without Ground”, and “The examples of systematic pattern deformations can be
Parquet Deformation” are described as derivations of seen in Lewis Foreman Day’s studies on textile orna-
Maldonado’s previous studio experiments.23 In an article ments in his books titled Pattern Design, A Book for
in the Ulm Journal, Huff explains Maldonado’s influence Students Treating in a Practical Way of the Anatomy,
on his exercise designs.24 Moreover, in the article titled Planning & Evolution of Repeated Ornament and The

Anatomy of Pattern. Day was a painter and indu to figure out a transformation rule by considering the
strial designer, and an important figure of the Arts symmetry operations. The transition from one shape to
and Crafts movement. In these books, Day does not another requires a repeated modification on the edges
directly address or name pattern deformations, but and vertices. In most examples, the continuous shape
introduces various patterns by drawing several varia- shifting starts from a simpler shape on the left side
tions near each other, denoting them “varieties of the and becomes increasingly complex while deforming to
same pattern”.32 The most well-known and mentioned the right side. In some of the designs, it is possible to
examples of the systematic and methodical deform identify several steps, or phases, where the deforma-
ation of patterns are the artworks of M. C. Escher, tion process changes. In this study, such key moments
especially Metamorphosis and The Day and Night.33 In will be identified and used as the geometric base of
Metamorphosis, Escher created many deformations of the parquet deformations. The major components of a
a square tessellation and manipulates the image by us- parquet deformation can be defined as follows:
ing shading and color.
The recent studies on parquet deformations are ⟩ Lattice: Lattice represents the underlying structure
mostly based on geometric principles. Craig Kaplan of a parquet deformation. The term is also used by
studied the mathematical underpinnings of pattern Huff while describing the types of tessellations that
deformations in his dissertation and various publi- a parquet deformation might be constructed onto.39
cations.34 He also studied Escher’s artworks and the In the original exercise, lattices are generally mono-
continuous deformations of Islamic patterns.35 Parallel hedral (Figure 4).
to Huff’s exercise, John Sharp studied similar compo-
sitions he named Morphing Tilings.36 In 2006, Andrew ⟩ Prototile: A prototile is a key cell that the parquet
Cooke developed BulliEpu, a computer application in variations are generated from. A regular systemat-
Java that generates parquet deformations. Although ic parquet deformation has at least two prototiles,
the Parquet Deformation and similar exercises are from which the parquets are generated while the
still being conducted in design studios, the academic shapeshifting from one prototile to another (Figure 4).
studies on the topic remain limited. Karen Li is one
of a few former students of Huff who succeeded her In this study, more components are added to better
background into an academic publication about the identify the proposed methodology:
exercise.37 For further reading and bibliography, read
ers are referred to David Bailey’s website (http://www. ⟩ Fundamental Curve: In most cases, the proto-
tesselation.co.uk/parquet-deformations) which shows tiles of parquet deformations are symmetrical. This
one of the most comprehensive sources on the litera- reduces the problem into the deformation of one
ture of Parquet Deformation exercise. or several curves. In this study, these curves will
be defined as the “fundamental curves” since they
Methodology contain the minimum information about a parquet
This study is a continuation of the author’s previous deformation. In the shapeshifting of a fundamental
study38 on the contemporary approaches for the para curve, new vertices and edges can be created (Fig
metric modeling of parquet deformations. An original ure 4). If the transformation process of a fundamen-
generative design approach and a computer tool will be tal curve is examined closely, it can be seen that
introduced and tested. As the testing case studies, two generally, there are linear interpolations between the
student works supervised by William Huff will be ana- vertices of these curves. During these interpolations,
lyzed and reconstructed by utilizing the aforementioned the appearance and disappearances of the vertices
tool. This is expected to help provide a framework for a can be observed.
contemporary interpretation of the exercise and its po-
tential connections with today’s popular design tools and ⟩ 3D Prototile: In the proposed approach, a three-
methods. This framework is believed to be an example of dimensional abstract object is created digitally, to
how today’s architects can use computer-aided tools not embed the variations of many two-dimensional
only in drawing but also in design exploration. prototiles. This object includes the sequential inter-
Since the first students of William Huff at Ulm had polations between the prototiles, represented in the
no computer access, these geometric computations XYT space (Figure 4d). Every section taken from this
were probably made by extensive work of sketching object would return a parquet variation. The design
on tracing papers. These compositions are generally of the 3D Prototile should be done depending on
based on a regular tessellation (square, triangular, or the underlying lattice, as in the original parquet de-
hexagonal grids). It would be possible for the students formations. This would require sketch modeling and
to use several layers of tracing papers while trying feedback, just like the original design process.

Fig. 4. Anatomy of a typical Parquet
Deformation (“I at the Center”). Image
by the author
arametric or explicit forms, or they can be created

p
⟩ Variation Space: After defining the 3D Prototile, it with different approaches such as vector fields.
is copied on the lattice, by specified numbers. This The way these surfaces are created and the data
three-dimensional waffle-like structure is named sources they use will provide a controlled environ-
“variation space” in this study. This set, as the name ment for the designer. Any digital input and surface
suggests, contains all the possible parquet deform creation method would be related to the design
ation compositions that can be derived from given process of a parquet deformation (Figure 4).
prototiles and fundamental curves (Figure 4).
The method described above provides fast results by
⟩ Deformer Surface: After the preparations ex- utilizing the basic functions of CAD such as surface
plained above, design outputs can be produced. modeling, intersection, and projection operations,
Projection of each section taken from the variation without revealing the complex mathematical calcula
space on the XY plane would generate a two-di- tions underneath. The deformer surface can be defined
mensional parquet deformation design. If sufficient independently from the lattice, prototile, and 3D proto
attention has been paid to the relationship with the tile. Therefore, the data source that constitutes the
lattice during the prototile design phase, no gaps deformer surface can be defined independently. More
or intersections would occur between the inner over, the geometric properties of the deformer surface
surfaces of the variation space. Thus, the sections would give information about the parquet deformation
produced from this structure wouldn’t contain any it creates. These properties include topological connect
gaps or overlaps. In most of the original examples, edness, draft angle, curvature, and critical points.40 For
the deformer surface is usually an inclined flat example, the Gaussian curvature of a deformer surface
plane. This enables a parquet deformation to represents a mathematical basis for explaining the lo-
progress at a steady rate of change, being evenly cal and overall regularities of a parquet deformation. If
distributed among the prototiles. However, the the Gaussian curvature of a deformation surface is zero
method proposed in this study makes it possible at every point, the surface is flat. This would result in a
to experiment with various surfaces with different parquet deformation in which the amount of change in
curvatures. If the surface is derived from a poly- every step of the deformation is constant. If the defor-
nomial function, this would sustain the surface mation surface has no Gaussian curvature and it is par
continuity, and the rate of change of the derived allel to the projection plane of the pattern, then there
parquet deformation will be more consistent would be no deformation on the pattern. While moving
and connected. These surfaces can be derived in across two points on a deformation surface, a positive

or negative change in the Gaussian curvature results in mation add-on includes two prototype components for
a variable rate of change in the deformation. Similarly, testing purposes.
the projection of the critical points on the deformer The add-on does not aim to generate random
surface would give the center points or attractor points parquet deformations automatically. Instead, it aims
of a parquet deformation. More studies are needed to to support designers in creating and analyzing these
utilize these properties of the deformer surfaces. compositions by solving several geometric construction
problems which are the essential parts of the meth-
Toolset odology presented in the previous section. This is why
The methodology explained above belongs to the the components of the add-on are not closed systems,
design process of a generalized parquet deformation. It but rather small problem-solving tools, intended to be
doesn’t represent a fully automated design process. As used in regular Grasshopper algorithms. The compo-
in the original compositions, a sketching process with nents were developed in Python Scripts and Grasshop
a feedback cycle is still necessary. However, since the per clusters and added to Grasshopper as user objects.
proposed method includes digital data and three-di- In the future, these prototype components would be
mensional digital models, it is not possible to perform further developed with C#.
it by using hand tools. In turn, it is better suited to be The first component, Loft Aligned, aims to help
fed by digital data, thus being used in generative design designers in the construction of the 3D prototiles. The
systems. In this study, it was decided to develop a main problem of this geometric construction is the
simple computer tool to increase the usability of this determination of the transitions between the vertices of
method and to facilitate the modeling process. Cur- the fundamental curves. Figure 5 shows that the regular
rently, the tool is being developed for the Grasshopper surface-creating commands such as loft, and sweep
in Rhinoceros Computer-aided Design (CAD) software. does not have enough inputs to automatically create
Grasshopper is a visual programming language special- such surfaces. It is possible to model these surfaces
ized in generating algorithmic designs and the analysis manually in CAD by adding more vertices. However, the
of geometric forms. The dataflow programming con- Loft Aligned component presents a general solution
cept is utilized in Grasshopper by introducing graph-like and helps designers in this modeling process. It takes
algorithms, composed of components (nodes), and the two input curves from the user, analyzes them, and
connections between the components (directed edg- matches their vertices so that they correctly create the
es). The algorithms created in this environment enables surfaces necessary for a 3D prototile. The vertex match-
designers to construct and explore geometric forms ing algorithm is derived from the observations made
in real-time. Grasshopper also offers an infrastructure on the original parquet deformations. The component
open to enrichment with new components prepared by can generate new vertices, if necessary. It is an essen-
the experts of the field. Currently, the Parquet Defor- tial phase of the methodology presented in this study
Fig. 5: The Loft Aligned component’s functionality and difference from a standard surface
modeling command, especially in protrusions. The object on the left is the result of a stan-
dard Loft operation in Rhinoceros. Since the command does not have enough information
about the general shape-shifting style of parquet deformations, it adds random edges along
the surface. The object on the right is the result of the Loft Aligned component. This com-
ponent calculates naked and overlapping vertices and generates new edges according to the
general rule of orthogonal foldings in parquet deformations. Image by the author.

because the outputs of this component return two be tested on the reconstruction of two original parquet
sets of points that are aligned and ready to be lofted as deformations.
a surface geometry. Currently, the Loft Aligned com-
ponent is under the component group of the add-on Strange Start, Startling Stop
named Utilities. “Strange Start, Startling Stop” was designed by stu-
The second component is called “p31m” and is dent Mary Purdy at the State University of New York in
under the component group named Symmetry. This 1985. The composition is based on a hexagonal lattice
component helps designers to explore not only parquet (Figure 6). There are four prototiles, marking the four
deformations but many pattern systems by enabling key moments in the shape-shifting process. The first
the application of symmetry groups. Currently, only one prototile is a regular hexagon, which is also the first tile
symmetry group called p31m was developed for testing of the composition. This prototile morphs into a shape
purposes. This component is developed as a native that is a composition of four smaller hexagons, creating
Grasshopper cluster, which means it is still under the the second prototile. This morphing process finishes
development phase. It takes geometric objects to be at the 9th tile from the left. The third prototile is then,
transformed, and two points to determine the size of slowly erected by enlarging a portion of every edge. This
the hexagonal lattice of the symmetry transformation. morphing process finishes at the 17th tile from the left.
These inputs are organized to match the outputs of the The fourth and last prototile is a continuation of the
Loft Aligned component so that an efficient parquet previous one. It expands the parts and creates snow-
deformation modeling workflow could be established flake-like spikes to finalize the composition.
quickly. In the future, the other symmetry groups could This final prototile expands the edges and changes
also be implemented to the add-on. In the next section, their directions, and the opposing edges merge into
the tool and the approach introduced in this section will single vertices at the end. These prototiles are created
Fig. 6. “Strange Start, Startling Stop”.

Redrawn by the author. The original
drawing made by the student, Mary
Purdy. © HfG-Archiv/Museum Ulm,
Fig. 7. “Strange Start, Startling Stop”.

(a) Fundamental curves, (b) prototiles,
(c) the 3D prototile, and (d) several
slices of the 3D prototile, creating the
tiles of the deformation. Image by the
author

by four fundamental curves. The fundamental curves utilizing the p31m symmetry operation. This operation
and corresponding prototiles are shown in Figure 7a repeats several rotation and reflection transformations
and Figure 7b. The designer would probably think of to create the closed polyhedron.
these four prototiles while designing “Strange Start, The geometric construction described above is test-
Startling Stop”. One of the important qualities of ed in Grasshopper (Figure 8). This algorithm can also be
parquet deformations is that they allow partial defor- regarded as a generalization, which means, it can gen-
mations on single tiles. In the example, the 10th and erate not only the particular case study but many other
18th tiles from the left show this kind of partial defor- alternative designs by providing it with different funda-
mations. The proposed methodology is expected to mental curves. The first part of the algorithm (Figure 8a)
simulate this behavior. After the definition of prototiles, takes the fundamental curves of the four prototiles.
their vertices are matched to create the lofting oper Then, these curves are fed to the Loft Aligned com-
ation (Figure 7c). Then, the 3D prototile is created by ponent to match their correct vertices. The outputs P1
Fig. 9. New interpretations of “Strange

Start, Startling Stop”, created by dif-
ferent sections of the variation space.
Image by the author.
Fig. 8. Parametric model of the variation space of “Strange Start, Startling Stop”. (a) Utilizing
the “Loft Aligned” component to align the vertices of the fundamental curves. (b) The appli-
cation of p31m symmetry transformation to create the 3D prototile. (c) The array of the 3D
prototile on the hexagonal lattice, creating the variation space. Image by the author

and P2 are used to draw line segments of the matching continuous deformer surface would result in a continu-
vertices on the prototiles. In the second part (Figure 8b), ously varying set of parquets. The rate of change of the
these line segments are used to create a surface shape deformation is related to the change in the curvature
by using regular modeling components of Grasshopper. of the deformer surface. Also, the local minimum and
Then, the symmetry transformation is applied to this maximum points of the deformer surface mark the
surface to create the 3D prototile. In the last part (Fig points or regions where the parquet deformation is
ure 8c), a hexagonal grid is created, and the 3D proto- attracted to Wiry Wonder.
tile is copied to this grid to finalize the variation space. The parquet deformation named “Wiry Wonder”
The result of this algorithm is a waffle-like structure of was designed by Michael Cuttita in William Huff’s
many copies of the 3D prototile, capturing all possible studio at the State University of New York in 1989.
parquet deformations that could be generated from Figure 10 shows the original drawing made by the
the given fundamental curves. student. Similar to the previous one, this compo-
In the last part of the reconstruction, various sition is based on a hexagonal lattice. In this compo-
deformer surfaces are used to cut the variation space sition there are three major prototiles, marking the
and generate new alternatives. In general, Huff’s studio three key moments in the shape-shifting process.
works utilize a continuous deformation on either one There is also one hidden prototile, which is created
or two axes. The methodology presented in this study to accommodate the transition between the last two
adds many different ways to it. Figure 9 shows two of prototiles. The first prototile is similar to the second
these re-constructions. One of the main qualities of prototile of “Strange Start, Startling Stop”, composed
parquet deformations is the seamless continuity of of four small hexagons.
all parquets. This is why it is possible to see variations This prototile morphs into a shape that has exten-
even within a single parquet. The proposed methodol- sions inside and outside, creating the second prototile.
ogy enables this quality to be implemented since any This morphing process finishes at the 8th tile from the
Fig. 10. Michael Cuttita, “Wiry Wonder”.

Basic Design Studio of William S. Huff, Spring
1989, SUNY at Buffalo. © HfG-Archiv/Museum Ulm,
Fig. 11. “Wiry Wonder”. (a) Fundamental curves,

(b) prototiles, (c) the 3D Prototile, and (d) several
slices of the 3D prototile, creating the tiles of the
deformation. Image by the author

Fig. 12: Parametric model of the variation space of “Wiry Wonder”: (a) utilizing the Loft
Aligned component to align the vertices of the prototiles, and the utilization of a hidden
prototile to achieve the transition between prototiles, (b) the application of p31m symmetry
transformation to create the 3D prototile, (c) the array of the 3D prototile on the hexagonal
lattice, creating the variation space. Image by the author.
left. The third and the last prototile creates saw-like has four fundamental curves, instead of three. The
details on the newly created edges in the second pro- hidden prototile is the extended version of the second
totile. This morphing process finishes on the right end. prototile, creating a basis for the transition between the
These prototiles are created by fundamental curves, second and third prototiles. The operations shown in
shown in Figure 11a. It is the same symmetry operation Figure 12a use the Loft Aligned component twice. One
with “Strange Start, Startling Stop”, that creates the of them is the actual matching of the vertices of the
prototiles shown in Figure 11b. After the definition of fundamental curves. The other component creates a
prototiles, the Loft Aligned component developed for temporary matching to add the information coming
this study was used to match the vertices of the proto- from the hidden prototile. A parametric evaluation was
tiles. This returns the one-to-one morphings of the line required to transfer this information to the actual ver-
segments of the fundamental curves, creating the 3D tex-matching process. match their correct vertices. In
prototile (Figure 11c). the second part (Figure 12b), a surface shape was cre-
Figure 12 shows the Grasshopper algorithm to ated by using regular modeling components of Grass-
generate the setup of “Wiry Wonder”. The first part hopper. Then, the symmetry transformation is applied
of the algorithm (Figure 12a) takes the fundamental to this surface to create the 3D prototile. In the last
curves of the three prototiles. The difference of this part (Figure 12c), a hexagonal grid is created, and the
composition from the previous one is the usage of a 3D prototile is copied to this grid to finalize the varia-
hidden prototile curve, which is between the second tion space. The result of this algorithm is a waffle-like
and third prototile. When analyzed, the transition from structure of many copies of the 3D prototile, captur-
the second prototile to the third prototile includes two ing all possible parquet de-formations that could be
different transformations. One of them is the extension generated from the given fundamental curves. Finally,
of the protrusions generated in the second prototile. various deformer surfaces are used to cut the variation
These protrusions are becoming much longer while space and generate new alternatives. Figure 13 shows
approaching the third prototile. At the same time, an two of these reconstructions.
other transformation occurs. This second transforma-
tion creates saw-like edges on the protrusions. This is Conclusion and Discussion
why the transition between second and third prototiles Architectural education must continue to redefine its
is a composite transformation. Since the current code interdisciplinary role with the help of newly develop-
of the Loft Aligned component can only process one ing technologies. This study focused on a historical
transformation at a time, this required the definition of studio exercise that is still valid for today’s paramet-
a hidden prototile. This is why the Grasshopper code ric modeling approaches. A novel methodology was

Fig. 13: New interpretations of “Wiry
Wonder”, created by different sections
of the variation space. Image by the
author.
introduced and tested. Different from the traditional new generation of designers trained in this field. More
two-dimensional drawing, the usage of a three-di- studies and studio experiences are needed to clarify
mensional variation space and a deformer surface these potential uses.
was introduced as visual ways of making complex One of the common starting points of many digital
calculations in the creation of parquet deformations. design theories is based on the relationship between
The toolset developed for this purpose should be complexity and harmony. In the broadest sense, most
further advanced and better tested in the future. In the computer-aided design tools are developed to help
proposed toolset, fundamental curves, and symmetry designers deal with the increasing complexity of design
groups were defined as the primary components of problems efficiently, and open new ways of study-
parquet deformations. The methodology presented in ing harmony. On the other hand, it should be noted
this study helps designers in executing complex and that the intellectual basis of these methods has been
repeating geometric constructions of parquet defor- establishing before the development of computer
mations. When the geometric construction of such technologies. The Parquet Deformation exercise is a
patterns is no longer a problem, it can be enhanced by perfect example of the non-computerized basis of a
additional inputs of any design domain. The code pack- contemporary digital design method, called paramet-
age being developed can be used as an example in the ric modeling. Although the Ulm Model has no major
current parametric modeling and coding-related design impact on today’s architectural theories, the element
courses and transformed with different data inputs. that still makes the Parquet Deformation exercise
This would break the limitation of the tool and widen useful and meaningful is the bridge it hosts between
the educational perspective. It would also be appropri- art and mathematics, similar to one of the prominent
ate to consider hybrid studio setups in which both tra- intentions of parametric modeling. Both parquet de-
ditional and digital tools are used to better understand formation and parametric modeling are the reflections
the underlying mathematics and to speed up creative of modern thought in science, art, and philosophy, in
thinking at the same time. It is not possible to state which a choreography of continuity, flow, dynamism,
that the developed code can automatically generate all and evolution are at the forefront.
possible parquet deformations. Human creativity can The modeling approach presented in this paper
always get out of the box and reveal unique patterns of increases the speed of formal exploration. Using
thought that cannot be predicted by pre-made code this method, a parquet deformation that would take
sequences. In this respect, the code presented in this several hours to be designed and drawn by hand
article is no different from a compass. It can be used can be d iscovered in minutes. However, speed is not
in new ways for different purposes in the hands of the necessarily a positive factor for design exploration.

Appearing to be a major advantage, at first sight, it is believed to emphasize the power of keeping d iverse
also esoteric and limiting from a wider perspective. attitudes together and keeping communication chan-
Besides, the inevitable result of this point of view is nels open between them.
the necessity of digitalization. Again, the digitization In Basic Design education, it is possible to see that
of architectural information raises the ontological and the constructivist perspective still lays the ground-
epistemological criticisms of reduction. This criticism work for short-term and contextually limited studio
is based on the fact that architectural knowledge exercises. But even Constructivism itself is not a
cannot tolerate reduction. Similar to the reduction of broad enough framework to include all the possibil-
architecture into “building science” in HfG, this educa- ities of architectural education. The contextual lim-
tional experience showed us that it is not possible to itations proposed by Constructivism should not turn
leave architectural education to a single point of view. into reduction. For this reason, these exercises should
If a lesson is to be learned from the history of HfG, be carefully linked to the general objectives of con-
it would be the importance of addressing different temporary architectural education and professional
perspectives from a holistic and pluralist perspective, responsibilities. The integration of these exercises with
and being able to stay together by understanding the the interdisciplinary roles of the architect continues to
value of those differences and benefiting from all. be one of the issues of every curriculum, studio, and
This is why the methodology presented in this study instructor. It is an open question how the future design
can be regarded as an alternative aiming to enrich the studios will take over this issue from the point left by
constructivist pedagogy of parquet deformations. It is William S. Huff and the parquet deformation.
Acknowledgments
The author would like to thank Werner Van Hoeydonck for providing the original drawings of
the Parquet Deformations which were analyzed in this study.
References
1
Tuğrul Yazar, “Revisiting Parquet Deformations from a Computational Perspective: A Novel
Method for Design and Analysis”, International Journal of Architectural Computing 15(4), SAGE
Publishing, 2017, pp. 250–267.
2
Alejandro Zaera-Polo, “Patterns, Fabrics, Prototypes, Tessellations”, Architectural Design 79,
Wiley, United Kingdom, 2009, pp. 18–27.
3
Rivka Oxman and Robert Oxman, “The New Structuralism: Design, Engineering, and Archi-
tectural Technologies”, Architectural Design 206, Wiley, United Kingdom, 2019, pp. 15–24.
4
Helmut Pottman, Andreas Asperl, Michael Hofer, and Axel Killian, Architectural Geometry,
Bentley Institute Press, United States, 2007, p. 451.
5
Mark Garcia, “Prologue for a History, Theory, and Future of Patterns of Architecture and
Spatial Design”, Architectural Design 79, Wiley, United Kingdom, 2009, pp. 6–17.
6
Yazar, “Revisiting Parquet Deformations”, pp. 250–267.
7
William S. Huff, “The Parquet Deformation”, Best Beginning Design Projects, The University
of Wisconsin-Milwaukee, United States, 1979, pp. 30–33; D’Arcy Wentworth Thompson, On
Growth and Form, Cambridge University Press, United States, 1945 edition, pp. 1026–1095.

8
Huff, “The Parquet Deformation”, pp. 30–33.
9
William S. Huff, “On Regulation and Hidden Harmony”, Harmony of Forms and
Processes: Nature, Art, Science, Society, International Society for the Interdisciplinary
Study of Symmetry, Ukraine, 2008.
10
William S. Huff, “The Landscape Handscroll and the Parquet Deformation”, Katachi U
Symmetry, Springer-Verlag, Japan, 1996, pp. 307–314.
11
William S. Huff, “Ordering Disorder after K. L. Wolf”, Forma 15, Proceedings of the 2nd
International Katachi U Symmetry Symposium, Scipress, Japan, 2000, pp. 41–47.
12
William S. Huff, “Defining Basic Design as a Discipline”, The Quarterly of the
International Society for the Interdisciplinary Study of Symmetry: Symmetry, Art and
Science 2, ISIS-Symmetry, Belgium, 2002, pp. 91–98; Carnegie Mellon University, “William
Huff, Buildings by Pedagogs”, exhibition catalog, Carnegie Institute, United States,
1965, p. 8; William S. Huff, “An Argument for Basic Design”, Journal of the Ulm School
for Design 12–13, Germany, 1965, pp. 25–38; Dénes Nagy, “Architecture, Mathematics,
and a Symmetric Link Between Them”, The Quarterly of the International Society for
the Interdisciplinary Study of Symmetry: Symmetry, Art and Science 2, ISIS-Symmetry,
Belgium, 2002, pp. 31–64.
13
Ibid.
14
15
Peter Kapos, “Art and Design: the Ulm Model”, exhibition at Raven Row, London, 2016,
http://www.ravenrow.org/texts/83/, accessed: 28 January 2021.
16
Isabel Clara Neves and João Rocha, “The Contribution of Tomás Maldonado to the
Scientific Approach to Design at the Beginning of Computational Era, The Case of The
HFG of Ulm”, Future Traditions: 1st Regional International Workshop, ECAADE, Portugal,
2013, pp. 39–50; Isabel Clara Neves, João Rocha, and José Pinto Duarte, “Computational
Design Research in Architecture: The Legacy of the Hochschule für Gestaltung, Ulm”,
International Journal of Architectural Computing 12(1), SAGE Publishing, 2014, pp. 1–25.
17
18
Kapos, “Art and Design”.
19
Tomás Maldonado and Gui Bonsiepe, “Science and Design”, Journal of the Hochschule
für Gestaltung, Ulm 10/11, Germany, 1964, pp. 10–29.
20
Kapos, “Art and Design”.
21
Cornelie Leopold, “Precise Experiments: Relations between Mathematics, Philosophy
and Design and Ulm School of Design”, Nexus Network Journal, Architecture and
Mathematics 15(2), Kim Williams Books, Italy, 2013, pp. 363–380.
22
Yazar, “Revisiting Parquet Deformations”, pp. 250–267; Huff, “An Argument for Basic
Design”, pp. 25–38; Neves and Rocha, “The Contribution”, pp. 39–50.
23
William Huff, “Students’ Work from Basic Design Studios of William S. Huff”, Intersight
1, The Journal of the School of Architecture and Planning, University of Buffalo, United
States, 1990, pp. 80–83.

24
Huff, “An Argument for Basic Design”: pp. 25–38.
25
Huff, “Defining Basic Design”, pp. 91–98.
26
27
Douglas Hofstadter, “Parquet Deformations, A Subtle, Intricate Art Form”, Metamagical
Themas: Questing for the Essence of Mind and Pattern, Basic Books, United States, 1983,
pp. 191–199.
28
Douglas Hofstadter, “Parquet Deformations, Patterns of Tiles that Shift Gradually in One
Dimension”, Scientific American Magazine, Springer Nature, United States, 1983, pp. 14–20.
29
Hofstadter, “A Subtle, Intricate Art Form”, pp. 191–199.
30
31
32
Lewis Foreman Day, The Anatomy of Pattern, B.T. Batsford, United Kingdom, 1887, p. 16;
Lewis Foreman Day, Pattern Design, A Book for Students Treating in a Practical Way of the
Anatomy, Planning & Evolution of Repeated Ornament, B.T. Batsford, United Kingdom, 1915,
pp. 28–47.
33
Craig Kaplan, “Curve Evolution Schemes for Parquet Deformations”, Proceedings of Bridges
Conference, Mathematics Music Art Architecture Culture, Tessellations Publishing, Hungary,
2010, pp. 95–103; Hofstadter, “A Subtle, Intricate Art Form”: pp. 191–199; Elaine Krajenke
Ellison and John Sharp, “Tiled Torus Quilt with Changing Tiles”, Proceedings of Bridges Con-
ference, Mathematics Music Art Architecture Culture, Tessellations Publishing, Hungary, 2010,
pp. 67–74; Yazar, “Revisiting Parquet Deformations”: pp. 250–267.
34
Craig Kaplan, “Computer Graphics and Geometric Ornamental Design”, Ph.D. dissertation,
University of Washington, Seattle, 2002, pp. 75–76, 208–212; Craig Kaplan, “Metamorphosis in
Escher’s Art”, Proceedings of Bridges Conference, Mathematical Connections in Art, Music, and
Science, Tarquin Publications, The Netherlands, 2008, pp. 39–46.
35
Craig Kaplan, “Islamic Star Patterns from Polygons in Contact”, Proceedings of Graphics
Interface GI’05, Canada, 2005, pp. 177–185.
36
Ellison and Sharp, “Tiled Torus Quilt”, pp. 67–74.
37
Karen Li, “Programmed Design, The Systematic Method and the Form of Pattern”, The
Quarterly of the International Society for the Interdisciplinary Study of Symmetry: Symmetry,
Art and Science 1–4, ISIS-Symmetry, Belgium, 2002, pp. 85–89.
38
39
Huff, “The Parquet Deformation”, pp. 30–33.
40

Tuğrul Yazar is an architect and computational design researcher. He received an M.Arch de-
gree in 2003, and a PhD in 2009 in the field of Computer-Aided Architecture. His dissertation
is a computational perspective on the short-term design exercises conducted in the early
years of architectural education. He has published on architectural geometry and design
technologies. In a recent publication, he explained the methodology of drawing Bézier curves
and B-Splines using only a compass. Since 2010, he has taught Architectural Geometry,
Design Mathematics, Parametric Modeling, and Digital Fabrication at İstanbul Bilgi University
Faculty of Architecture. In the Parametric Modeling and Design Mathematics courses, he has
been teaching Grasshopper and Python languages to design students. In the Digital Fabri-
cation course, he is introducing robot technologies to design students. In addition to these
technical courses, he has also been a studio instructor at the first-year computation-based
basic design studios of the same faculty. This studio is based on a combination of compu-
tational thinking and physical material performances. Between 2010 and 2018 he shared a
portion of his studies on his blog at designcoding.net. In 2016, he published the first Turkish
book in this field, Parametric Modeling with Grasshopper. In a recently completed scientific
project, he was a member of a team that developed specialized computer software based on
the principles of space syntax theory. In a current scientific research project, he is studying
the effects of different geometric qualities on the physical performances of rammed-earth
structures. Apart from the academic studies, he is a computational design consultant and a
workshop tutor. The works of the POTPlus design/research group, which he established with
Fulya Akipek, have been exhibited in various exhibitions and biennials. In 2018 their rammed-
earth structure exhibited at the Antalya Architectural Biennial received the Sustainable
Architecture (S-ARCH) Award.
Pattern Manipulation
through Hinged Tessellations
Jay Bonner
Introduction corollary pattern with 7-pointed stars that makes use

In 1948, Buckminster Fuller discovered polyhedral of the same triangular faces and rhombic interstice
“jitterbug” transformations, whereby specific vertices of regions (Figure 5).
a uniform polyhedron act as hinges, allowing the faces This chapter expands upon this two-dimensional
of the polyhedron to rotate in a progressively increas corollary by examining the use of progressively rotating
ing angle.1 This progressive angular increase causes hinged tessellations to modify existing geometric pat-
the overall size of the polyhedron to expand. At the terns and, with greater and greater angles of rotation,
zenith of this expansion, the continued rotation of the creating altogether new designs.
faces draws them gradually back together through the When asked by Werner Van Hoeydonck to contrib-
decreasing angles of the opposite side of each rotating ute a chapter for this publication, and knowing that a
vertex, thereby shrinking the size of the polyhedron prime focus of this book would be the pioneering work
until it is back to its original form. In 2018, I published of William S. Huff and others who have contributed so
a paper titled “Doing the Jitterbug with Islamic Geo- significantly to the remarkable study of parquet defor-
metric Patterns” in the Journal of Mathematics and mations, including the works of my co-authors of this
the Arts.2 This paper described how the geometry of book, I chose to write about the metamorphic potential
jitterbug transformations can be used to apply geomet of pattern manipulation that can result with the use
ric designs with Islamic characteristics onto spherical of hinged tessellations. Even if somewhat tangential to
surfaces, and by setting the degree of rotation of the the primary subject of parquet deformations, I believe
transformation to conform with the angles associated my contribution is nonetheless relevant, if for no other
with specific polygons, one can create spherical geo- reason than the fact that the groundbreaking and
metric designs that exhibit the uniform distribution of beautiful works of such luminaries as Huff, Hofstadter,
unexpected regular star forms. In that paper, I demon- and Escher3 have directly inspired my exploration of
strate two examples: a design with a uniform distribu- hinged tessellations as a means of creating new geo-
tion of regular 7-pointed stars; and one with regularly metric designs.
placed 13-pointed stars. Each of these employs the My focus upon hinged tessellations as a means of
jitterbug transformation of the octahedron, with the creating geometric designs through the gradual vertex
eight triangular faces rotated to angles associated with rotation of the polygonal modules of a tessellation is in
the heptagon and tridecagon, respectively. Rotating marked contrast to the gradual modification of a grid
the octahedron’s triangles to these specific angular through the process of parquet deformation. Parquet
conditions creates interstice regions comprised of isos- deformations involve a spatial evolutionary metamor-
celes triangles that are edge to edge with one another phosis that is either one-dimensional (linear) or two-
upon their short edges. The spherical design is created dimensional (e.g., radial) in symmetrical makeup. The
by applying carefully constructed pattern lines to the changes in angular growth of hinged tessellations are
triangular and isosceles triangular faces and projecting more a function of time than space and are best rep-
these decorated faces to the surface of the sphere resented through animation, such as examples by the
(Figure 4b). On the two-dimensional plane, the same Dutch geometric artist Rinus Roelofs,4 rather than fixed
edge-to-edge isosceles triangles produce a rhombus, illustrations. However, for the medium of the printed
and this paper also demonstrates a two-dimensional page I am herein representing this process of seamless

angular expansion as a series of isolated, or one might points for these rotational changes remain fixed upon
say frozen sequential angular intervals. These illustrate the midpoints of the formative tessellation throughout.
the gradual metamorphosis of the initial tiled design As these changes move from left to right, the angles
with 0° rotation of the polygonal module. (Of course, of the pattern move from being more acute to be
it is worth noting that it is entirely possible to also coming more obtuse. The progressive design in Figure
represent the gradual changes that result from parquet 2 employs two points upon each polygonal edge of
deformations through the vehicle of animation.) the formative tessellation and starts with the forma-
tive tessellation itself (A) before rotating the edges
Parquet Deformations and both clockwise and anti-clockwise in increments of
Islamic Geometric Patterns 6°. Throughout this gradual process, several distinct
As a visual artist and specialist in Islamic geometric designs emerge that are noteworthy. The pattern at (D)
design, I enjoy exploring new methodological ap- is characterized by two sizes of regular decagon; the
proaches to creating geometric patterns. When I first pattern at (G) includes regular pentagons and is, in fact,
saw Craig Kaplan’s parquet deformations of Islamic a well-known Islamic geometric pattern; the example
geometric patterns,5 I was both excited by their aes- at (J) shares characteristics with a variety of fivefold
thetic impact and beauty, but also disappointed for historical patterns; the pattern at (M) has parallel lines
not having thought of the idea myself! He and I have leading to the 10-pointed stars, and lines that converge
both worked extensively with the use of polygonal tes- upon a single point at the center of the initial format
sellations as the substructure for generating Islamic ive pentagons (see A); design (P) has parallel lines that
geometric patterns, and when I saw his parquet de- connect the 5- and 10-pointed stars; the pattern in
formations of Islamic geometric design, I immediately (S) is characterized by regular pentagons and is also
comprehended his gradual opening and closing of the a historical Islamic pattern; and example (V) includes
angular conditions of the crossing pattern lines that regular decagons. The use of two points of contact be
are located at the midpoints of the generative, but tween the edges of the formative tessellation and the
not shown, polygonal tessellation. Along with a more applied pattern lines in this sequence is in keeping with
general interest in parquet deformations, this work the two-point historical pattern family associated with
sparked my own interest in working with geometric the tradition of Islamic geometric design. The method
designs that are characterized by gradually changing for placing and rotating these pattern lines is demon-
rotational conditions. strated in Figure 3. The horizontal line in (A) represents
It is worth mentioning that the examples of such any one of the polygonal edges that make up the initial
gradually changing Islamic geometric designs do not tessellation. This edge line shows the two points in
strictly adhere to the two criteria for parquet defor- contention, and the diagram above this line shows how
mations as set by Huff.6 In particular, the patterns that the division of these lines is based upon the golden
are being modified are comprised of many distinctly section. This is a purely arbitrary decision on my part. In
different polygonal shapes that lock together to create the subsequent illustrated edge conditions, the dashed
a larger repeat with translation symmetry, rather than lines represent the original polygonal edges, while the
each modified cell of a parquet deformation having the two sets of crossing pattern lines can be seen at the
ability to fill the two-dimensional plane on its own. And two division points. These sequentially increase with
yet the aesthetic character of these gradually changing 6° rotations until they become perpendicular with the
Islamic geometric designs, as well as the modifying formative polygonal edge (P).
process itself, is close enough to be regarded, at least in As these illustrations demonstrate, parquet
my opinion, as a loose interpretation of Huff’s criteria. deformations of Islamic geometric designs involve
In order to better differentiate parquet deforma- the rotation of the crossing pattern lines where they
tions of Islamic geometric designs with patterns that intersect with the midpoints of the polygonal edges of
are modified via the process of hinged tessellations, the formative tessellation. This rotational process is
I have prepared a couple of examples of the former. I similar to, but nonetheless distinct from, the rotational
have also included additional visual information that process used in modifying designs with hinged tessel-
illustrates some of the principles used in their con- lations. They both have their own aesthetic merits and
struction. are both worthy of further exploration.
The gradually changing geometric design in Figure
1 is derived from the 3.4.6.4 – 4.6.12 two-uniform, or Hinged Tessellations
demi-regular grid of regular triangles, squares, hexa- Hinged tessellations are essentially the two-dimen
gons, and dodecagons. The angular conditions change sional corollaries of three-dimensional jitterbug trans
by 10° increments from 12-pointed stellar center to formations. By way of example, the octahedron is com-
the adjacent 12-pointed stellar center, yet the control prised of eight triangular faces with a 34 configuration
144 Pattern Manipulation through Hinged Tessellations Jay Bonner

Fig. 1: Progressive 10° angular transfor-
mations of an Islamic geometric design
constructed from the 3.4.6.4. – 4.6.12
2-uniform tessellation of triangles,
squares, hexagons and dodecagons.
Fig. 2: Progressive angular

transformation of a two-point Islamic
geometric pattern.
Fig. 3: The sequential modification of

the initial polygonal edges (A) uses two
points of rotation, thereby creating the
progressive angular transformations of
the two-point geometric patterns.

at each of the six vertices. By rotating each of the faces to illustrate its properties for the printed page, I have
by 60° at the vertices, the angular openings conform to arbitrarily selected specific rotational sequences.
the geometry of the icosahedron with a 35 configura- However, the reader is encouraged to keep in mind that
tion at each vertex. With a 90° rotation at the vertices, the sequences of rotating tessellations, along with their
the angular opening creates the cuboctahedron with associated geometric patterns, that are represented
the 3.4.3.4 vertex configuration. Similarly, the 120° ver- in this chapter are merely snapshots of a continuously
tex rotation of the equilateral triangles of the 36 regular changing geometric structure. As demonstrated herein,
two-dimensional grid transforms this tessellation into this rotational process has a dramatic effect on the
the 3.6.3.6 semi-regular grid of triangles and hexagons overall visual quality of the geometric patterns.
(Figure 8).
The two images in Figure 4 were both published in Pattern Application
my 2018 paper7 concerning jitterbug transformations The application of the pattern lines onto the hinged po-
as a vehicle for spherical Islamic geometric patterns. lygonal modules that make up the tessellation involves
These demonstrate how a specific rotation of the octa- the placing into each polygonal module that makes up
hedron’s triangular faces by an amount associated with the initial non-rotated tessellation a repetitive design
the heptagon (25.7142…°) allows for the construction of with lines that either flow seamlessly and identically
a geometric pattern that is characterized by a regular into the lines of the adjacent module or reflect into the
distribution of 7-pointed stars. The yarn temari in Figure adjacent module along their shared edge. Theoretically,
4b is a spherical projection of the jitterbug polyhedron these pattern lines can have any degree of geometric
in 4a. This lovely object was made by my friend Maude complexity, but for the purposes of this demonstration,
Rayburn in Santa Fe. I have kept the applied pattern lines very basic. Even
Since their discovery by Buckminster Fuller in 1948, with these simple repetitive motifs, and the resulting
jitterbug transformations have received considerable simplicity of the basic overall pattern with 0° of rota-
scholarly study.8 Similarly, the science of two-dimen- tion, once the process of hinged rotation takes effect,
sional hinged tessellations has been explored exten- the resulting changes to the basic pattern can become
sively and is well understood as an accordion-like quite interesting.
expansion and contraction mechanism applicable to The manner I have chosen to populate the inter
a diverse variety of polygonal tessellations, including stice regions that open up as the hinged tiles progres-
regular grids, semi-regular grids, and grids made up of sively rotate employs one of two stratagems: Either
non-regular polygons.9 the lines along the edges of the rotating modules
Vertex rotations of polyhedral tessellations invari- extend into the interstice regions until they meet with
ably increase in area as the angle of opening increases. another extended line, or the lines along the edge of
This is a distinctive feature that differs from parquet the rotating modules are mirrored into the interstice
deformation structures. The shape and geometric regions. In either case, the lines are cleaned up by
character of the interstice regions are a product of the either extending them until they meet with other lines
rotating polygonal modules and their degree of rotation. or are trimmed at a place that is aesthetically pleas-
As with jitterbug transformations, after the rotation has ing. Invariably, this process of introducing pattern lines
reached its zenith, continued rotation draws the polyg- into the interstice regions requires a degree of aes-
onal modules back toward themselves. This expansion thetically driven decision-making, and the examples
and contraction takes place as a temporal continuum; illustrated in this chapter are a combination of the
Fig. 4: A jitterbug transformation of the

octahedron with angular openings set
to 25.7142…° rotation, thereby allowing
for the equal distribution of 7-pointed
A B stars. The spherical projection.

Fig. 5: A two-dimensional pattern with
7-pointed stars that also employs
a rotation of the repetitive double-
triangles by 25.7142…° (1/14 of 360°),
thereby allowing the placement of
regular 7-pointed stars.
Fig. 6: Two historical patterns that

employ the principle of modular
22.5° rotation 25.7142° rotation (1/14 of 360°) rotation.
obvious, the more interesting, and the aesthetically “polygons in contact”). Unlike other proposed methods,
pleasing. In other words, the rotated designs depicted this highly refined design methodology is capable of
herein are, to a lesser or greater extent, arbitrary, even producing all levels of complexity found throughout this
if constructed using established formulae. discipline, and is ideally suited to producing patterns
Figure 5 shows a two-dimensional design, originally in each of the recognized pattern families associated
from my 2018 paper,10 that rotates the congruent edge- with Islamic geometric art. Moreover, there is an abun
to-edge double-triangle modules by an amount that dance of historical evidence for the use of the polygonal
also corresponds to the heptagon, thereby creating a technique, something that other proposed methods of
rhombic interstice region that allows for the construc- construction lack. However, one thing that most spe-
tion of a geometric design with regular 7-pointed stars. cialists agree on is the fact that an individual design can
There are a number of historical Islamic geometric pat- frequently be created identically with multiple meth-
terns that exhibit this form of rotation symmetry, and I odological techniques—whether historical or not. In the
have examined multiple examples in my book on Islamic case of the two historical examples in Figure 2, these
design methodology.11 Two particularly interesting histori- can easily be produced without using an approach that
cal examples from this book are shown in Figure 6. involves the willful rotation of a primary repetitive mod-
Despite such historical examples of patterns with ule to an arranged angle as per hinged tessellations.
rotated repetitive modules, it is important to emphasize These two examples are included because they closely
that the use of hinged tessellations to create geomet- resemble designs created with hinged tessellations, and
ric designs is not being proposed as a historic design their underlying geometry is essentially identical.
methodology. The methods used to construct geomet-
ric patterns by the countless Muslim geometric artists A Selection of Geometric Patterns
of the past is debated by specialists in this field. My own Created with Hinged Tessellations
belief is that the predominant method used traditionally Figure 7 illustrates the progressive hinged rotation of
was the polygonal technique (sometime referred to as the 44 regular tessellation made up of squares. The

squares are rotated in 5° increments—from 0° to 90°. the incremental increases to every 5° is an arbitrary
After 90°, further 5° increases in the rotation angle determination that was chosen as it is a divisor of
(not shown) cause the formerly obtuse angles of the 45°, 60°, and 90°, and therefore works nicely with the
interstice rhombuses to sequentially decrease, and the symmetry of the square. It is worth noting that the 60°
resulting patterns and tessellations are mirror images rotation creates the angular conditions of the 3.3.4.3.4
of the 5° increasing progression. In this way, the rota- semi-regular tessellation of triangles and squares. Once
tion to 95° is the mirror image of the rotation to 85°. the rotation reaches 90°, the square modules are once
For this reason, I have only demonstrated the rotation again in an orthogonal arrangement, and the geomet-
of the hinged tessellations in this chapter up to their ric design is identical in every respect save its surface
rotational zenith. coverage to that of the initial unrotated tessellation.
The applied pattern lines on each of the square Figure 8 shows a progressive rotation of the 36
modules are an 8-pointed star, with every other point tessellation of regular triangles. This also employs an
intersecting the midpoints of each edge of the square. angular increase of 5°, thereby allowing for the demon-
This repetitive module was well known to the historical stration of designs with 30°, 60°, 90°, and 120° rotation-
record, producing the classic “star and cross” deal conditions.
sign. The applied pattern lines extend into the inter- The applied pattern lines are a simple hexagon
stice regions to create the modified designs. Setting placed so that three of the vertices of the hexagon are
0° rotation 5° rotation 10° rotation 15° rotation 20° rotation
75° rotation 80° rotation 85° rotation 90° rotation

Fig. 7: A sequence of 5° rotations of the
44 tessellations of squares. Each square
is decorated with an 8-pointed star.

Fig. 8: A sequence of 5° rotations of the

36 tessellation of triangles. Each trian-
gle is decorated with a simple hexagon.
located at the midpoints of the triangle’s edges. On triangular modules with ditrigons that have three 90°
the 36 tessellation without rotation this produces the internal angles, and the tessellation with 120° angles is
classic threefold Islamic geometric pattern comprised the 3.6.3.6 semi-regular grid of triangles and hexagons.
of 6-pointed stars and hexagons. It is interesting to Figure 9 demonstrates a series of patterns creat-
note that the 60° rotation produces a pleasing tessel- ed by the sequential rotation of the 3.6.3.6 semi-reg-
lation of small equilateral triangles separated by larger ular tessellation. In a sense, this can be regarded as
equilateral triangles with edges that are double in a continuation of the previous example in its 120°
length. The tessellation with 90° rotation separates the rotation. However, in this hinged tessellation the hinged


Fig. 9: A sequence of 5° rotations of
the 3.6.3.6 tessellation of triangle and
hexagons. The applied pattern lines are
perpendicular to the edges of the grid.
vertices are not the same as those of the previous semi-regular tessellation of triangles, squares, and
120° example. These patterns are also derived from a hexagons.
5° sequence of rotations, from 0° to 90°, after which Figure 10 is a variation of the examples in Figure 9,
the triangular and hexagonal modules close back into with different applied pattern lines to the same 3.6.3.6
themselves, with resulting mirror images as discussed rotating tessellations. This demonstrates the range
previously. of diversity that can be achieved through varying the
The applied pattern lines are perpendicular to the applied pattern lines, and anyone interested in work-
edges of both the hexagonal and triangular modules ing with hinged tessellations as a means of producing
and are located at 1/4 divisions of the polygonal edges. geometric designs should play with all manner of
This produces a 6-pointed star within the hexagonal pattern line applications to a single rotating tessella-
module and a hexagon with extended edges within tion. As this illustrates, changing the applied pattern to
the triangular modules. The patterns in this set are the repetitive modules will result in very different, and
particularly pleasing. sometimes very satisfying, designs. Both the triangu-
The 60° tessellation produces the conditions of lar and hexagonal modules in this set of designs have
the 34.6 semi-regular tessellation of triangles and hexagons as their applied pattern lines. These applied
hexagons, and the 90° rotation produces the 3.4.6.4 hexagons connect the midpoints of each edge of the


the 3.6.3.6 tessellation of triangles and
hexagons. Each triangle and hexagon is
decorated with a simple hexagon.
modules. The patterns produced with this alternative The rotational tessellations in Figure 12 stem from
pattern line application are likewise very successful. the semi-regular 3.4.6.4 grid comprised of triangles,
The designs in Figure 11 are also produced from squares, and hexagons. In these hinged tessellations
the same 3.6.3.6 semi-regular tessellation in 5° rota only the square and hexagonal modules are hinged,
tional increments. The applied pattern lines of the with the triangles of the initial tessellation (0° rota-
triangular module simply connect adjacent midpoints tion) gradually morphing into new trifold shapes with
of each triangle’s edges, producing a 60° triangle within each sequential 6° rotation. These centers of threefold
each triangle. These 60° angles within the hexagonal rotation produce distinctive rotational design features
modules produce a 6-pointed star. Once again, the un- within the applied pattern lines.
rotated 3.6.3.6 tessellation with this specific application Figure 13 applies a different set of pattern lines to
of pattern lines produces a design that is known to the the same sequential 6° rotations of the 3.4.6.4 semi-
historical record. regular tessellation that was used in Figure 8. In this
The design produced from the 90° rotated tessella- series of designs only the square modules have applied
tion, with its combination of 6-pointed stars, 4-pointed patterns lines. These consist of a simple cross of two
stars, and triangles is, once again, known to the histori- perpendicular lines that connect opposite corners of
cal record (Figure 21). each square. The pattern that these applied pattern


the 3.6.3.6 tessellation of triangles and
hexagons. The triangles are decorated
with a simple triangle and the hexagons
with a 6-pointed star.
lines produce on the initial tessellation with 0° rota- octagons into each square module. Four of the ver-
tion is a well-known historical design comprised of tices of these octagons are located at the midpoints
6-pointed stars that have 90° angles at their points, of each edge of the squares. The 135° interior angles
surrounded by ditrigons with three 90° internal angles. of the octagons are mirrored into the hexagons and,
As with the previous example, the regions of threefold in the initial tessellation with 0° rotation, into the
rotation symmetry in the rotated tessellations pro- triangles. Within the hexagons, this mirroring is further
duce very obvious and interesting design features with elaborated with the introduction of a 6-pointed star.
this simple application of crossed pattern lines. The Within the triangle, the mirroring produces a ditrigon.
design created by the 90° rotation is a superimposi- This particular design set within the unrotated tessel-
tion of the regular hexagonal grid with its dual trian lation was popular among Mamluk artists in Egypt.12
gular grid. Figure 15 is a departure from the previous eight
Figure 14 is a third design variation based upon sets of rotational designs in that the rotating module
the same 3.4.6.4 initial tessellation. The pattern lines is not a regular polygon. Rather, it is the well-known
in this set of rotational tessellations place regular rhombus associated with fivefold symmetry that has


the 3.4.6.4 tessellation of triangles,
squares and hexagons. The decoration
places 90° pattern line upon the mid-
point of each polygonal edge.
72° and 108° interior angles. The translational unit of ple, the design produced from a rotation of 108° (not
many fivefold Islamic geometric patterns employ this shown) will be a mirror image of the design produced
rhombus, and indeed, the depicted initial design with from 72° rotation. The 36°, 72°, and 90° tessellations
0° rotation is very well known to the historical record. are particularly interesting. The interstice rhombuses of
I have chosen to illustrate the increased growth of the 36° rotated tessellation have 36° and 144° inte
rotation at 6° intervals. This allows for 18°, 36°, 54°, rior angles. This rhombus is likewise associated with
72°, and 90° increments, each of which is associated fivefold symmetry and was also used historically as a
with fivefold symmetry. Once again, continued rotation translational unit in Islamic geometric design. It is also
beyond 90° (not shown) causes the hinged tessellation worth noting that both of these two rhombuses pro-
to close back into itself, with further 6° increments vide the basis for Sir Roger Penrose’s aperiodic rhombic
being mirror images of the depicted designs. For exam- tiling with matching rules. It therefore stands to reason


places corner-to-corner pattern lines
within each square of the tessellation.
that the extended pattern lines that populate the inter- associated with fivefold symmetry (such as the regular
stice regions of this 36° hinged tessellation produce a pentagons) within an orthogonal repetitive structure.
particularly satisfactory geometric pattern (see Figure The formative initial tessellation with 0° rotation
25). Somewhat surprisingly, this does not appear to in Figure 16 is comprised of pairs of edge-to-edge
have been used historically. The tessellation of the 72° pentagons that connect with other pairs at their ver-
rotation of the rhombic module creates an interstice tices. This network of pentagons has two varieties of
rhombus that is identical to the rotating module itself, rhombic interstices. It is worth noting that these two
with 72° and 108° interior angles. In this example the rhombuses are the same fivefold rhombuses dis-
geometric design results in regular 10-pointed stars cussed previously (with 72°/108° and 36°/144° interior
located at each vertex of the tessellation. Despite the angles). The applied pattern lines for these pentagonal
non-regular stars at each vertex, the example with modules are a simple pentagon that connects each
90° rotation is interesting in that it combines features adjacent midpoint of the pentagonal module. The

unrotated initial tessellation with its applied pattern midpoints produces patterns in the obtuse family. (For
lines provides a good, if rather simple, example of the those interested in historical Islamic geometric design
polygonal technique whereby key points of an under- methodology, please refer to my book on this sub-
lying polygonal grid (most frequently the midpoints of ject.13) The absence of 10-pointed stars in the fivefold
the polygonal edges) are used to apply pattern lines. non-rotational initial design that is produced from this
This discipline is made more elaborate by the use of underlying tessellation identifies this as a field pattern,
four historical conventions for applying such pattern and although it does not appear to have been used
lines—each varying the angular conditions of the historically, is shares multiple aesthetic characteristics
pattern lines, thereby creating distinct pattern families. with the many fivefold field patterns used by Muslim
Within fivefold Islamic geometric patterns, the appli- geometric artists during the Seljuk Sultanate of Rum
cation of 108° crossing pattern lines at the polygonal in Turkey.

places 135° crossing pattern lines at
each polygonal midpoint of the tessel-
lation.


a grid of rhombuses comprised of 72°
and 108° angles. The decoration places
108° crossing pattern lines at the mid-
point of each rhombic edge.
This rotational sequence is also set at 6° intervals
and the brevity of incremental examples is due to an
interesting and unusual phenomenon: at 36° rotation
the tessellation becomes identical to its 0° start- Particularly Successful Designs
ing point except that the overall tessellation is now Created from These Hinged Tessellations
rotated 90°. Concomitantly, the geometric pattern that The many geometric patterns created by the hinged tes-
results from the 36° rotation of the pentagonal mod- sellations shown in the previous figures are rather small
ules is identical to that of the 0° rotation, except that it and, in some cases, difficult to appreciate. In order to
is now rotated 90°. A further 36° of rotation of the pen- emphasize how successful this method of pattern mani
tagons would bring the tessellation and pattern back pulation can be, I have selected examples from each
to its original starting point. It is also worth noting that of the previous sets that I find particularly interesting
the conditions of the 18° rotation create an orthogonal and appealing for further aesthetic elaboration. Spe-
tessellation and geometric pattern that repeats upon cifically, I have increased the scale, dispensed with the
the square grid (Figure 26). underlying formative tessellations, widened the pattern

24° rotation 30° rotation 36° rotation

a grid made up of pentagons and two
varieties of rhombuses. The decoration
places a simple pentagon within each
pentagon of the grid.
lines, provided the widened lines of most examples with approaching my explorations into hinged tessellations as
an interweave, and introduced color. These examples a means of developing new geometric designs.
are intended to help demonstrate the design potential It is, therefore, my intention that the material in this
of hinged tessellations for generating new and original chapter should demonstrate how the process of apply-
geometric patterns with interesting symmetrical charac- ing pattern lines into the interstice regions that result
teristics. from hinged rotations of polygonal modules is a highly
flexible method of creating original geometric patterns.
Conclusion Such designs will frequently bear the hallmark of indi-
A fascinating feature of the contents of this book is the vidual creativity, and potentially idiosyncratic aesthetic
expansion of two-dimensional parquet deformations into sensibilities. When the as-yet-unrotated tessellations
the realm of their three-dimensional corollaries. Such have applied pattern lines that are typical of historical
gradually changing three-dimensional space tessellations Islamic geometric patterns, as per the examples in this
have very real potential for application to architectural chapter, the patterns created through the hinged rotation
expression, including the architectonics of spatial layout, process will exhibit varying degrees of traditional Islamic
vaulting designs, and especially spaceframe structures. aesthetic standards. While I do not suggest that the use
Perhaps ironically, my contribution to this discussion of hinged tessellations was a historical methodology, or
moves from the three-dimensionality of jitterbug transfor that matter, a particularly effective means of produc-
formations to their two-dimensional corollary: hinged ing “traditional” Islamic geometric designs, it is through
tessellations. Despite my work with kinetic architectural innovative experimentation that the exceptional range of
features such as open-and-closing shade structures and stylistic and geometric diversity found within the Islamic
domes that slide into an open or closed position, it is geometric arts flourished. In short, experimenting with
difficult to imagine a practical architectural application of new approaches to design methodology such as parquet
such hinged tessellations. Could their gradually increas deformations and hinged tessellations is both enjoyable
ing and decreasing interstice regions serve as a means and frequently highly worthwhile. And just as innovative
to increase or decrease light penetration into a build- flexibility was undoubtedly an aspect of the historical
ing—either as window screen or shade structures? And discipline of Islamic geometric design, so also can con-
if so, what would one do with the ever-changing pattern temporary artists and designers find great satisfaction in
lines within the interstice regions? Such questions of working with these somewhat unusual methodological
real-world relevance should not hinder one’s inspira- variants. Parquet deformations and hinged tessellations
tion while working with new ideas. It is often a fact that are certainly avenues that can lead to worlds of further
inspiration precedes application, and this is how I am design exploration.

Fig. 17: Three patterns from
Figure 07 with widened
30° rotation 45° rotation 70° rotation interweaving lines.


Fig. 20: Four patterns from Figure 10 with widened interweaving lines.

30° rotation 75° rotation 90° rotation interweaving lines
Fig. 22: Two patterns from Figure 12

48° rotation 72° rotation with widened interweaving lines.

48° rotation 72° rotation with widened lines.
48° rotation 90° rotation

with widened interweaving lines.

Figure 15 with widened inter-
36° rotation 72° rotation 90° rotation weaving lines.
Fig. 26: Two patterns from

Figure 16 with widened inter-
0° rotation 18° rotation weaving lines.
References
1
Joachim Krausse and Claude Lichtenstein, Your Private Sky: Discourse R. Buckminster Fuller,
Lars Muller, 2001.
2
Jay Bonner, “Doing the Jitterbug with Islamic Geometric Patterns”, Journal of Mathematics
and the Arts 12, 2–3, pp. 128–143.
3
William S. Huff, Parquet Deformations “Best Problems” from Basic Design, State University
of New York, 1979; Douglas Hofstadter, “Parquet Deformations: A Subtle, Intricate Art
Form”, Metamagical Themes, New York, 1985; Doris Schattschneider, Visions of Symmetry:
Notebooks, Periodic Drawings, and Related Works of M. C. Escher, New York, 1990.

4
Rinus Roelofs: http://www.rinusroelofs.nl/animation/avi-dyn-tilings/avi-dyn-tilings.html.
5
Craig Kaplan, Islamic Patterns, ACM SIGGRAPH Art Exhibition, 2008; Craig Kaplan, “Curve
Evolution Schemes for Parquet Deformations”, Bridges Proceedings 2010, Mathematics, Music,
Art and Culture, 2010.
6
Huff, Parquet Deformations.
7
Bonner, “Doing the Jitterbug”, pp. 128–143, Figures 14 and 15.
8
Hugo F. Verheyen, “The Complete Set of Jitterbug Transformation and the Analysis of their
Motion”, Computers & Mathematics with Applications, 17, 1989, pp. 203–250; Duncan Stuart,
“Polyhedral and Mosaic Transformations”, Student Publications of the School of Design,
University of North Carolina, Raleigh, NC, 1963.
9
David Wells, The Penguin Dictionary of Curious and Interesting Geometry, Penguin: p. 199;
Robert Williams, The Geometrical Foundation of Natural Structure: A Sourcebook of Design,
Dover (reprint, 1979), Joseph Clinton, “Let’s Make a (36)D (36)L Chiral Tessellation Dance”,
Bridges Proceedings 2012, Mathematics, Music, Art and Culture, 2012.
10
Bonner, “Doing the Jitterbug”, pp. 128–143, Figure 9.
11
Jay Bonner, Islamic Geometric Patterns: Their Historical Development and Traditional
Methods of Construction, Springer, New York, 2017.
12
Bonner, Islamic Geometric Patterns.
13
Ibid.
Jay Bonner is a specialist in multiple Islamic design disciplines, including geometric patterns,
muqarnas, rasmi star-vaulting, as well as the floral idiom. He has an international reputation
for his work with Islamic geometric patterns, including particularly complex designs that
meet the modern mathematical criteria for self-similarity and quasi-periodicity. As an
independent scholar of Islamic geometric design, Jay Bonner has published multiple peer-
reviewed papers. He is the author of Islamic Geometric Patterns: Their Historical Development
and Traditional Methods of Construction, with a contributing chapter from Craig Kaplan and a
foreword by Sir Roger Penrose, Springer, 2017. At 595 pages, with over 100 photographs and
over 500 illustrations, this book is a significant contribution to this field of study. Jay Bonner
has taught design workshops and given lectures on the topic of Islamic geometric patterns
in North America, Europe, North Africa, and Asia. He was the opening keynote speaker at the
2003 Bridges Conference in Granada, Spain (Three Traditions of Self-Similarity in Fourteenth
and Fifteenth Century Islamic Geometric Ornament). Jay Bonner is currently working on a
series of online Islamic geometric design courses that will be available very soon. He is also
a professional design consultant specializing in Islamic architectural ornament, with some
38 years of experience working on projects in the Middle East, Asia, Europe, and the United
States. The many projects he has participated on include: the expansion of the Masjid an-
Nawabi (Prophet’s Mosque) in Medina; the expansion of the Masjid al-Haram (Grand Mosque)
in Mecca, including the minbar for the Kabba courtyard; the Abraj Al-Bait Clock Tower in
Mecca; the International Medical Center in Jeddah; the Tomb of Sheikh Hujwiri in Lahore; the
New Senate House in Rawalpindi; and the Ismaili Centre in London.

Parakeet3D: Algorithmic
Re-Envisioning of Geometrical
Pattern Morphogenesis
Esmaeil Mottaghi and
Arman Khalil Beigi Khameneh
1. Introduction of the generation process. The data we have today on

Computational tools have created new linguistics in the progression of these patterns is minimal. Barring
different disciplines related to design and fabrication. a few exceptions, what has been gathered to date are
The algorithmic nature of these tools offers a critical mere morphs and shapes without associated data or
possibility to encode the form-generation process algorithms. The missing link is the logic behind these
and embed various constraints, which results in a rich morphological processes, forming the idea behind
and dynamic platform for early exploration of design Parakeet3D. This research aims to decode or approx-
possibilities. Furthermore, the innate procedural notion imate the generation process behind some of the old
in these tools coupled with tracking methods and and authentic patterns in an algorithmic syntax. This
maintenance of the associated data of each alternative approach enables the designer to generate, apply, and
facilitates the process of evaluation and optimiza- analyze patterns not only for ornamental designs, but
tion using real-time feedback. Algorithms enable the also for modern architectural purposes, such as thin
designer to explore the design space interactively and shells, free-form surfaces, and performative envelopes.
efficiently. Parakeet3D is a cross-platform design tool. Parts of
The computational approach is a relevant discourse it have been released for Grasshopper3D™, an algorith-
under the subject of morphological methods for geo- mic modelling tool associated with McNeel Rhinocer
metrical pattern generation. There is a notable gap in os™. In addition, Parakeet3D offers methods of creat
the research on [traditional] geometrical patterns. What ing and modifying patterns and biomimetic patterns
is being accumulated and studied for most geomet- (Figure 1), and some geometrical processes inspired by
rical patterns is simply the outcome or final product computer science are also included.
Fig. 1: Examples of bio-inspired

patterns generated using
Parakeet3D:
a b c d
a) Venation algorithm I
b) Venation algorithm II
c) Differential growth algorithm
d) Floral (arabesque) patterns
e) Fracture (crack) pattern
f) Fractals
g) Diffusion-limited aggregation
e f g h h) Flow-path patterns

a b c
Fig. 2: Basic concepts and methods

for geometric pattern generation using
Parakeet3D:
a) Tilings (base grids)
b) Modification methods
c) Pattern generation methods
(Genotypes)
2. Heterogeneous Pattern Generation interventions by advanced designers. The intention

via Parakeet3D was to design the pipeline not in a black box, but in an
To create parquet deformations or heterogeneous adaptable manner. Aligned with a generic interpretation
patterns with Parakeet3D, a number of basic con- of these patterns, the methods for generating patterns
cepts must be considered, namely tilings, modification are called “genotypes”. This approach is highly valuable
methods, and pattern genotypes (Figure 2). To obtain for multiple reasons, including the ability to keep track
a simplified design method, the generation proce- of each outcome via “identifiers” or “genes”, to substi-
dure of networks is divided into four discrete steps: tute some of the time-consuming geometrical pro-
first, selecting the base grid or tiling; second, selecting cesses with simpler textual operations, and to achieve
modification methods which can optionally be applied an enhanced optimization process.
on the base grid; third, deciding the pattern-generation To design patterns or, in particular, parquet defor-
method applied on each cell; and fourth, applying op- mations, the base grid or tiling must first be selected.
tional modification or post-processing methods upon The tiling category of Parakeet3D currently consists of
the resulting network. a variety of tilings, such as uniform and semi-uniform
This approach simplifies the process for intermedi- tilings, (some) k-uniform and irregular filings, and pen-
ate users and offers numerous possibilities for custom tagonal and non-edge-to-edge space-filling networks.
b c
Fig. 3: A Parakeet3D 2-Uniform Tiling;
a) Basic tiling (with vertex configuration
of [3.4.6.4 and 3^2.4.3.4]
b) Tiling after mirroring quad subdivision
c) Dual graph of the tiling
a d) Tiling after truncation
d e e) Tiling after complex transformation A.
164 Parakeet3D: Algorithmic Re-Envisioning of Geometrical Pattern Morphogenesis E. Mottaghi, A. K. Beigi Khameneh
a b
Fig. 4: Application of different

Parakeet3D genotypes and a tiling:
a) Genotype K
b) Genotype C
c) Genotype A
b) Genotype B
c d
a b
Fig. 5: Application of a Parakeet3D
genotype on modified tilings:
a) Genotype B on a tiling with mirroring
quad subdivision
b) Genotype B on a tiling’s dual graph
c) Genotype B on a tiling with
truncation
d) Genotype B on a tiling with complex
c d transformation
At this point, an opportunity for post-processing is ⟩ Dual graph: Based on a notion in graph theory, a
offered, or for modifications that can be applied to the dual graph is a graph (network) in which the nodes
grid. These modifications increase the complexity and are located inside the faces (closed polylines) of the
diversity of the results (Figure 3). initial network (in this case, at the centroid of the
The modification methods—including truncation, existing polylines). The connectivity is derived from
mirroring quad subdivision, dual graph, and complex the topological relation of the initial cells. As the
(non-Cartesian) operations—can be applied on any type dual graph of any base grid results in different and
of network consisting of closed polylines. Each method is diverse cells, it expands the options for the base
detailed below. network (Figure 3c).
⟩ Mirroring quad subdivision: Derived from some ⟩ Truncation: Commonly perceived to be similar to
traditional geometric patterns, mirroring quad sub- the concept of Archimedean solids in 3D space,
division is a particular method for subdividing the truncation refers to essentially shrinking the end
base grids. Though the subdivision itself is based on points of each linear element towards its mid-
the common method of drawing the perpendicular points, which creates a new cell (closed polyline)
bisector on each edge, the key resides in the order of at each node with several segments equal to the
the points in resulting shapes (Figure 3b). Thus, points node’s valency. This process can be repeated recur-
on each sub-cell are ordered so that horizontal and sively, as needed, which also smoothens the angles
vertical mirroring axes are created. This particular among elements (Figure 3d).
order of points makes seamless patterns when used ⟩ Non-Cartesian operators: Several numerical
with the majority of Parakeet3D’s pattern genotypes. methods can be applied on geometries for complex

transformation, namely Möbius and hyperbolic majority of genotypes are designed to create seam
transformation. For instance, a simple method is less networks. The logic can be applied on any closed
converting the Cartesian coordinates to non-Carte- polygon and therefore applies to every type of tiling
sian coordinates, followed by processing the result (Figure 5). After this point, the designer can perform
with complex operators and returning the results to the modifications again, if needed.
Cartesian space again. This is not an affine trans- By selecting the base grid or tiling, the generation
formation, meaning that the parallelism, angles, method or genotype, and optional modifications on
and distances will not remain constant (Figure 3e). each step, a homogenous pattern can be created (Fig
After the optional application of the modifications, pat- ure 6). Though changing the input genotype parameters
tern genotypes are subsequently selected (Figure 4). The generates various patterns, all cells are associated with
a
Mirroring Quad Subdivision
Application of Genotype rules
c
Mirroring effect
d
Pattern propagation
Fig. 6: Homogenous pattern generation procedure

using Parakeet3D Genotype B and mirroring quad
subdivision on a hexagonal tiling: a) Subdivision of
a cell into quadrilateral sub-cells, b) Application
of genotype rules, c) Alteration of point orders in
adjacent cells to create mirroring effect,
d) Pattern propagation.
Mirroring Quad Subdivision Selecting a GenoType
Mirror Line
δ= 0 δ = 0.2 δ = 0.4 δ = 0.6 δ = 0.8 δ=1
Fig. 7: Variation of (homogenous) pat-

terns created using a sample genotype
(G) on subdivided base grids.
a constant state of genotype; therefore, the resulting simple techniques can be used. Widely used meth-
shapes are uniform and unvarying (Figure 7). ods include using X or Y coordinates for each cell’s
In order to create heterogeneous variations, the centroid (Figure 8) or the distance from each cell to
genotype input parameters have to vary. Therefore, in- certain distinct geometries in space. More advanced
stead of associating all of the cells with a single state users may create unique values for each cell based on
of a genotype, each cell is linked with a unique num- numerical equations, graphs, or values derived from
ber. To achieve a unique value for each cell, several the design environment.

δ0 δ1
δ
1
Fig. 8: Parquet deformation

generation by a continuous
change in genotype inputs.
Fig. 9: Parquet deformations

based on Parakeet3D
genotypes.
To summarize, having selected tilings, genotypes, mation parquets can be generated using Parakeet3D.
and modifications, homogenous geometrical patterns Moreover, algorithmic complexity augments the result
are generated. Then by associating different values ing patterns with tools embedded in Parakeet3D or
for each cell, heterogeneous networks, or parquet methods from the host environment.
deformations, emerge. At this point, as most of the Non-linear deformations are exemplary outcomes
Parakeet3D genotypes are standardized, numerous of this approach. Thus, utilizing integrated digital tools
variations of parquet deformations can be generated can push conventional limitations. Examples include
(Figure 9). transitions of higher degrees, polynomial or spline-
based (graph-based) transitions (Figure 10), or tran-
3. Beyond Linear Deformations sitions based on characteristics of initial surface or
The computational approach in Parakeet3D pattern mesh, such as Gaussian or mean curvature. Another
genotypes aligned with the generative context of method for creating patterns with non-linear propaga-
Grasshopper™ enables the designer to easily explore tion is the use of specific tilings, for example, hyperbol-
advanced types of interpolation. The linear transition ic tilings such as the Poincaré disk model or substitu-
between the cells to create a metamorphosis outcome tional tilings such as Penrose (Figure 11).
has been widely explored. Conventional linear defor-
1
Delta (δ)
0
x
1
Delta (δ)
0
x
1
Delta (δ)
0
x
1
Delta (δ)
Fig. 10: Effect of non-linear

interpolation of genotype
0 x
values.
Fig. 11: Parquet deformations

on non-linear tilings:
a) Parquet deformation on
hyperbolic disk,
b) Parquet deformation on
a transformed variant of
regular square tiling,
c) Parquet deformation on
Penrose tiling.

4. Towards New Dimensions area. Therefore, the corresponding three-dimensional
Several approaches can be used to investigate geo shape has sections with a constant area (Figure 12).
metries derived from patterns beyond the two-di- The transition speed (rate) between various sec-
mensional plane. One widely studied method is tions is a further parameter for examination. Non-
stacking shapes inside each cell upwards to create a linear transformation of input parameters forms
2.5-dimensional shape. A more profound approach, various three-dimensional geometries. The transfor-
though, comprises using a non-Cartesian coordinate mation speed is not constant, and the conversion
system—that is, mapping coordinates to cylindrical or is performed using non-linear interpolations. For
spherical systems. instance, interpolation may be based on arbitrary
graphs, trigonometric functions, or polynomial equat
4.1. [Pseudo] Three-Dimensional Geometries ions (Figure 13).
This method substitutes linear propagation in a two-
dimensional plane with a horizontally assembled 4.2. Non-Cartesian Geometries
order. In this widely investigated approach, the third Excluding the Cartesian methods and transformation
dimension is not actively generated and lacks the rates that can generate complex and novel patterns,
amount of data embedded in other dimensions. (It we conducted extensive research using various coor-
is a plain accumulation of two-dimensional shapes.) dinate systems with Parakeet3D. The logic behind this
Therefore, this category may more accurately be tool enables designers to explore geometric designs in
called 2.5D (two-and-a-half dimensional) or pseudo- different coordinate systems. Genotypes are designed
three-dimensional geometries. to work with simplified numerical inputs. There-
In this method, a number of issues are still fore, these parameters can easily be interpreted as
under-studied, including “transition rate” and “de- non-Cartesian coordinates. The generator parameters
sirable phenotypes”. Desirable phenotypes are a set can be translated into polar coordinates in two-dimen-
of characteristics resulting from the interaction of sional space or cylindrical and spherical coordinates in
their genotypes. These phenotypes (expression or three-dimensional space (Figure 14). This novel approach
emergent outcomes of genotypes) can offer benefi- generates greater complexity than the conventional
cial architectural/design features. For example, some method of stacking 2D layers of geometries upon each
genotypes have regions that maintain a fixed inner other to create 3D geometries (Figure 15).
=1
A
=1
A
A=1
=1 A=1
A
) A=1
(δ
ta =1 Height (H) A=1
Del A
A=1
=1 A=1
A
A=1
=1
A
Fig. 12: [Pseudo] three-dimensional

geometries created by the accumulation
of 2D patterns. Depicting an emergent
phenotype of maintaining a constant area
throughout the transformation.
δ
1
1 1 1 1 1
Delta (δ)
Delta (δ)
Delta (δ)
Delta (δ)
Delta (δ)
0 0 0 0 0
Height (H) Height (H) Height (H) Height (H) Height (H)
90 90 90 90 90
Angle (θ)
Angle (θ)
Angle (θ)
Angle (θ)
Angle (θ)
0 0 0 0 0
Height (H) Height (H) Height (H) Height (H) Height (H)
Fig. 13: [Pseudo] three-dimensional

geometries created by the accumula-
tion of 2D patterns. Depicting the effect
of non-linear transition rate.

P(1, 90°, 0°) P(1, 72°, 11°) P(1, 54°, 22°)
P(1, 36°, 31°) P(1, 18°, 39°) P(1, 0°, 45°)
Fig. 14: Interpretation of

input parameters of a geno-
type (I) to spherical coordi-
nates.
Fig. 15: Interpretation of

input parameters of different
genotypes to spherical
coordinates.
5. Discussion: Limitations and Opportunities One of the concerns regarding digital tools is at
Through the radical advancement in personal com- the level of user intervention. On one hand, tools like
puters and the emergence of advanced computational Parakeet3D, which are usually referred to as “high-lev-
tools at designers’ disposal in the last several decades, el” tools, need to limit the number of user inputs/
a conspicuous tendency towards building free-form modifications to keep the tool user-friendly and sim-
geometries or projects with augmented performative ple. On the other hand, “low-level” tools offer a higher
capabilities has developed. These phenomena reflect degree of freedom for more competent users. Thus,
the urge of designers to solve increasingly complex a limitation in digital tool development is managing
problems. The level of complexity and the integrat- a trade-off between being user-friendly and at the
ed data in modern projects require an inevitable shift same time providing intervention possibilities. Another
in design tools and problem-solving methods. The limitation of morphological tools is the coordinate sys-
algorithmic context offers possibilities to tweak and tem. Current studies on pattern generation methods
develop for advanced users. Digital tools have even primarily use a Cartesian coordinate system and then
greater potential in their integration with tools from morph, map, or project the outcome shapes onto arbi-
other disciplines. Digital tools can be adjusted with the trary geometries. These transformations are problem-
requirements and constraints of complementary fields. atic in many cases, for instance, in mapping a flat pat-
For instance, with a generative tool like Parakeet3D, data tern onto a manifold mesh. A possible solution is using
for digital fabrication can be generated directly from local coordinates systems, such as using parametrized
the generation platform or optimization criteria can be UVW coordinates of a surface or mesh instead of using
considered to optimize the fabrication process of waste Cartesian positions and calculations. This approach
or time management. requires a major revision of generation procedures, as
Geometry is a small portion of what a competent even the simplest methods of calculating intersections,
digital tool must offer a computational designer. At the angles, and distances are fundamentally different in
lowest level, generating morphs and shapes is expected local UVW coordinate systems.
from such tools. Integration and coherence is another Parakeet3D represents an effort to revise pattern
significant feature. Using standard protocols and meth- generation methods. Algorithmic thinking can be used
ods for storing or transferring data enables digital tools to re-code the generative process. The subject of
to create a powerful gestalt. For example, instead of geometrical patterns began long before computational
saving the geometric components in simple plain data, tools were in designers’ toolkits, yet the geometrical
Parakeet3D handles data in more advanced data struc- and mathematical concepts behind it make it high-
tures, such as half-edge data structures. This practice ly compatible with modern computational geometry
allows Parakeet3D to keep topological data associated syntax. Computational tools offer a vast opportunity
with morphological data and thus tightly integrated with for designers to effortlessly generate, represent, and
other major computer science libraries, which makes evaluate their designs.
future developments much more manageable.
Biographies of the Authors
Arman Khalil Beigi Khameneh is a digital architect. He holds a master’s degree in architec-
tural technologies. He is a design technician, and his teaching focuses on design computa-
tion and integration of cutting-edge or customized fabrication technologies into the design
process. He pushes the boundaries of his designs to the intersection of computational ge-
ometry, digital fabrication, and material technologies. He is a co-founder of Paragen creative
studio, where he provides algorithmic solutions for complex design and fabrication.
Esmaeil Mottaghi is a computational designer, architect, and computational geometry re-

searcher based in Tehran, Iran. He graduated with a master’s degree in computational design
from the University of Tehran and has experience as an expert in digital manufacturing and
as a computational design tutor. He has also been a director of multiple digital fabrication
and algorithmic design workshops organized by Tehran University and other architectural
centers. He is a co-founder of Paragen creative studio.

Teaching Perspectives
Presenting the Experiments’ Outcomes

Editor's note
Approximately 450 students completed the experiment in each of the two years—a selection
of their works form the core of this book and are displayed in the following chapters carry-
ing the names of the respective experiments: “3D Parquet Deformation” (winter semester
2017–2018) and “Cellular Space Sequences” (winter semester 2018–2019).
Both chapters progress chronologically so that readers can follow the logic of the assign-
ments given to students. In both cases, initial two-dimensional exercises are shown, leading
to a selection of three-dimensional models built at the end of the semester.
Following the notion of the artistic experiment, this book focuses on presenting inves-
tigations rather than judging the results obtained by students as right or wrong. The topics
investigated by the students’ final presentation model overlap in many instances, and they
have been grouped accordingly during the editing process of the book.
The “3D Parquet Deformation” chapter features works that explore “Composition”, “Dis-
solving”, “Gradual Changes”, “in Motion”, “Materiality Matters”, “Multiplication”, and “On Stage”.
The “Cellular Space Sequences” chapter collects works examining “Balance”, “Crystal-
line”, “Gradual Changes”, “Materiality Matters”, “Multiply”, and “Opening Up the Inside”.
Themes do not follow any (scientific) criteria chosen beforehand; instead, they have
arisen from carefully studying the students’ works and distilling ideas from them—ideas
that reoccur, that blur into one another, and that do not present a systematic evaluation
regarding a particular direction, be it materiality or geometric rule sets. Rather, the topics
extracted from the works do not follow any given order; they stand next to one another to
build relationships.
The models were realized with many different materials and processing types, the draw-
ings were created partly by hand, partly digitally. This and also the exact size is not always
documented and is therefore not indicated. Short explanations offer the reader some back-
ground information about the design process and essential features of its structure.
175
The Tiling and The Whole
Christian Kern
As the person responsible for the “three-dimensional Shapes and surfaces are necessarily abstracted, bodies
design and model making” research field at the Vienna initially represented in monolithic form.
University of Technology and as one of the editors of this In the 1:1 implementation, depending on the mate-
book, I would like to begin with a few cursory notes on rial and its processing, contiguous seamless surfaces
the task, the process, and the results presented here. as found in the model are not possible. A discretization
Our research center is affiliated with the Institute of the surfaces is necessary—that is, a division into
for Art and Design, part of the Faculty of Architecture elements of limited size that can be produced and
and Spatial Planning. We train architecture students, assembled. The geometry of these elements and their
among others, in fundamental questions of form. Due to interaction produce an aesthetic effect that either sup-
its size and durability, architecture is a very visible and ports the form of the architecture or space or disturbs
permanent symbol of culture; it is “building culture”. As and questions it. The division and the structure of these
a creative achievement of a community, it interacts with elements should therefore not be decided in a techni-
other artistic, intellectual, and creative disciplines. It may cal-pragmatic manner in implementation, but should be
hurry ahead or lag behind or even be self-referential or related to the aesthetics of the building.
consciously take itself out of the temporal context in Perhaps this discretization of surfaces is already
which it repeats. An essential difference from other cre- shaped by rhythms, by rules that live from many or a
ative cultural achievements lies in architecture’s sheer few complex variations of the repetition and which can
dimensionality. be easily described mathematically. In this case, it would
As a rule, it is not possible to conceptually grasp be obvious to apply these rules also in the sub-level
the future effect of architecture before it is physically of the form, in the design of elements. Yet perhaps the
represented in any way. In the development of a design, architecture is complexly curved or dissolved into ob-
visual media and other tools, are therefore necessary jects with different geometries and directions, as in the
to convey an impression and allow the assessment of work of Zaha Hadid, Frank Gehry, or Daniel Liebeskind.
design approaches. This process does not happen at In such cases, the development of the division becomes
the real scale of architecture; this would be too slow, an exciting design task in its own right, a task that we
too complex, and would allow too few variants. As such, approached with students over two semesters as part
the design is scaled down, and work is carried out at of the subject “three-dimensional design” in architecture
this smaller scale—in the sketch, the plan, or the model. training at the Vienna University of Technology.
Fig. 1: Cinthia Anton, 2017;

overlapping effects.
Space Tessellations Teaching Perspectives 177

Legato and crescendo Three-dimensional effects were observable even in
in the deformation of the parquet the first drawings. Overlaps, and thereby depth, could be
Let us consider the parquet in more detail. In mathe interpreted depending on the geometry of the elements,
matics, a parquet is defined as the gap-free, over- because human perception automatically simplifies
lap-free division of an area or plane. For functional rea- shapes for reasons of effectiveness. Shapes are com-
sons, this definition does not have to exclusively apply pleted even if the corresponding area is not visible or
in architecture. Gaps can be interesting as perfora is mentally overlaid by another element. If one was
tions (light, ventilation), and overlaps can be useful for aware of this principle, one could consciously increase
technical reasons (e.g., with shingles). Nevertheless, it the depth effect by, for example, using convex shapes
makes sense to first stick to the mathematical defi- together with an additional expressive shape, leading to
nition in order to be able to develop design rules and an incision in the adjacent field. The latter was ignored
to start experiments in which not only results are as- in the perception as an essential feature of the actual
sessed, but targeted changes are made with increasing shape and interpreted as an overlapping element of an
knowledge gain. adjacent element.
A wide variety of often surprising patterns can be A paradox emerges here. When the shape develops
accomplished within these rules. With small changes outwards, it turns inside out on the opposite side at the
in the geometry of the elements, there are sometimes same time, since all parts used are congruent in a tiling.
decisive changes in the appearance of the whole. In the The creative impact of these effects is difficult to pre-
spirit of experimentation, the following design questions dict, even with some experience. The design process in-
can be asked: What if convex shapes are chosen? What volves emergences, appearances, and effects that arise
if partial shapes become expressive? We explored these in the making and that were not considered in advance.
questions with a large number of students, and it was Good examples are directions, rotation effects, and
amazing how different and enriching the results were, depth effects, which become visible when a sufficient
even with simple specifications. On the basis of these number of elements work together.
investigations, the experiments, as described in more If there are more elements, the effect does not
detail elsewhere in this book, became more complex. fundamentally change, but may repeat or lengthen. This
We advanced from legato to crescendo, from a se outcome can be useful if a designer does not want to
quence to a deformation and thereby to an increase in distract from the geometry, the area being played on, or
the effect. The elements were continuously changed, or the architectural context. If parquet may be effective as
the grids in which the elements developed were warped. a design element, the deformation becomes interesting:
This was initially a two-dimensional development with a a continuous change in the basic structure or the pro-
three-dimensional effect. At the end of the discussion, tocell in which these interventions become more and
however, there was a need to think about and apply til- more expressive and complex. This process can occur
ing in three dimensions as an abstract principle of order radially or linearly, for example, and a curve or a centric
for sculpture and architecture. effect is created accordingly. The students’ examples
Fig. 2.a: Sarah Reithofer, 2017, depth Fig. 2.b: Daria Lanina, 2017, depth
effect through overlapping effects. effect through different densities.
178 The Tiling and the Whole Christian Kern

Fig. 3: Anja Bezjak, 2017,
wave motion due to grid
distortion.
ppear particularly dynamic when the changes take

a Yet, what if important rules like the absence of gaps
place in two directions at the same time. For various and the exit from a basic grid still apply? As is always
reasons, in these cases the three-dimensional effect the case with the transition from a two-dimensional
tends to be stronger than with simple tiling in only to a three-dimensional design investigation, the effort
one direction. Overlapping effects can occur again, involved in exploring this question is significantly greater
though. Another cause is the depth perception due to than in a pure drawing experiment. Form studies have
the so-called “aerial perspective”. Based on experience, to be built and material processed. Furthermore, gravity
the assumption is that objects further away are seen is of great relevance in the process, as joints must be
as brighter, as the contrasts decrease due to reflective form-fit or force-fit.
floating elements in the air (aerosols, dust, water vapor). A particular problem arises in the visual or haptic
Through a compression and thus a change in the gray detection of the configuration or the deformation. If
value, the corresponding areas are interpreted as closer. made from opaque surfaces, the solid packs are initially
With alternating gray values, three-dimensional wave invisible. As such, strategies had to be developed to
movements can be perceived. look inside or to present elements in such a way that an
The deformation of the basic grid had a dramatic interesting composition did not just appear as an un-
effect (Exercise 3). Due to these distortions, elements defined collection. One way was to break up the bodies
are interpreted as lying diagonally in space. A strong into edges. Another method was the definition of gaps,
depth effect can be expressed by means of a parquet which—together with the solids—were subject to a rule.
deformation in the plane. The French Op Art artist Victor In some cases it was sufficient to show the basic solids
Vasarely has often used this effect in his pictures and detached so that an observer constructed the spatial
justified the use of such optical effects with the state- representation of the geometry in his or her head, even
ment: “Art is artificial and by no means natural: creating though it was not directly visible.
does not mean imitating nature, but equaling it and The wish that individual elements and their config-
even her by means of an invention which, among all uration could be experienced tends to lead to additive
living things, only humans are capable of surpassing it”1. “element swarms”. These can have any effect or, as a
As previously noted, this examination of the rules whole, can acquire a meaning and become “gestalt”
and design possibilities of tiling formed the basis for a (form + meaning = gestalt). Students developed basic
three-dimensional, object-like, and spatial investigation. design ideas so that a sculpture could be created that
It was important to us that we not only use this princi- is more than the sum of its parts. This idea was often a
ple for the design of surfaces, but also for the shaping development or transformation of basic elements and
and configuration of solids. For functional and efficien- thus the overall form, partly linear, partly radial. There
cy reasons, architecture is often thought of as a “tight were variations of bodies that remained similar in size,
package”. Residential buildings, for example, primarily but there were also works with a dramatic crescendo.
consist of a three-dimensional arrangement of rooms. Others exhibited little or no variation and lived off a
These spaces, together with their envelopes, can also rhythm that is made interesting by light and shadow.
be understood as stacked solids, as in the case of North There were works that seemed self-contained and
American pueblo architecture. others that represented an excerpt from something
What if these spaces are not just orthogonal? What larger. Deliberately chosen locations, the decision for or
if they change in the sense of a “parquet deformation” against a base, and elevations or suspensions supported
and break away from the monotony of 3D parquet? or weakened the design intent. Color, gloss, and strong

Fig. 4: Victor Vasarely, Vega,
1956. © Bildrecht, Wien
2021.
material structure occurred, but could distract from the but interior and intermediate spaces were considered
shape. In their conception and implementation, the works from the start. Although these spaces were geometri-
were very enriching and impressive. What they have in cally related to the outer shape, they were not neces-
common in a positive sense is a mathematical, geomet- sarily identical in terms of their formal characteristics.
ric logic with individual characteristics and high design The topic was expanded by connecting the spaces, a
quality. sequence that can be walked through or flown through.
Emergences, phenomena that were not consid-
Staccato of the objects and ered in advance, again arose from the mathematical-
legato of the spaces and the space sequence geometric principles. A conscious and high-quality
For the initiators of the previous exercise, an import- choreography in the spatial sequence, the sequence,
ant question of perception remained open, which was and the transitions between the spaces was required—
reflected in the work described above. The objects are a conscious design within the rules and a commitment
generally viewed from the outside as sculptures, partic- to individual authorship. The shape should be edited
ularly because of their dimensions. There are sometimes
gaps, when the solids are broken up into bars or trans-
parent surfaces so that insights are possible. However,
the spatial quality of the interior is not easy to grasp and
was not particularly addressed. In architecture, it is not
only the formal quality of the object that is essential, but
also the space inside the solid or the spaces in between.
This aspect became formative for the program in
the following year for the “Cellular Space Sequences”
exercise. Here, too, the basic assumption was taken from
tiling and its geometrical-mathematical fundamentals,
Figure 5: Peter G. Auer, 2018;

Example of a solution.
180 The Tiling and the Whole Christian Kern

and defined on four levels: the shape of the cell, then are s ubject to different requirements. A basic design
the overall shape resulting from the combination of exercise in our faculty is also a basic exercise in archi-
cells, the space inside the cell, and the spatial contin- tectural thinking.
uum resulting from the connection. With some results, The three-dimensional development or investiga-
there is another form that requires attention, which is tion of the topic of cellular spatial sequences leads
the spaces between the cells (i.e., spaces in between). to different design priorities than with the previous
This makes judgements as well as predictions parquet deformation. The focus, as indicated in the title
within this design exercise a complex and demanding of this exercise, was on the sequence of spaces, the
task, especially because it is extremely difficult for choreographed transition from one space to the next.
human perception. Not least for evolutionary reasons, The solids themselves were mirrored or rotated, but
people usually concentrate on objects (or subjects). basically retained their geometric features. This continu-
Like interior spaces, these are delimited, while spaces ity resulted in compositions in which a staccato of the
in between often expand without limits in one or more objects was accompanied by a legato of the spaces.
directions. In the two-dimensional graphic represen- These “tilting solids”, in which the viewer’s percep-
tation, this perception problem can be avoided by re- tion constantly fluctuates between two basic charac-
versing it, keeping the cut surfaces of the bodies dark teristics, required a critical number of elements in order
(e.g., by hatching). This approach makes the spaces for the composition and sequence to be effective. It
more visible as a form. However, concentration is re- therefore took some effort to physically manufacture
quired to be able to perceive the shapes of the objects them. In addition, openings, resolutions, or distances
in this type of representation. had to be developed and implemented with reference
As in a tilt image by psychologist and phenom- to a higher-level aesthetic so that the spatial sequence
enologist Edgar John Rubin, the space or the object could be sufficiently perceived. The selection of the
appear alternately. In such a situation, design is only resulting objects presented here deals with these de-
possible iteratively. A person works on one and checks manding requirements in different ways. However, as in
the effects on the other. If the changes are assessed the previous exercise, all examples achieved a very high
negatively, this intervention is withdrawn and a different quality of design. They are able to intensively occupy
approach is attempted. Only with increasing knowledge the observer, but—due to the inherent geometric log-
about the impact can a more targeted and holistic ic—they can be grasped both as overall compositions
approach be taken. This task is therefore very close and on the level of the individual elements, a sculptural
to an architectural problem in which different param- achievement that is also a fundamental characteristic
eters have to be mutually dependent, although they of architecture.
References
1
Victor Vasarely, Gespräche mit Victor Vasarely, Jean-Louis Ferrier, Spiegelschrift 8, Verlag
Galerie der Spiegel, Köln 1971, p. 155.
Christian Kern was born in 1964 in Wipperfürth, West Germany. Apprenticed as a machine
fitter, studied architecture at the TU Stuttgart and the Curtain University Perth, West
Australia. Collaboration at Stirling and Wilford Stuttgart, Ken Yeang Malaysia, Behnisch and
Partner Stuttgart, Auer + Weber Stuttgart, Meier-Scupin & Petzet Munich. Scientific assistant
at the chair for building theory and product development, Prof. Richard Horden, TU Munich.
Since 1998 own office in Munich, 2001 founding of BLAUWERK Architects with Michael
Schneider, since 2008 with Tom Repper. Board member of Europan Germany e.V. Since
2007 Professor at the Vienna University of Technology, head of the department for three-
dimensional design and model making.

3D Parquet Deformation
The aim of this exercise was to extend the discussion of 2D tiling and 2D
parquet deformation (William S. Huff), which is well anchored in systema
tized basic design theory, to the concept of 3D tiling and 3D parquet
deformation. Students were to become familiar with the groups of con-
gruent figures filling a plane or space, their topological relationships,
and the principles of continuous deformation (see, for example, Dürer
and D’Arcy Thompson), but also to develop a coherent and aesthetically
pleasing 3D composition. Finally, a three-dimensional form study was
developed based on the principle of parquet deformation. The semester-
long exercise was divided into four individual exercises, the results of
which are presented in the following pages.
Exercise 1
2D Parquets and 2D Parquet Deformation
The first exercise started with lattice structures based on regular/platonic

polygons (triangle, square, and hexagon) and other simple polygons (e.g.,
rhombus, rectangle, isosceles triangle), and their transformation possibili-
ties were studied. Based on one of the simple polygons (basic element), a
grid was created by hand drawing. In three steps, students “breathed life”
into the grid by deforming or transforming it in different ways.
1. Transformation of the Basic Element
The goal was to develop new protocells apt for tiling, based on a constant
grid structure. This goal could be achieved by different operations: short-
ening or lengthening a line; introducing a “kink”, a “bump”, or a “nub” (pene
trating or protruding a simple shape); subdivision (subdividing); assigning a
different location for a corner point; changing a straight edge to a curved
one; or drawing a diagonal or circles. These newly developed cells were cut
out of cardboard to playfully explore their possible combinations (study of
symmetry groups based on rotation, translation, reflection).

Marijana Zivkovic
Tutor: Werner Van Hoeydonck
Mehmet Semih Özcelik

Tutor: Martina Kögl

Irem Akcay
Tutor: Manuela Fritz
Zeynep Dikmen
184 3D Parquet Deformation Exercise 1 , 2D Parquets: 1. Transformation of the Basic Element

Jakob Kandelsdorfer
Tutor: Christoph Meier
Leonardo Haglmüller
Tutor: Christian Kern
Michael Robert Jimenez

Eunice Gomes Alexandre

Tutor: Fridolin Welte
Sarah Bochis

Sarah Bochis
Noura Omar
Tutor: Nora Fröhlich
Johanna Himmelbauer
Daniel Koller
Niklas Hörburger

Cinthia Anton
Jan Wucherpfennig
Ruben Mahler
Laurenz Katamay
Tutor: Anita Aigner
Ye-Ryun Kim
Tutor: Anita Aigner

Ye-Ryun Kim
Tutor: Anita Aigner
Ye-Ryun Kim
Tutor: Anita Aigner
Tobias Dirsch
Tutor: Anita Aigner
Ceylan Elenor Ergelen

Maximilian Wolfram
Tutor: Anita Aigner

Laura Huber
Tutor: Anita Aigner
Laura Huber
Tutor: Anita Aigner
Markus Biel
Tutor: Anita Aigner
Daria Lanina
Tutor: Anita Aigner
Alexander Ladentrog

Exercise 1
2. Continuous Deformation
Within the existing basic structure, the cells could be transformed in one
or two directions or even from a center. The considerations of transform-
ing the basic element in Step 1 often already included an approach to
“movement” and suggested a stepwise transformation of the individual
elements. However, a transition between two (or more) periodic tilings
could also be created (i.e., an interpolation between initial patterns).
Alexandra Konstantinova
Tutor: Anita Aigner
196 3D Parquet Deformation Exercise 1 , 2D Parquets: 2. Continuous Deformation

Leonidas Peithner Damjan Veličković
Tutor: Anita Aigner Tutor: Judith P. Fischer

Tobias Speckner
Alexander Keil
Tutor: Anita Aigner
Tobias Dirsch
Tutor: Anita Aigner

Anja Bezjak
Alexandra Konstantinova
Tutor: Anita Aigner
Ye-Ryun Kim
Tutor: Anita Aigner

Theresa Thaler
Tutor: Peter G. Auer
Karlo Keca
Sana Halimovic

Merve Vural
Moira Ruppert
Mahir Kurtalić

Christian Mitschdörfer
Tutor: Judith P. Fischer
Valentina König
Roman Morozow

Chiara Huf
Andreas Frank
Damjan Veličković

Christoph Paul Hofmann
Tutor: Anita Aigner
Daria Lanina
Tutor: Anita Aigner

Sarah Reithofer
Michael Bachmeier

Exercise 1
3. Deformation of the Basic Structure
A “moving” tiling could also be achieved by transforming the constituent

lattice structure. Different possibilities were explored for this purpose.
Zorana Sotirov
206 3D Parquet Deformation Exercise 1 , 2D Parquets: 3. Deformation of the Basic Structure

Steffen Alexander Blickle

Susanne Stampf
Tutor: Markus Bauer
Damjan Veličković
208 3D Parquet Deformation Exercise 1 , 2D Parquets: 3. Deformation Of the Basic Structure

Andrea Di Tommaso
Ye-Ryun Kim
Tutor: Anita Aigner

Samuel Huber-Huber
Mahir Kurtalić

Karlo Keca
Patrik Marchhart
Roman Morozov

Claudia Pitterle
Anja Bezjak
Isa Kirchberger

Sacha De Simony
Christian Mitschdörfer
Michael Haidinger

Exercise 2
The study of tiling and parquet deformation was then continued in

three-dimensional space. For this purpose, simple polyhedral (e.g., cube,
triangular and hexagonal prism, twisted double wedge (gyrobifastigium),
truncated octahedron, rhombic dodecahedron) and the spatial lattice
structures based on them were assumed. As in Exercise 1, the initial
cell was to be transformed first, followed by the entire packing (periodic
space filling). This exercise was divided into three steps:
1. Transformation of the Basic Element (Polyhedron)

The shape of the initial cell was changed to create a penetration-
free and gapless 3D-tiling of new, congruent (proto)space cells. The
reshaping could be of varying complexity (i.e., in one, two, or three
directions) and could be brought about by different operations (e.g.,
bending in or out, dividing, subdividing). Space fillings based on two
protocells were also allowed. Enough cells should be generated
to occupy the aggregation of the transformed protocell, and the
combination possibilities could be investigated.
2. Deformation of the 3D Space-Filling Structure

In the second step, a 3D space-filling structure consisting of the same
elements was continuously deformed (at least in one direction, possibly
also starting from a center). For this purpose, rules for a continuous
transformation had to be established, and consideration had to be
given to which surfaces, lines, or points could be “moved” or displaced
step by step. Different variants and construction methods were
experimented with. A transformation of the constituent spatial grid
structure was also possible.
3. Representation/Mode of Construction
Transformed basic elements could be represented, for example, as a
sheath unwrapped from cardboard or as a mass model (e.g., XPS/extruded
rigid polystyrene foam). However, since a “closed” stacking of solids
consisting of massive volumes cannot be viewed, the representation
of 3D lattice structures as a rod model was recommended. In addition,
alternation of empty space and mass volumes or, in the case of
aggregations based on two protocells, representation of only one group of
cells should be considered.
214 3D Parquet Deformation Exercise 2, 3D Parquets and 3D Parquet Deformation | Exercise 3, Design Concept
Exercise 3
Design Concept
The third exercise step was to develop a form study based on 3D parquet
deformation. The design was to be developed from previous studies, a
completely new approach was also allowed.
Concept
It was essential that the design was based on a content-related idea, a
design concept. It had to be possible to name the concept in the design
of the 3D parquet deformation. The concept could be the representation
of a movement process, a certain theme or image (e.g., interlocking,
growing out of a cell), or an abstract, mathematical-geometric rule or
code. A central requirement was that the monotony of a 3D parquetry
was broken up. The element of deformation or transformation should
also have a visible effect (i.e., not be a minor matter).
Composition
The 3D parquet deformation, which was in principle infinite, had to be
limited according to compositional aspects. Depending on the design
approach, longitudinally aligned compositions, cubically framed com-
positions, or compositions organized from a center were conceivable.
The elements should be balanced in number and size (not too small,
not too large) and above all help to express the conceptual idea. Visually
incomprehensible deformations or unintelligible aggregations were to be
avoided. In this respect, the construction (representation of edges, sur-
faces, or volumes) was also an essential component of compositional
considerations. The form study was to be conceived as all-view; that is, it
was not to be designed for one main viewpoint but had to be attractive
from all sides. Different spatial positions and presentation possibilities
(e.g., plinths) were to be explored, as they are part of the composition and
should be proportioned and designed in relation to the object. The dimen-
sioning of the form study was free, but the dimension 24 × 24 × 24+n cm
was considered as an orientation for the composition space.

Jennifer Berger
Mahir Kurtalić
extrusion of the shape

initial form
(2D tessellation) different depths and heights
progress create an overall impression
pull upwards (deformation) “heartbeat”
pull downwards
push inwards
Noura Omar
Tobias Speckner
Selma Dervisefendic

Steliyana Chipeva
Christoph Paul Hofmann Jakob Kandelsdorfer
Tutor: Anita Aigner Tutor: Christoph Meier

Theresa Steinberger

Exercise 4
Presentation Model
The last step of the exercise was the material realization of the design
under the premises of 3D parquet deformation into a sophisticated
presentation model.
For the compositional fine-tuning, we first checked whether
the essential aspects of the design idea had been concisely conceived.
In some cases, the object was optimized with regard to dimensioning,
cut-out/boundary, number of cells, and formal expression. The next step
was to find the ideal construction method and the ideal material for
the form study. With the construction method, students had to make
compositional decisions. Depending on whether edges, surfaces, or
volumes were represented, the formal expression changed. The decision
of whether adjacent cells were represented individually or combined into
surfaces also had an effect. With the decision for a certain representation,
a design idea could be aesthetically elevated, but also weakened.
Therefore, one had to carefully consider which type of construction (rod
model, structural model, or shell unwinding or moldable materials for the
representation of fully plastic volumes) was best suited for the respective
form study. The effect of the form also heavily depended on the spatial
position as well as on a reference system (plinth or base plate).
222 3D Parquet Deformation Exercise 4, Presentation Model

Jennifer Berger
Pyramids are inserted in a cube and placed on
opposite faces in a take-away-and-add process.
The resulting solid is deformed along the x-axis.
Subsequent and subtle downscaling occurs until half
of the initial height generates a series of connectable
solids, which is done using threads, allowing the
wall-like stacking to move and bend.
Space Tessellations Teaching Perspectives Multiply 223

Andre Nadtochyi
Tutor: Christoph Bruckner
A wave line is doubled and straightened, divided in
three parts with two hinges and three rotation points.
A sequence of differently-hinged elements is created
and stacked in a step-like manner reminiscent of a fly-
through cathedral vault on three levels, defined by the
initial wave lines appearing on the outer side.
224 Multiply 3D Parquet Deformation Exercise 4 , Presentation Model

Kristina Hendker
Tutor: Markus Bauer
A solid 3D interpretation of the so-called Miura
folding, named for its inventor, the Japanese
astrophysicist Miura. The crease pattern, a
combination of straight and zig-zag lines, is used
to create a stacking that is able to contract and
expand, a true 3D origami.

Barbara Posch
Two opposing vertices of the rhombic dodeca-
hedron are pushed inward, creating a new shape
that is subdivided into eight parts and rearranged,
respecting the original space-filling aggregation of
the rhombic dodecahedron.
226 Composition 3D Parquet Deformation Exercise 4 , Presentation Model

Amer Mahmoud
A constellation based on the rhombic dodecahedron
that is divided into smaller interlocking cells so that
the totality of deformed interlocking cells has the
same volume as the original shape, which becomes
indiscernible.
Space Tessellations Teaching Perspectives Composition 227

Mirkovic Radovan
Inspired by mushrooms, palm trees, or bones,
acrylic and plaster cast spheres interact by reflec-
tion and rotation operations according to a square
grid. Additional elements are introduced in an
upward-spiraling arrangement.
228 Composition 3D Parquet Deformation Exercise 4 , Presentation Model

Lukas Hansmann
The truncated octahedron is subdivided into four
and six identical shapes. Recomposed, they form
a rhombic dodecahedron. The void in the middle
can be filled with three half-parts of the truncated
octahedron.
Space Tessellations Teaching Perspectives Composition 229

Selma Dervisefendic
Inspired by a Penrose tiling of golden rhombi, the
rhombic hexecontahedron is dissected into 20 acute
golden rhombohedra, the rhombic triacontahedron
into ten acute and ten obtuse golden rhombohedra,
which are recombined into an infinitely extendible
crystal-like structure.
230 Dissolving 3D Parquet Deformation Exercise 4 , Presentation Model

Ovcina Hajrudin
The Weaire-Phelan structure consisting of pyritohe-
dra and truncated hexagonal trapezohedra is sliced
into half multiple times, replicated, and—through a
process of addition and subtraction—recomposed
into an arch-like structure.
Space Tessellations Teaching Perspectives Dissolving 231

Mahir Kurtalić
An asymmetrical distortion of the cube is cut in half.
Inspired by aerodynamic and aquadynamic shapes
such as feathers and shark scales, its subparts are
switched in position and rearranged to give a maxi-
mum of tension and movement toward dissolution.
232 Dissolving 3D Parquet Deformation Exercise 4 , Presentation Model

Steliyana Chipeva
Hexagons and (invisible) squares are connected by
concave and convex bends of copper, resulting in a
dissolving spatial composition reminiscent of an ab-
stract bouquet of flowers, reflecting its surroundings
in different color nuances.
Space Tessellations Teaching Perspectives Dissolving 233

Anja Bezjak
A distorted hexagonal prism is multiplied, mirrored,
and recomposed. Inspired by quarries and rice
terraces, a dynamic landscape is created in an ever-
changing play of light and shadow complemented by
gentle surfaces, balancing the work as a whole.
234 Gradual Changes 3D Parquet Deformation Exercise 4 , Presentation Model

Steffen Alexander Blickle
A sequence of three vertex motion operations along
the diagonal of the cube generates sharp sides,
infiltrating its adjacent cells in a gradual give-and-
take process along the diagonals of the cube, made
visible by distancing the four aggregations.
Space Tessellations Teaching Perspectives Gradual Changes 235

Philip Kaloumenos
Tutor: Anita Aigner
In a 3 × 3 × 9 modular cubic structure, starting with
completely closed cubes, a progressive subtle vertex
motion operation results in foldings of the cubes’ sur-
faces, generating gradually growing triangular openings
in a give-and-take process of lightened-up space.

Berkay Yozgyur
Tutor: Anita Aigner
From rest to movement in six steps: Some of the
cube’s vertices travel in space in a forward-upward
continuous movement, transforming the cube’s
static faces in a space-filling aggregation of ever-
changing spatial triangles, inspired by a sprinter
coming into action.

Christoph Paul Hofmann
Tutor: Anita Aigner
Two rows of five cubes undergo a twofold left-and-
down deformation. The cube’s gradual multidirec-
tional vertex motion is reflected in an equivalent
concave deformation of the upper side. The spiky
concave drift down instigates its concave counter-
part above.

Thomas Ran
A three-step gradual intrusion of the rhombic
dodecahedron in opposite vertical directions
generates negatives spaces that look like positive
spaces. The viewer’s eye is challenged to experience
a 3D flip-over reminiscent of the figure-ground
problem in 2D.

Alexander Keil
Tutor: Anita Aigner
A cube is quartered at its top and bottom, mimicking
a checkerboard of positive and negative space. The
two diagonally opposite surfaces shift upward and
the grid structure elongates: The previously shifted
surfaces become progressively extruded. Gradually,
the initial cube becomes a loop-like shape.

Theresa Gasteiger
Eight faces of the twelve-sided rhombic dodeca
hedron are rotated along their short diagonals,
generating five new faces, partly concave, partly
convex. The reclined 1 × 3 × 3 module stacking shows
the original form at its base and two rows deformed
by gradual face rotations.

Tobias Speckner
Extended tetragonal bipyramids combined with
(invisible) tetrahedra fill space. A vertex motion
operation by half a width on the x-axis on all four
corners (in opposite directions) of the bipyramids
fuses with a one-unit-high rotation of the
composition’s two upper horizontal axes.

Markus Biel
Tutor: Anita Aigner
Cubes are distorted by different interrelated vertex
motion operations and stacked in a distorted
3 × 3 × 5 modular cubic structure. By omitting cubes
in a 3D chessboard-like way, an interplay of light
and shadow enhances the spatial exploration of the
connected voids.

Noura Omar
Three rhombi are extruded. Deformed cubes and
rhombic prisms of different sizes and lengths are
created by folding edges up and down, inward
and outward, in order to express a dynamic
multidirectional 3D heartbeat.
244 In Motion 3D Parquet Deformation Exercise 4 , Presentation Model

Isa Kirchberger
Inspired by animal scales, a cube is gradually
reshaped, allowing for multiple and subsequent
changes in size, direction, and overlap, permitting a
scale structure to open and close and re-orientate
according to seemingly effortless, instinctive
movements.
Space Tessellations Teaching Perspectives In Motion 245

Lisa Hirsch
All lines, starting from the outer square, are bent
inward; all lines starting from the inner square are
bent outward. In this way, the two-dimensional
tessellation becomes a three-dimensional sculptural
form that can be expanded infinitely.
246 Materiality Matters 3D Parquet Deformation Exercise 4 , Presentation Model

Jakob Kandelsdorfer
Sixty-four identical oblique three-sided prisms in mas-
sive grained wood are packed in two directions, giving
every cell its identity. The regularity is broken through a
tilted box which trims away three parts of the stacking,
revealing new connections between the cells.
Space Tessellations Teaching Perspectives Materiality Matters 247

Maximilian Greinwald
Opposite faces of the cube are respectively bent in
and bent out, resulting in a cubic brass rod structure
that kinks and bumps through space according to
a growth rule. Different levels are made visible by a
series of parallel connecting rods. Three predefined
positions on its plaster base induce drastic
perception shifts.
248 On Stage 3D Parquet Deformation Exercise 4 , Presentation Model

Christine Bosse-Büchling
A subdivision of the cube along diagonals creates
five smaller forms that can be recomposed in eight
different ways. One of eight possibilities is selected
and deformed in several directions, a larger cube, the
initial shape of the deformation frames and holds
the composition together.
Space Tessellations Teaching Perspectives On Stage 249

Andreas Frank
The rhombic dodecahedron is divided into four
parallelepipeds along all its space diagonals. Extruded
parallelepipeds change direction and length in order to
make them intertwine. These “threads” soften the strict
geometric shapes into a kind of spatial textile.
250 On Stage 3D Parquet Deformation Exercise 4 , Presentation Model

A 3D tiling is created out of a continuous
deformation of three 2D parquets stacked on top of
each other, whereby the top and bottom slices are
the same form. A “loop” is created by starting at Tile
A, going to B and C, and then backward from C to
B to A. To interlock, a second 3D tile is created. This
time the loop is defined by CBABC.
Space Tessellations Teaching Perspectives On Stage 251

Cellular Space Sequences
Looking at the densest possible packing of space, the cell (from Latin
cella = small space) is the epitome of a determining individual element.
Its occurrence is manifold in nature and culture. In Western architecture,
the term appears early in history (Romans). Ideas of urban densification
in the course of the urgent need for housing after the First World War
more intensively examined the notion of the cell. This investigation ranges
from beginnings in the interwar period (Bauhaus) to high points in the
1950s/1960s with structuralist approaches (Herzberger, Tange, Candilis)
and the radical Raumstadt concepts (Constant, Friedman, Schultze-
Fielitz, etc.). At present, the desire for urban densification, mixing of
functions and population pressure seem to be creating space again for
approaches regarding the notion of the cell.
In architecture, the connection of spaces does not only serve
functional purposes. The quality of the connections with a choreographed
sequence of homogeneous or heterogeneous spaces plays a weighty role,
as can be seen in many architectural path concepts (e.g., Egyptian temple
complexes, Baroque enfilade, Le Corbusier’s Promenade Architecturale, J.
Frank, R. Koolhaas). In the winter semester 2018/19, we devoted ourselves
to designing and creating interesting spatial sequences within a specific
matrix. This matrix is fundamentally cellular in structure and determined
by a gapless, three-dimensional tiling of space with the help of convex
polyhedra. We worked in a defined systematics and, in order to get size
variance for the space-filling polyhedra, methodically made use of the
principle of three-dimensional parquet deformation and subdivision, such
as the self-similarity of fractals.
The semester-long exercise was an elementary three-dimensional
design task on the above topic, freed from functional constraints. This
exercise included, among other things, the creation of spatial hierarchies
in the sense of a scenographic sequence as well as the conception
of large and small spaces and objects in a given systematic structure
(cellular matrix).
The exercise started two-dimensionally with clear rules in order
to gradually arrive at more individual design strategies, rules, and formal
solutions. At the end of the assignment, real three-dimensional, allround
attractive form studies with a sculptural effect were produced. The exer-
cise was divided into four individual exercises, the results of which are
presented in the following pages.

Exercise 1
Figure Ground
The first exercise started with the choice of one of the following simple
polygons: square, rhombus, equilateral triangle, regular hexagon, or cairo
tiling. The chosen polygon served as a protocell for a tiling of the plane.
Following a comprehensible concept, an inner space with an opening
to the outside on only one side of the polygon was drawn in the chosen
protocell. The shape of the internal space and the opening could be
freely chosen, but both referred to the inherent structure of the initial
cell. Attention was paid to the quality of this inner space in relation
to the protocell in terms of proportion, tension, and figure/ground
relationships.
The protocell complemented with an internal space was to be
used in a planar tiling, large enough (not less than 20 elements) so that
the figure/ground relationships as well as the visual effects of this surface
filling could be understood and assessed. When the choice of a certain
protocell allowed it, the possibilities of different congruence mappings
(sliding, rotating, mirroring, glide-mirroring) were used. Even if the tiling
was theoretically possible to infinity, only a limited section with a simple,
clear outer contour was to be used. This section could correspond to
the contour of the protocell (e.g., square for squares). The result was a
hand drawing in which the contour lines of the protocell should be clearly
legible and the “wall” areas should not be blackened, but highlighted with
pencil as hatching.
The result of such tiling with only one lateral opening resulted in
inaccessible interior spaces. In order to generate spatial sequences in the
tiling, individual cells (maximum four to six) were first regularly shifted
outwards from the bond to create gaps. This process created new spatial
possibilities in the outer area that broke up the strict outer contour. In
a final step, simple interventions were implemented to connect the
remaining interior spaces that were not yet connected to create an
interesting spatial sequence. This spatial sequence became visible by
drawing a dotted line (trajectory). The spectrum of such interventions
ranged from conceptually well-conceived to completely arbitrary. The
latter was to be avoided in order to get an interesting result.
254 Cellular Space Sequences Exercise 1 , Figure Ground

Cristina Cazacu
Antonia Maisch

Maria Oikonomou
Gabriel Esposito

Nikola Stevanovic
Elena Thöni

Zsofia Arnhoffer
Tutor: Anita Aigner
Charlotte Hemmen

Svetoslava Svetoslavova
Tutor: Anita Aigner
Johanna Grabner
Tutor: Markus Bauer

Karin Riedl
Tutor: Anita Aigner
Stefan Binder

Tanja Punz
Tutor: Anita Aigner
Jana Riernössl

Exercise 2
Solid and Void
From the following convex polyhedra, which allow a gapless and over-
lap-free tiling of the three-dimensional space, one was chosen for further
processing: cube (hexahedron), twisted double wedge (gyrobifastigium),
or rhombic dodecahedron. The dual form was realized by connecting the
centers of adjacent polyhedron faces of the initial solid. These straight
connecting lines form the edges of the dual solid.
Two separate working models were created, a rod model of the
initial polyhedron (size approximately to fit in a sphere Ø15 cm) and a
solid model of its dual. The spatial and plastic qualities of the connection
between the two solids and their spatial positions were then examined
with sketches. In the next step, the dual solid was subtracted from the
initial polyhedron. The result was a shell solid with an interior without
openings. From this solid, a working model divided by a deliberate cut
was built as a shell model, which we called the "avocado" model.
The dual solid was only one of many conceptual possibilities in
the search of an interesting interior space. To explore their own ideas, the
students selected another polyhedron from the group outlined above. For
this polyhedron, interiors were developed experimentally and following
individual rules. It was possible to fall back on approaches from Exercise
1 or to experiment with the dual solid. The shape of the interior could be
freely determined. The aim was to create an exciting and conceptually
well-conceived interior space that corresponded to the individual formal
inclinations in relation to the form of the original solid. Experiments were
conducted not only on the conceptual level (e.g., drawings, sketches), but
also on the level of different materials and construction methods.
262 Cellular Space Sequences Exercise 2 , Solid and Void

Thomas Emil Rasmus

Antonia Maisch

Patrick Neuwirth

1. 2.
3. 4.
5.
Aia Metnan

Exercise 3
Composition and Design
In exercise 2, different interior spaces were investigated and developed

for a chosen initial polyhedron. These polyhedra were multiplied to cre-
ate a space filling with an appropriate number of cells in all three spatial
directions. On the one hand, the result could be a dense packing, cell to
cell. On the other hand, by omitting certain cells, loosened conglomera-
tions could be formed without the disintegration of the structure.
For the compositional development of the form study in space,
it was necessary to develop and define a basic idea or design concept.
For this purpose, a virtual compositional space of 24 cm × 24 cm ×(24+n)
cm was proposed as a basic contextual guide (cube or cuboid). Based on
the conditions of gravity, a multitude of possible spatial positions arose.
Some possible cases were investigated with the help of sketches and on
the basis of working models.
In order to produce well-designed sequences of spaces, the inte-
rior spaces had to be deliberately opened up and connected. Based on
the idea of being able to “fly” through these spaces, the spatial properties
of the interior spaces were examined by means of perspective sketches.
Essential parameters influencing the quality of the spatial sequences,
such as the size and shape of the spaces (scaling), could be controlled,
among other methods, by three-dimensional parquet deformation or
by fractal subdivision. The form study was to be designed all-view. An
important problem was the representation of these spatial sequences
for an outside observer. By what creative means could the glances and
staged sequences of glances allow for conveying an understanding of the
choreographed deep nesting of the interior spaces?
268 Cellular Space Sequences Exercise 3 , Composition and Design

Gisela Eder

Patrick Neuwirth
Alea Sokya

Antonia Maisch
Cristina Cazacu

Franziska Veith
Joline Imwolde

Kim Gubbini
Tutor: Anita Aigner
Elisabeth Anna Prantner


Exercise 4
Presentation Model
In the last step of the exercise, the cellular space sequence was to be
converted into a sophisticated and aesthetically convincing presentation
model. The task was to check whether the essential aspects of the
design idea had been concisely elaborated on. Was the object already
optimized in terms of spatial position, cut-out/boundary, number of
cells, and expression of form? Could an outside observer see and
understand the quality of the spatial sequences? The compositional
fine-tuning had to be related to the chosen construction method. Which
type of construction was most suitable for the particular form study:
rod/structure model, or representation of fully plastic volumes through
shell unwinding or malleable materials? The effect of a form also heavily
depends on the spatial position/positioning and a reference system,
such as a plinth or base plate.
The aesthetic value of the form study was largely determined
by the appropriate treatment of the material chosen in each case. This
included precise craftsmanship in the execution, degree of abstraction,
and also the surface features determined by light and shadow (e.g.,
structure and texture, edges, contours, convexities and concavities,
roughness, gloss). A homogeneous appearance was desired, preferably
in white. Supporting and auxiliary constructions foreign to the object
were to be avoided so that the presentation model could convince in
its compositional development in space, in relation to form and content
and speak for itself.
274 Cellular Space Sequences Exercise 4 , Presentation Model

Antonia Maisch
Starting from a concrete base, the rhombic
dodecahedron is arranged to maximize view axes.
Brass rods define the borders and allow insight
into the carefully designed crystalline inner spaces.
Every inner space and every connection is designed
individually to optimize the potential of multiple
spatial sequences.
Space Tessellations Teaching Perspectives Balance 275

Thomas Emil Rasmus
An octagonal bipyramid is subtracted from a
gyrobifastigium, creating an interior space with four
linear openings. Increasing dimensions results in
wider connecting openings. An interrupting “wall”
acts as a borderline where bipyramids take over
control in their contrasting solidity.
276 Balance Cellular Space Sequences Exercise 4 , Presentation Model

Gisela Eder
Two overlapping and deformed square bipyramids
carve out a multidirectional complex space into the
gyrobifastigium. Triangular “over the edge” openings
bring in light, fly-troughs, and see-troughs. By
twisting and displacing this concept, a constantly
varying dynamic spatial sequence is created.
Space Tessellations Teaching Perspectives Balance 277

Cristina Cazacu
By connecting midpoints, thirds, and fourths of edges
and diagonals, the cube becomes almost indiscernible.
Created by a similar procedure, every interior space has
its unique play of light and shadow. In aggregation, a
complexity of perspectives, meant to be navigated by
close-up visual analysis, displays a multiplicity of insights.
278 Crystalline Cellular Space Sequences Exercise 4 , Presentation Model

Donart Gallapeni
In order to direct visual connections, the outer
surfaces are deliberately left closed or open up.
Spatial sequences branch into different directions.
Crystalline and convex carvings subdivide the cube;
spaces and spatial sequences emerge in contrast
to massive leftovers.
Space Tessellations Teaching Perspectives Crystalline 279

Patrick Neuwirth
The rotation and translation of the dual solid of the
gyrobifastigium around its square faces is explored,
the initial shells are broken up, and a set of 18
different interior spaces appears, four of which are
chosen and put in aggregation by turning them 90°.
Extrusions of polygonal faces of their interiors allow
complex, unique connections.

Lena Roth
The cube is decomposed into 15 polygonal parts
by a Voronoi operation. Seven parts are removed
to create an interior space that can be further
deformed by taking away more parts. Asymmetric
repetitions and mirroring operations build an arch-
like portal intended to draw the viewer in.

Jakob Fitz
Two identical double wedges that are rotated by 90
degrees intersect, resulting in two identical cutting
patterns on opposite faces. Aggregated four times,
this creates a nested, complex route inside and
negative spaces between the individual modules and
the stand area.

Orsolya Nyulas
A crystal-shaped space is extracted from the
cube, creating multiple openings on the sides.
The cut cube is first mirrored along the x-y axis,
then multiplied by rotations of 90°, 180°, and 270°
degrees, resulting in a larger cube with a twisted
continuous interior space.

Theresa Steinberger
Center points of the cubes’ edges are connected
to form different triangles, creating two interior
spaces with dissimilar triangular surfaces, indi
vidually connected in a 3 × 3 × 3 modular graduated
arrangement. To allow insights, certain elements of
the volume are omitted.

Baraa Hachicho
Named by Michael Goldberg after a playing card,
the “ten-of-diamonds” decahedron consists of
eight triangles and two rhombi. By connecting
vertices with one-third of their opposite edges, an
interior space is created in complete geometrical
concordance with its initial shape, interacting in a
mysterious mathematical balance.

Elisabeth Anna Prantner
Intersections of important connecting lines within
a stacking of rhombic dodecahedra defined by fine
brass rods create a multitude of fascinating small
crystal-like brass solids, not to be entered, but to be
experienced from the outside in an ever-changing
play of light, shadow, and reflections.
286 Materiality Matters Cellular Space Sequences Exercise 4 , Presentation Model

Bernarda Ehrenhöfer
A sphere on the cube’s midpoint—with increasing
radius from bottom to top—intersects with eight
spheres placed with their midpoints on the vertices
of the cube, thereby creating circular insights. In a
stack of three such cubes, diagonal connections
generate a concatenation of spatial sequences to be
explored.
Space Tessellations Teaching Perspectives Materiality Matters 287

Franziska Veith
The dual of the gyrobifastigium is deformed by
mirror, rotation, and vertex motion operations, then
scaled up, penetrating its initial form. Small openings
lead to large inner spaces and large openings into
a void. Cold concrete meets warm gold, static
massiveness meets decay, and smooth meets
uneven in a union of opposites.
288 Materiality Matters Cellular Space Sequences Exercise 4 , Presentation Model

Esraa Metnan
On three locations, center points of surfaces and
edges of the truncated octahedron are connected
with vertices, resulting in three openings. The tetrakis
hexahedron, the dual of the truncated octahedron, is
joined with these openings to create an inner space
that allows for a great variety of spatial sequences.

Mathias Winder
The cube—inert, static, without any notion of
movement—is opened up on all six surfaces, two
vertices are removed, and a series of geometric
operations leads to space-defining polygonal convex
solids of various sizes. Mirrored in all possible
directions, the crystal-like protocells are aggregated
to convey an upward motion of spatial sequences.
290 Multiply Cellular Space Sequences Exercise 4 , Presentation Model

Emma Katarina Kaufmann-LaDuc
Tutor: Anita Aigner
Triangles are extruded into prisms within the cube,
the apex of each triangle a quarter notch away from
the cube’s face. What is left of the cube remains
solid; the prisms define the openings. Inspired by
the philodendron, meaning “tree-loving”, the base
assumes the role of the tree, allowing the units to
(visually) descend again.

Regina Bednar
Tutor: Markus Bauer
A symmetrical connection of vertices and midpoints
of the edges of the cube results in points of inter-
section on the cube’s faces. Joining them results in
an inner space that is blown up four times in in a
stack of a 3 x 3 x 3 modular structure of cubes until
inner spaces become outer space.
292 Gradual Changes Cellular Space Sequences Exercise 4 , Presentation Model

Kim Gubbini
Tutor: Anita Aigner
Twelve cubes stacked in a 3 x 4 modular structure
provide the basic structure. Five of six faces of each
cube are removed. A rotated square is connected
by plane surfaces with different smaller square
openings on the opposite corner side of the cube
to create a variety of torsions, negative space, and
insights.

Gabriel Esposito
Midpoints of four of the cube’s vertical faces are
connected with vertices and with the midpoint of the
cube’s horizontal faces, resulting in eight solids with
four triangular sides and four apexes. Connected only
with their apexes, a spatial sequence of 128 of these
solids creates a symmetrical fragile-looking space.
294 Open Up the Inside Cellular Space Sequences Exercise 4 , Presentation Model
Denise Redl
A spatial sequence of twisted double-wedges
is opened up on its square faces to allow a
free view on icosahedral inner spaces that are
continuously deformed by dropping one or more
of its triangular faces and then connected to
allow a fly-through path.
Space Tessellations Teaching Perspectives Open Up the Inside 295

Milena Vogl
The twisted double-wedge consists of both
triangular and square faces. The squares are rotated,
size-wise, around a corner of the triangular face and
reduced as they go more inward. If one of these
twisted double wedges is lined up with another,
other spaces are created that lead from the small to
the large and back to the small again, or vice versa.
Aia Metnan
The hendecahedron, a space-filling polyhedron with
eleven faces, has three rectangles in its interior.
These are connected through the introduction of
new connecting lines from certain midpoints. Com-
ponents are taken away to create openings. The
space sequence shows a 3D Cairo tiling, from which
selected solids are removed.
Space Tessellations Teaching Perspectives Open Up the Inside 297

Maria Oikonomou
Points on each of the gyrobifastigium’s surfaces are
connected to form an inner shape. Distorting the
inner shape so that it breaks out of the fastigium
creates openings on some of its surfaces and results
in a new unit. This protocell serves as a building
block for a 3D space-filling, which is cut in the
middle to expose its inner chaos.
Epilogue
Christian Kern
In Research Perspectives, the first part of the book, we have tried to explore the historical
context and possible futures of 2D parquet deformations.
In Teaching Perspectives, the second part of our book, we have selected inspiring works that
we judged to be a valuable answer to our research question: to express the idea of a subtly
developing tile into a subtly developing 3D-tiling.
William Huff pointed out that a parquet deformation forces a more spatiotemporal
perception instead of an immediate overviewing understanding. If we had defined too strict
rules without allowing interpretations, the results could maybe have been more systematic
and consistent, coming closer to what a parquet deformation in 3D should look like. Since it
was our first experiment in this domain, our primary goal was to experiment, to give freedom
in strategies on how this spatiotemporal idea could be provoked. In our second experiment
“Space Sequences”, the idea of perception was transferred into a more physical, architectural
experience, as if we were able to “fly” through these forms. This second assignment made
students more aware of figure-ground relationships, in between spaces, what can be done
with the interior of polygons and polyhedra and how to connect them to design a complex
choreographed space sequence. In both experiments we tried to encourage the students to
refrain from (without forbidding it) the use of CAD in order to enable a more lasting hands-
on experience and thus generate a greater understanding of strategy and form.
Student feedback indicated they were very content with the learning effect generated
and the positive impact of the fusion of geometrical constraints (the grids, lattices and
space-filling structures) and personal creativity. Great benefits to both assignments were
reaped by the students: not only the new insight gained concerning 2D-tilings and space-
filling structures but also how to subtly transform, perceive and connect them.
William Huff’s legacy is so important because it forces students to explore parametrical
design without the aid of computers. In the last century, mathematicians contributed
immensely to 2D and 3D topology and geometry. Although many architects today do use
(mostly computer-generated) patterns in their designs, basic concepts of topology and
tessellations are still underrepresented in architectural and artistic education. We truly hope
that our book will help to fill this gap and inspire students, architects, artists and designers
all around the world.
301
Acknowledgments
Christian Kern
The observation and analysis in the work shown here is very inspiring and enriching for me
as a designer and architect in relation to my own work and research in the field. The text
contributions enable me to gain a deeper understanding of the subject in mathematical,
artistic, and philosophical terms. I hope that the readers feel my same joy about the
experiments, results, and reflections shown here. It has always been important to us to
share these results with designers, mathematicians, artists, and all other interested parties.
I would particularly like to thank Werner Van Hoeydonck, who archived and structured
the results with great personal commitment and thus made the realization of the book
possible. His expertise in geometrical patterns results from many years of creative and
theoretical research in this area. A large portion of the texts and important contributions
about the concept of the book came from Werner. In publications and at congresses, Werner
also established a connection between our discussions and the contemporary research
environment. In doing so, contacts were established or deepened with researchers and
artists who later offered contributions to this book.
My thanks also go to Eva Sommeregger, who—though originally not involved—took on
the task of co-editing the present book with great enthusiasm and competence. I am also
grateful to Anita Aigner and Peter Auer, who developed the tasks in coordination with me;
the tutors of the exercises, who, especially if they had an artistic background, suffered
from the rigid geometrical rules; and especially the students, who study in a world in which
ECTS points count more than the content and who generated passion and attention for this
exercise. Last but not least, content has to be shared. I am delighted that we have been able
to win over the artist and graphic designer Marie Reichel.
302
Imprint
Editors Werner Van Hoeydonck, Christian Kern, Eva Sommeregger
Acquisitions Editor David Marold, Birkhäuser Verlag, A-Vienna
Content & Production Editor Bettina R. Algieri, Birkhäuser Verlag, A-Vienna
Proofreading Ada St. Laurent
Layout and cover design Marie Reichel
Cover illustration Steffen Alexander Blickle
Image editing Mona Torsan, Augustin Fischer
Texts p. 183, 196, 206, 214–215, 222 based on the assignment sheets by Anita Aigner.
Texts p. 253, 254, 262, 268, 274 based on the assignment sheets by Peter G. Auer.
Texts p. 223–251, p. 275–298 by Werner Van Hoeydonck
Printing Holzhausen, die Buchmarke der Gerin Druck GmbH, A-Wolkersdorf
Paper Condat matt Périgord 135 g
Typeface Work Sans
Library of Congress Control Number 2021948985
Bibliographic information published by the German National Library

The German National Library lists this publication in the Deutsche Nationalbibliografie; detailed
bibliographic data are available on the Internet at http://dnb.dnb.de.
This work is subject to copyright. All rights are reserved, whether the whole or part of the
material is concerned, specifically the rights of translation, reprinting, re-use of illustrations,
recitation,broadcasting, reproduction on microfilms or in other ways, and storage in databases.
For any kind of use, permission of the copyright owner must be obtained.
ISBN 978-3-0356-2517-2
e-ISBN (PDF) 978-3-0356-2518-9
© 2022 Birkhäuser Verlag GmbH, Basel

P.O. Box 44, 4009 Basel, Switzerland
Part of Walter de Gruyter GmbH, Berlin/Boston
987654321 www.birkhauser.com
303
Many thanks for the financial support:
Three Dimensional Design

and Modelmaking
E264/2
Institute for Art and Design
University of Technology Vienna
Harry Schmidt
DESIGN&FUNCTION

Space Tessellations

Uploaded by

Copyright:

Available Formats

Space Tessellations

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Space Tessellations

Uploaded by

Copyright:

Available Formats

space tessellations

7 In Memoriam: William S. Huff (1927–2021)

Research Perspectives Teaching Perspectives

143 Pattern Manipulation through

163 Parakeet3D: Algorithmic Re-Envisioning

William S. Huff lecturing at the

8 In Memoriam: William S. Huff Dénes Nagy

10 In Memoriam: William S. Huff Dénes Nagy

This “signature”, where “Dénes” (five letters) flows into

Dénes Nagy is a Hungarian-Australian mathematician, historian of science, and theoretician

12 In Memoriam: William S. Huff Dénes Nagy

16 Introduction Werner Van Hoeydonck and Eva Sommeregger

18 Introduction Werner Van Hoeydonck and Eva Sommeregger

Biography of the Authors

Werner Van Hoeydonck

Space Tessellations Research Perspectives 23

24 Parquet Deformations: A Subtle, Intricate Art Form Douglas R. Hofstadter

Space Tessellations Research Perspectives 25

26 Parquet Deformations: A Subtle, Intricate Art Form Douglas R. Hofstadter

Fig. 4: “Consternation”, by Scott Grady.

­ ttempt an analysis of it. In its intermediate regions I

Space Tessellations Research Perspectives 27

Fig. 7: “Quirky Cogs”, by Arne Larson.

Fig. 8: “Trifoliolate”, by Glenn Paris.

28 Parquet Deformations: A Subtle, Intricate Art Form Douglas R. Hofstadter

Space Tessellations Research Perspectives 29

30 Parquet Deformations: A Subtle, Intricate Art Form Douglas R. Hofstadter

Fig. 12: “Beecombing Blossoms”, by

Space Tessellations Research Perspectives 31

32 Parquet Deformations: A Subtle, Intricate Art Form Douglas R. Hofstadter

Space Tessellations Research Perspectives 33

34 Parquet Deformations: A Subtle, Intricate Art Form Douglas R. Hofstadter

Space Tessellations Research Perspectives 35

36 Parquet Deformations: A Subtle, Intricate Art Form Douglas R. Hofstadter

Space Tessellations Research Perspectives 37

Fig. 1: “Circlewaves”, 2016,

Space Tessellations Research Perspectives 39

Fig. 2: Marcela Quijano, curator at the HfG archive

Space Tessellations Research Perspectives 41

Space Tessellations Research Perspectives 43

Space Tessellations Research Perspectives 45

Fig. 5: Moritz Kreitz, “One-handed Two-handed

Fig. 6: Parkettierungsaufgabe 1956/57,

Space Tessellations Research Perspectives 47

Fig. 7: “Parquetry Study in Three

Space Tessellations Research Perspectives 49

Fig. 9: Thomas C. Davies, “Leather of

Fig. 11. Peter Hotz, The Original

Space Tessellations Research Perspectives 51

Fig. 13: M.C. Escher’s Liberation, April

Space Tessellations Research Perspectives 53

Fig. 14: Huff’s analysis of the

Space Tessellations Research Perspectives 55

Space Tessellations Research Perspectives 57

Space Tessellations Research Perspectives 59

Space Tessellations Research Perspectives 61

Biography of the Author

Werner Van Hoeydonck received a master’s degree in Architecture in Ghent, Belgium in

Space Tessellations Research Perspectives 63

We must complain about the over-schematic and

ttempt an analysis of it. In its intermediate regions I

Fig. 5: Creating parabola by moving a right angle,

Rhombic dodecahedron—8 edges dual solid Tetrahedron+truncated