How We Talk(ed) About It: Ways of Speaking About Computational

Architecture

Abstract

If we understand architecture as a three-part system formed by the building, its

image, or drawings and images describing buildings, and the critical discourse
around architecture, then the texts or ways of speaking about computational
architecture play a key role in understanding the field and its development. By
analysing a corpus of around 4.6 million words from texts written between 2005
and 2020 that form a part of critical discourse in the field, this paper aims to map
ways of speaking about computational architecture. This contributes to
architectural theory and might help gain a better understanding of the evolution
of the field. Findings show that computational architecture is surrounded by a
specific way of speaking, hybridised with words from fields such as biology,
neuroscience, arts and humanities, and engineering. While some topics such as
‘sustainability’ or ‘biology’ come up consistently in the discourse, others, such as
‘people’ or ‘human’, have periods when they are more and less popular. The
paper tracks and documents trends and illuminates patterns and concludes by
presenting a map of periodic and recurring topics in ways of speaking about
computation in architecture over the last 15 years, discusses them within a larger
context and highlights open research questions.

Keywords: architectural design, computational architecture, digital construction, natural

language processing, digital architecture.

As computation is retooling most fields (1), over the past 30 years, the avant-
garde in architecture has been connected to the use of technology. The digitalisation that
architectural design has been going through has had an impact on the profession’s
conceptual agenda, design, and materialisation. Repurposing software tools built for
other industries and using them for architecture has created star practices such as Gehry
Partners with CATIA (2), and Zaha Hadid Architects with Maya (3). Moreover, some
architects from the younger generation have started to code as they discover that there is
a lack of tools for specific tasks, or that existing tools are inadequate or insufficient (4),
(5), (6), (7), (8). As new tools are developed, so are new ways of thinking, writing,
designing, and doing. The logic of tools feeds back into the mentality of the operators
(9), (10) as design is always affected by the choice of tools (11), (12), (13), (14).
All this has produced many terms referring to architectural projects that make
extensive use of digital technologies such as: ‘digital architecture’, ‘parametric
architecture’, ‘computational architecture’, ‘algorithmic’, ‘generative architecture’, and
‘advanced architecture’. These terms are explained and discussed extensively in
architectural texts.
Menges and Ahquist (15) define computational architecture as the explicit use of
scripting and/or programming in the design and/or the fabrication phase. According to
Leach (16), algorithmic architecture involves the use of programming languages and/or
paradigms. One definition for parametric architecture is that it implies working through
software interfaces that allow relational design: virtual objects contain interconnected
features and changing one feature will change the others automatically (16). In this case,
the designer produces objects as well as the relationships between objects. The debate
around what parametric, computational, algorithmic, and digital architecture mean (17)
is ongoing. As has been shown in (18), all these terms have been used ambiguously,
inconsistently, and interchangeably.

In this article, computational architecture is used as an umbrella term to discuss

architecture shaped by technological advancements, and this includes digital,
parametric, algorithmic, and parametric architecture. The term ‘computational’ is used
instead of digital or digitalisation because the focus is on early adopters of advanced
technologies for architectural design and not on how software applications designed for
architecture are being implemented across the discipline. Therefore, this study deals
with the early phases of the digitalisation of architecture.
Hensel identified a series of problems that computational architecture faces:
fragmented discourse, exhausted idiosyncrasy, redundant form-function dialectic, and
shallow ecological and sustainability approaches. However, most importantly,
contemporary discourse does not reflect on the larger context in which computational
architecture exists (19). Similarly, Cash (20) makes a compelling case on the poor state
of theory and meta-theory building in design research in general.

Forty (21) describes architecture as a three-part system formed by the building,

its image (drawings and photographic representations), and its accompanying critical
discourse. Modernist architecture was not only a new style of building, but also a new
way of talking about architecture, ‘instantly recognizable by a distinct vocabulary’ (21).
By studying how architects write, Medway explains how much of writing is done to
motivate action, stating that ‘architects finish a sentence with a sketch’ (22).
Furthermore, according to Damron (23), sketches are illuminated by sentences, as

1
writing is part of the doing. Language then becomes an integral part of architecture (24),
(21), (22), (23). Therefore, mapping and investigating the vocabulary of computational
architecture becomes important for understanding the practice in general, for building
theory and meta-theory for architecture, and for reflecting on the larger context in which
the field evolves.
This paper investigates how computation is changing architecture by studying
writings about architecture and is guided by the following research questions:
RQ1: What is the language of computational architecture?
RQ2: Does this language change over time and in what ways?
This study aims to add to the body of work that investigates the digital turns in
architecture (9), (25),. In order to answer these questions, a corpus linguistics
representative for computational architecture was built. This corpus contains texts
written over a 15-year period between 2005 and 2020 from two sources: the journal
Architectural Design and the eVolo skyscraper competition.
The rest of the paper is structured as follows: after related work is presented in
Section 1, the tools, methods, and research framework for investigating the research
questions are introduced in Section 2, and the findings are presented in Section 3.
Finally, in Section 4, the main topics found in the corpus and surrounding
computational architecture are discussed, and a conceptual map of the topics
surrounding computational architecture over time is presented.

1. Analysing Ways of Speaking in Architecture

The term ways of speaking is sometimes used to refer to the discourse a certain
community creates (26), (27), (28). The ways of speaking of an academic community
help to build discipline-specific knowledge and establish its cultural identity (29), (28).
In ‘Words and Buildings’, (21) argues that the ways of speaking of modernist architects
were integral in helping them frame their vision, while (30) goes so far as to say that
modernist architecture was ‘more basically, a body of documents defining modernism
and interpreting those buildings’. In ‘The Words Between the Spaces: Buildings and
Language’, (31) read the history of architecture through the development of
architectural texts discussing the role of language in producing buildings. In (23),
Markus argues that ‘the use of language should be investigated in design simply
because language is involved at every stage’.

2
 It is generally accepted that architecture has a specific vocabulary (32), (33), yet
little work has been done on analysing ways of speaking in computational architecture.
(18) study a corpus of texts trying to find unified definitions for parametric, generative,
and algorithmic design. (34) and (35) both collect and analyse corpuses of texts about
architecture in general and report findings related to the particularities of these texts:
architecture has a specific vocabulary impregnated by topics which come from
connected fields, the language is technical and often metaphorical, and new words are
created ‘with ease’.

2. Materials and Methods

To analyse the ways of speaking about computational architecture, a corpus

linguistics in English, built to be representative of the subfield, was created. This
corpus, ComPara, is different from previous work by (34) and (35) in two ways. First,
ComPara looks at a specific area of architecture, namely computational architecture.
Second, ComPara covers the period between 2005 and 2020. This section describes the
design, collection, and analysis methods for ComPara. After selecting relevant sources,
data was extracted and analysed quantitatively to extract the main topics and trends in
the text. The data were then assessed qualitatively. The analysis of ComPara represents
a corpus-based interpretative study.

2.1. Selection of relevant sources - criteria for corpus design

The first step in building a corpus is to select relevant sources. (34) uses three criteria in
designing her corpus - representativeness, accessibility and contemporariness.
Representative sources are those that are relevant in describing the professional
discourse of architecture. Accessible sources are those that are available for
professionals and those that can be found and placed in a digital database.
Contemporariness refers to up-to-date sources.
Two sources that fit the three criteria were selected: the journal Architectural
Design (AD), and winners and honourable mentions of the eVolo Skyscraper
Competition. AD and eVolo were chosen because they both specifically deal with
technological innovations in relation to architecture. Established in 1930, AD is widely
considered to be at the forefront of architectural thought. Over the last three decades,
AD has also featured many articles about technology and architecture to such an extent

3
that Mario Carpo states that ‘Not all things related to computational design have been
published in AD, but a large part of them have’ (25). On the other hand, eVolo’s About
section describes the journal as ‘focused on technological advances in architecture and
design’ (36). The eVolo Skyscraper Competition is arguably one of the most popular of
its kind worldwide, with around 1200 yearly submissions from over 150 countries (37).
This is why AD and eVolo are representative of computational architecture. The period
between 2005 and 2020 was chosen because of the accessibility of digital texts from
this time. For AD, only issues starting in 2005 are available digitally on the journal’s
page (38), while the first edition of the eVolo Skyscraper Competition was released in
2006.

2.2. Collecting the data and size of ComPara

The following inputs from the period between 01/2005 and 12/2020 were
collected from AD for use in the corpus: (a) all issue titles, (b) the titles of all 1795
articles from each issue, (c) the text of these articles, and (d) keywords associated with
the Introduction article. This forms a corpus of around 4.5 million words. The keywords
were collected from the Information section next to the article on the journal’s webpage.
It seems that they are generated automatically using a language processing algorithm,
but details of the algorithm are inaccessible to external users. Keywords represent
‘words which are statistically characteristic of a text’ (39).
Data from 2006 to 2020 were obtained from eVolo using the following bases:
(a) titles of all winning and honourable mentioned projects and (b) the descriptions
(abstracts) submitted by authors for these projects. This forms a corpus of around
100,000 words.
The current total size of ComPara is around 4.6 million words and forms a
special purpose medium-sized corpus (34).

2.3. Tools for processing ComPara

The study presented here is corpus-driven (40), meaning there were no pre-
assumptions or hypotheses before the analysis was conducted.
The analysis of ComPara was done in two stages. First, the corpus was analysed
quantitatively with the use of natural language processing (NLP) tools. Next, a
qualitative analysis of the results that emerged from the quantitative analysis was
carried out.

4
 An array of NLP tools exists, with each tool implementing different algorithms
derived from statistical techniques for topic modelling (41), (42), or (43). In this study,
two browser-based text analysis applications were used, namely Voyant Tools (44) and
Infranodus (45). These applications implemented well-known algorithms, such as the
Latent Dirichlet Association (42), (41), as well as proprietary algorithms.
Voyant Tools includes a large collection of tools. The ones used here are
Summary, Trends, Phrases and Cirrus. Using Cirrus, word clouds were created to
display words that were dimensioned based on their frequency in a text (46), (47).
Common connection words and punctuation are excluded. Word clouds are useful for
seeing key terms in a text and have been successfully used as tools for the preliminary
analysis of texts (48). However, in classical word clouds, all connections between words
are lost.
Infranodus is an NLP tool that transforms pieces of text into contextual word
clouds (49). Infranodus is based on a text network analysis algorithm, similar to the
Latent Dirichlet Association (but described as better), that represents any text as a
network and identifies the most influential words in a discourse based on terms' co-
occurrence. An algorithm is applied to identify different topical clusters, which
represent the main topics in the text as well as the relations between them (45). Thus,
contextual word clouds represent the most common words in a text, the connections
between the words, and topics, which are words that appear next to each other in text,
but not with the other words.

2.4. Data analysis

The data in ComPara contains titles, keywords, and prose text. These different
data types were analysed using two different approaches. Keywords are words without a
context, and titles are only short sentences. As a result, contextual word clouds would
either not be created or the connections between words would be too weak to produce
meaningful results. The titles and keywords were transformed into word clouds to
address this while the prose text was directly transformed into contextual word clouds.
Word clouds were created from titles and keywords from AD and titles from
eVolo. These clouds were then printed out. After a period of becoming familiar with the
data, the clouds were coded all together in initial subsets using an emerging coding
approach (50). Then, these emergent codes went through a period of analysis, where
Voyant tools’ Trends and Phrases were used to query the data for different keywords

5
that were traced back to their original contexts. This ensured that the meaning in context
was understood correctly and helped in the production of the final list of codes. The
generated list was used to code all word clouds. Afterwards, the codes were affinity
diagrammed (51) until a final theme structure was created.
The prose text from AD articles and eVolo project descriptions was transformed
into contextual word clouds and main topical groups, and the most influential elements
were generated automatically using Infranodus NLP.

3. Findings: Architectural Design

Issue titles, article titles, article text, and the Introduction keywords from entries
between 2005 and 2020 were retrieved from AD. The titles and keywords were
transformed into word clouds and categorised under six main themes that emerged after
affinity diagramming: (a) profession-specific terms, (b) places, (c) time periods and
currents in art and architectural history, (d) technology, (e) sustainability, and (f)
mathematics, physics, and biology. The texts in the articles were transformed into
contextual word clouds, and the topical clusters and most influential elements from each
issue are presented.

3.1. Data from AD issue titles (2005–2020)

Fig. 1 shows the word cloud made from the titles of the 96 AD issues. These are
discussed below under the six main themes.
Profession-related terms include: ‘architecture’, ‘design’, ‘space’, ‘cities’,
‘urbanism’, ‘urban’, ‘buildings’, ‘research’, ‘space’, ‘landscape’, ‘city’, ‘site/non-site’,
‘housing’, ‘megastructure’, ‘pavilion’, ‘territory’, ‘local’, and ‘hyperlocal’. ‘Rural’ and
words related to it such as ‘countryside’ or ‘pastoralism’ are a lot less frequent
compared to words related to ‘urban’.
References to places in AD’s issue titles include ‘Europe’ (2006), ‘India’ and
‘Italy’ (2007), ‘China’ (2008, 2018), ‘Turkey’ (2010), ‘Latin America’ (2011), ‘Iran’
(2012), ‘London’ (2012), ‘UAE’ and ‘the Gulf’ (2015), and ‘Brazil’ (2016).
Words related to time periods that stand out are ‘contemporary’, ‘21st century’,
‘2050’, ‘now’, ‘1970s’, ‘1960s’, ‘age’, and ‘third age’. ‘Future’ is a lot more frequent in
the titles than the term ‘past’. Words related to current affairs in art or architectural
history are ‘radical post-modernism’ (2011); ‘new structuralism’ (2010), which refers to
a symbiosis between design, engineering and architectural technologies; ‘parametricism

6
2.0’; ‘surrealism’ (2018); and ‘avant-garde’ (2019). By 2019, two new ‘posts’ had
replaced the post-modern of 2012, namely ‘post-digital’ and ‘post-Anthropocene’.

Fig. 1 – Word cloud of all words, scaled according to frequency, forming the 96 issue titles of
AD (01/2005 – 12/2020). 550 total words and 293 unique words.

Words relating to technology were abundant in AD’s issue titles, and they
include the following terms: ‘digital’, ‘computation’, ‘interactive’, ‘software’, ‘robots’,
‘open-source’, ‘machine’, ‘virtual’, ‘robots’, ‘algorithmic’, ‘programming’, and ‘3D
printed’.
Then, there are words related to sustainability such as ‘sustainable’, ‘ecology’,
‘food’, ‘ecological’, ‘sustaining’, ‘scarcity’, ‘green’, ‘ailing planet’, ‘depleting’,
‘resources’, ‘environment(s)’, ‘post-traumatic’, ‘ecoredux’, and ‘resilient’. It is
interesting to note that the word ‘sustainable’ is less frequent than either ‘digital’ or
‘computation’ in AD issue titles.
References to the field of mathematics include ‘mathematics’ and ‘4D space’.
References to physics include ‘morphogenetic’, ‘morpho-physical’, ‘vicissitude’, and
‘flows’, while references to biology include ‘protocell’ and ‘neo-plasmatic’.

7
3.2. Data from AD article titles (2005–2020)
The titles of the 1795 AD articles are made up of 12,929 words with 4,146 unique
words. The most frequent words in the titles are ‘architecture’ (199), ‘design’ (194),
‘new’ (97), ‘urban’ (90) and ‘city’ (86). The word clouds from the article titles are
relatively similar to the word cloud made from the issue titles, as each issue called for
articles fitting these themes. However, analysing the titles year by year reveals an
interesting progression, which is also visible in the analysis of the texts of these articles.
This will be discussed in succeeding subsections. Word clouds of article titles year by
year are available at (52).

3.3.Keywords associated with the AD Introduction (2005-2020)

There are 13,835 keywords with 5,961 unique words associated with the Introduction of
each of the 96 AD issues, and the most frequent keywords are ‘architecture’ (166),
‘university’ (85), ‘architects’ (75), ‘design’ (70) and ‘new’ (63). Fig. 2 illustrates the
500 most frequent keywords scaled according to their frequency and loosely grouped in
the six thematic clusters.
Profession-related terms such as ‘architecture’, ‘design’, ‘house’, and ‘building’
lie at the centre. The most mentioned architectural functions are ‘museum’, ‘pavilion’
and ‘residential project’. Other common functions are ‘hotel’, ‘campus’, ‘office’,
‘airport’, ‘station’, ‘hospital’, ‘library’, ‘arena’, ‘hall’, and ‘square’.
Names of places sit at the lower right corner placed on a map that paints a
polarised picture. The USA, Europe, China, Japan, and Australia are relatively well
represented, while Latin America is only represented due to mentions of Mexico and
Columbia. No African country, city, or place make the top 500 keywords. Zooming in
and looking into the keywords year by year, we see that Africa appears twice in the
keywords, once in 2015 and once in 2017. In comparison, China is mentioned 21 times,
while London has 50 mentions, some in every year between 2006 and 2020. Europe is
only represented by a few places or institutions. London and the Bartlett dominate the
representation, with mentions of the Architectural Association (AA) and its Design
Research Laboratory (DRL), the RIBA, and the Serpentine. Next comes the Venice
Biennale, followed by Paris, the Pompidou, and France. Then there are some mentions
of German places and institutions: Berlin and Stuttgart with the Institute for
Computational Design (ICD). Finally, Switzerland and Zurich, Vienna, the Netherlands
and Barcelona-Spain are the least frequently mentioned places. Eastern Europe,

8
Northern Europe, and Russia are not mentioned at all. Istanbul is mentioned a few
times, while the Middle East is only represented through Abu Dhabi, Beirut and the
Gulf. Mumbai is mentioned, although only a few times, followed by Singapore and
Hong Kong, as well as China with Beijing, Shanghai, and Shenzhen. The rest of Asia is
only mentioned through Japan. Even the word ‘west’ is more frequently mentioned and
is consequently larger on the representation than the word ‘east’.

9
Fig. 2 – Word cloud showing 500 most used keywords associated with the Introduction in the 96 issues of
AD (01/2005 – 12/2020) scaled according to frequency and grouped based on thematic clusters. There are
13,835 keywords with 5,961 unique words.

10
 Words related to periods in art and architecture are ‘modernism’ and ‘post-
modernism’, ‘future(s)’, ‘history’, ‘contemporary’, and ‘functionalism’. The term
‘industrial revolution’ is found in the keywords in 2009, 2015, and 2017–2020. The
word ‘gothic’ is mentioned three times in the keywords (2013, 2016, 2018); in contrast,
‘baroque’ is only mentioned once, in 2011. The word ‘new’ is a lot more prevalent than
the word ‘old’ throughout the years.
Color-coding the names of people (upper right corner in Fig. 2) shows a male-
dominated scene apart from some notable exceptions such as Zaha Hadid, Neri Oxman,
and Jane Burry. Names of practices such as OMA or Arup are double coloured while
‘BIG’ and ‘Happold’ from ‘Buro Happold’ are left blue because the names of the
practices are of male architects. Both place and name analysis pictures look a lot more
diverse when zooming in to the keywords year by year. Looking at the names
mentioned in AD’s Introduction keywords year by year, philosophers include
materialists such as Deleuze, Deleuze-Guattari, and DeLanda. ‘Deleuze’ is a keyword in
2006, 2009, 2012, and 2014, ‘Guattari’ in 2009, 2012, 2014, while ‘body without
organs’ is mentioned in 2008. ‘DeLanda’ is a keyword in 2009, 2012, 2015 and 2016.
These three are the most popular philosophers whose names are included in the corpus.
Next come the words ‘deconstructivist’ (2007, 2009, 2014) and ‘Derrida’, which were
mentioned in 2009. Third, and more recently, Tim Morton was mentioned in 2012, then
Harman (2016, 2019) and Heidegger (2019). Other philosophers include Kant (2014,
2019), Foucault (2006, 2008, 2012), Lefebvre (2009, 2012, 2013), and Merleau-Ponty
(2012, 2019). Edward Soja (2011–2012), Roland Barthes (2009, 2016), Žižek in 2010,
McLuhan (2006, 2012) and Latour (2006, 2014) are also mentioned. Scientists
mentioned include Wolfram (2016), Freud (2008, 2016), and Darwin (2009, 2012,
2019). John Ruskin (2009, 2019), Heinrich Wöflin (2016), Arthur Danto (2009), and
Duchamp (2009, 2013, 2019) are some of the included art historians. Finally, among
architects, Rem Koolhaas and Le Corbusier are the most popular. They were part of
keywords in 11 out of the 16 years. Gregg Lynn is mentioned six times (2006, 2007,
2009, 2014, 2015, 2020), Bucky Fuller is mentioned six times (2006, 2008, 2010, 2011,
2012, 2015) as well, and Frei Otto is mentioned seven times (2006–2010, 2015–2016).
The upper left corner of Fig. 2 has grouped together terms which have to do with
technology such as ‘digital’, ‘technology’, ‘computational’, ‘media’, ‘network’, and
‘internet’. Software families (‘BIM’), programming languages (‘grasshopper’), and
manufacturing technologies (‘CNC’ and ‘robotic fabrication’) were mentioned as well.

11
The interest in engineering is also seen in the frequency of ‘Arup’ as a keyword
between 2010 and 2012. This keyword comes back in the periods of 2014–2015 and
2017–2018, albeit less frequently.
Words related to sustainability placed around the top centre of Fig. 2 include
‘ecologies’, ‘green’, ‘environmental’, and ‘homeostasis’.
Finally, there are words related to mathematics, physics, and biology. ‘Geometry’
appeared six times in 2010 and 2011. There are few words that can be connected to
physics, and they include ‘air’, ‘energy’, ‘sky’, and ‘physics’. Words which can be
connected to biology include ‘bio’, ‘biological’, ‘growth’, ‘natural’, ‘organic’, and
‘life’.

3.5. Data from the text of AD articles (2005-2020)

The texts forming the 1795 AD articles are made up of 4,544,090 words and
92,963 unique words. The most frequent words are ‘design’ (19,892), ‘architecture’
(16,915), ‘new’ (16,701), ‘building’ (10,528) and ‘city’ (9,668). Fig. 3–7 shows the
main topical clusters and most influential elements in each AD issue. These were
retrieved from the contextual word clouds generated using the Infranodus NLP tool.
Most of the topics shown are profession-specific words, such as ‘architecture’,
‘design’, ‘building’ or ‘city’. The words ‘form’ and ‘system’ (marked in italics in Fig.
3–7) often come up either in the main topical groups or as the most influential elements.
References to places make up the main topical groups sometimes, and they corelate to
the AD issue titles, article titles, and Introduction keywords. Sometimes the words
‘local’ and ‘hyperlocal’ are significant topics. ‘Time’ comes up eight times in topical
clusters, sometimes appearing close to the word ‘space’ (see Fig. 4: 2008-4, Fig. 5:
2013-5). ‘Future’ comes up three times.
Words that have to do with technology are coloured purple, and they are
relatively evenly distributed throughout the years and include ‘virtual’, ‘software’,
‘parametric’, ‘robotic’, and ‘BIM’. ‘Artificial intelligence’ comes up 145 times in the
texts, and most mentions are in 2019–2020.
Words that could be connected to sustainability are coloured blue. These
include ‘environment’, ‘scarcity’, ‘resource’, and ‘sustainability’.
Words that can be associated with biology are coloured in dark blue and include
‘protocell’, ‘biomimicry’, and ‘DNA’. In 2019, some references to neuroscience were
made (see Fig. 7: 2019-5). Topics about mathematics include ‘geometry’ and

12
‘mathematic’, and these were concentrated in 2011. Words that can relate to physics
include references to outer space exploration, such as ‘Moon’ and ‘Mars’, but also
‘flow’ and ‘energy’.

Fig. 3

13
Fig. 4

14
Fig. 5

15
Fig. 6

16
Fig. 7

17
Fig. 3, 4, 5, 6, 7, 8 – texts making up the 1795 AD articles between 2005-2020. Main topical groups and
most influential elements as analysed using the Infranodus NLP. The topics are presented for each AD
issue, year by year. The texts total 4,544,090 words and 92,963 unique words.

Lastly, words which can be associated with human are coloured in orange and they
include ‘social’, ‘community’, ‘human’, and ‘people’. It is interesting that not a single
word that could be connected to humans was part of the most used topics between 2009
and 2013. However, they have been frequently used in the last three to four years (see
2016 through 2020 in Fig. 5-7). The word ‘human’ itself appears a total of 4006 times
in the texts of the AD articles, but it is used significantly more often in 2014, 2019, and
2020.

4. Findings: eVolo Skyscraper Competition (2006–2020)

There are 42 winning projects and 307 honourable mentions in the eVolo Skyscraper
competition between 2006 and 2020, which in total form 349 projects. Some of these
projects have been described and categorised in detail over the years in (53), (54) and
(37). Below, the topics forming the titles of eVolo projects are presented under the same
main themes used for AD. The abstracts describing the projects were transformed into
contextual word clouds. The main topical clusters and the most influential elements
from these contextual word clouds are presented year by year.

18
4.1. Data from the titles of eVolo winning projects and honourable mentions
(2006–2020)
Fig. 9 shows the words forming all titles of the winning projects and honourable
mentions of the eVolo skyscraper competition between 2006 and 2020.
Terms connected to architecture, and more specifically to high rises, such as
‘vertical’, ‘tower’, ‘skyscraper’, ‘city’, ‘urban’, and ‘structure’ stand out at first glance.
Architectural functions that have high frequencies are ‘airport’, ‘bridge’, and strangely,
‘pyramid’.

Fig. 9 – Word cloud of all words, dimensioned according to frequency, in titles of winning
projects and honourable mentions for the eVolo skyscraper competition (2006–2020). 1,483
total words, 753 unique words.

The names of places of high density, such as New York, Hong Kong, Paris,
London, India, and Shanghai, have high frequencies. But ‘Babel’ is used as often as

19
these real places and appears as part of titles six times in total (twice in 2012 and 2014,
and once in both 2016 and 2017). Babel is connected to a skyscraper under perpetual
construction (The New Tower of Babel (55)), a home built at almost any height with the
help of aerostatic construction (House of Babel (56)), an ecological structure designed
as a scientific facility and tourist attraction for the desert (Sand Babel (57)), a massive
collage of cultural symbols (Taiwan Babel Tower (58)), and a memorial for workers in
the building industry (The Scaffold of Babel (59)). A series of projects look at outer
space as a place to build human habitats. The word ‘Mars’ appears relatively frequently
in the titles: twice in 2013 and once in 2017. ‘Moon’ is also part of titles with the
Moonscraper in 2011 (60), while ‘stratosphere’ is mentioned in 2013 (61). Generally,
these projects describe concepts of terraforming that would save humanity in the face of
overpopulation, depleting resources, and the negative effects of climate change.
On the other hand, maps of geographies that produced successful eVolo
submissions can be found in (52) but also under (53), (54), and (37). For the winning
submissions, the 42 projects came from 16 countries. When looking at the countries of
both winning projects and honourable mentions, 48 countries are represented, but the
distribution is uneven. The United States is clearly dominating (88 projects), with China
(51 project) second, the United Kingdom (41 projects) third, France (26 projects) fourth,
South Korea fifth (19 projects), and Poland (11 projects) and Russia (11 projects) sixth.
Africa is almost off the map, with only two honourable mentions from Egypt, while
South America is only represented by Chile (3 projects), Peru (one project), and
Venezuela (one project).
The most referenced period in eVolo’s titles is the 21st century. The term
“future” also appeared frequently and was used to refer to the year 2016 (for an entry in
2010) and to more distant ones such as 2100 or 3015. The only reference to the past that
comes up in eVolo titles is ‘the 70s’.
When it comes to names, the prevalence of ‘Babel’ is complemented by other
Christian religious references such as ‘Noah’ and ‘Moses’ (in the context of depicting
apocalyptic scenarios). An honourable mention from 2011 called Rhizome Tower: A
Thousand Underground Plateaus (62) makes the influence of both Deleuze and Guattari
explicitly present in the titles of eVolo projects.
References to technology include words such as ‘machine’, ‘algorithmic’ and
‘parametric’, ‘3d printed’, ‘drone’, and ‘data’.

20
 While direct technology references are not as common in the eVolo titles as they
are in AD, there are more words that can be connected with sustainability in eVolo. For
example, the terms ‘ecology’, ‘climate’, ‘sustainable’, ‘living’, ‘earth’, ‘clean’,
‘pollution’, and ‘recycling’ frequently appear in eVolo titles.
There are no direct references to mathematics in the project titles. However,
references to biology are ample and include ‘geno-tower’, ‘bioclimatic’, ‘peristal
living’, ‘cell’, ‘geno-matrix’, ‘bio-city’, ‘bionomic’, ‘bio-pyramid’, ‘bio-habitat’, and
‘biomorph’. Physics is also referenced, although less often than biology. For example,
the word ‘quantum’ is part of titles with ‘Quantum City’ in 2007 (63) and ‘Quantum
Skyscraper’ in 2013 (64).

4.5. Data from eVolo abstracts of winning projects and honourable mentions
(2006–2020)
The abstracts of winning projects and honourable mentions in the eVolo skyscraper
competition between 2006 and 2020 have 96,016 words and 9,988 unique words. The
most frequently used words are ‘city’ (482 mentions), ‘building’ (371 mentions), ‘new’
(360 mentions), ‘water’ (322 mentions) and ‘structure’ (298 mentions). Fig. 10 presents
the most influential topics and elements in the eVolo abstracts year by year between
2006 and 2020.
Among the most common topics in the abstracts are profession-specific words,
such as ‘tower’, ‘building’, ‘space’, and ‘structure’.
There are no names or periods that come up in the main topical groups or most
influential elements, and the only topic that can be connected to technology is ‘drone’
(Fig. 10: 2016).
It is interesting to note that the word ‘water’ is among the words that are part of
both the most influential topics and the most influential elements in the abstracts, and
this has a rather uniform distribution throughout the years (see 2008–2010, 2012–2014,
2018–2020 in Fig. 10). ‘Water’ is used in connection to sustainability and framed as a
problem that needs to be addressed through architectural projects for a sustainable
future. Less frequently, ‘carbon’ and ‘air’ appear among the most influential words in
the abstracts (Fig. 10: 2014, 2020).
Words that can be connected to ‘human’ (in orange) have a growth in frequency
from 2008 onwards. This can be seen both by looking at the relative frequency of
‘human’ in the abstracts, but also by looking at the most influential topics in the

21
abstracts (see Fig. 10: 2018–2020). In the years 2006 through 2015, the most influential
topics in the abstracts were ‘structure’, ‘building’, ‘skyscraper’, ‘space’ or ‘project’.
‘Structure’ is very often among the most influential words in the abstracts (see 2006 to
2008, 2013, 2015 in Fig. 10). From 2016 onwards, ‘people’ is used more often (see
2016–2018, 2020 in Fig. 10), although a trend towards this was already indicated when
‘resident’ was included in the most influential topics in 2012. This is a similar trend to
what we saw in AD article texts, and it shows that the topics surrounding computational
architecture change over time and that there is a transition of interests from building (as
a noun and as a verb) to the act of habitation, people, and humans.

22
Fig. 10 – eVolo abstracts of winning projects and honourable mentions 2006-2020. Main topical groups
and most influential elements as analysed using the Infranodus NLP. There are 96,016 total words and
9,988 unique words. The topics are presented year by year.

23
5. Discussion

Computational architecture makes use of a specific vocabulary that allows for the
refinement of ideas and the cultivation of culture around the field. This section
discusses the topics that consistently appear in ways of talking about computational
architecture and the topics that come in a periodic fashion.

5.1. Recurring topics in ways of speaking about computational architecture

Computational architecture is more interested in the future rather than the past, in the
new rather than the historical, and in the urban rather than the rural—all this broadly
follows the field of technology. In architectural theory, the ‘rejection of history’ has
been well debated, at least since modernism. The discourse is dominated by the West,
although projects from China have won mentions in eVolo in recent years (see Section
4.2.). The avant-garde of the 70s is mentioned across the corpus, and there are
references to the Moon, Mars, and space exploration.
In general, the discourse is developed and hybridised with topics coming from
the natural sciences, specifically biology and physics. Topics from mathematics are also
prevalent in, but the contribution of explicit mathematical topics was concentrated in the
period between 2010 and 2012. Importantly, the topic of sustainability comes up often
and consistently.
Strangely, eVolo contains words related to biblical characters such as Noah and
Moses, and biblical places such as Babel—which comes up in titles with a surprising
frequency. Almost every year, a number of eVolo Skyscraper Competition winners or
honourable mentions have the word ‘Babel’ in their titles. It might be interesting to
investigate the origin of projects which make these biblical references.
The words ‘form’, ‘space’, and ‘system’ often appear as main topics in
ComPara. While ‘space’ and ‘form’ are traditional concerns in architectural theory, the
word ‘system’ might be newer in architectural discourse, and understanding how and
where it is used could uncover interesting patterns.

5.1.1. Sustainability
In general, sustainability is described as a problem to which architecture (many
times enhanced by technology) is seen as a solution. Investigating how sustainability is
understood in the field of computational architecture over time is a possible direction

24
for future research. The following are potential research questions that can be
investigated: What does it mean to be sustainable? Can sustainability be achieved? How
will we know when we have achieved it? Can sustainability be described without
reaching tensions about diverging interests?
Sustainability comes up as a topic more often in the eVolo corpus, where most
projects state problems related to the environment and climate change that the project
can solve. Generally, the projects start with stating a problem that is dramatic and large,
and continue with suggesting highly technological, built (conceptual) solutions that can
solve the problem. The problems mostly deal with high population density and its
associated issues of over-population and pollution of the sea, earth, and sky. Stressed
infrastructures, desertification, the depletion of natural resources, potential nuclear
disasters, or the melting of polar caps are frequently mentioned. This results in a series
of words hinting at rather pessimistic realities and futures such as ‘cemetery’, ‘landfill’,
‘Chernobyl’, ‘garbage’, ‘plastic waste’, and ‘pollution’ (see Fig. 10). However, these
futures are saved by the solutions suggested through the projects. But starting in 2015,
the word ‘problem’ becomes more frequent than the word ‘solution’ in the abstracts.
This might show a transition towards a different understanding of sustainability as a
more complex or wicked problem (65). To exemplify the problem-solution dynamic,
Noah’s Ark: Sustainable City (an honourable mention from 2012) is a floating city that
could support all living species once they have been evicted from land ‘by natural
disasters, warfare, whatever disasters the end days may bring’ (66). Oceanscraper (67),
is a large underwater architectural structure that ‘does not have to abide by the laws of
gravity’ and would use decommissioned Russian submarines lying on the sea bed as
nuclear power sources. Moses: A Decentralized Floating Network of Skyscraper Cities
(68) and The Promised Land Waterscraper (69), are solutions to rising sea levels. The
metaphor of the ark appears a couple of times in the projects as a solution to apocalyptic
futures. All of these conceptual projects clearly articulate a real-world problem, usually
related to sustainability, and then continue to offer solutions to that problem. The
solution does not have to be feasible, or even realistic, in any way, but the problem
needs to be real and of monumental proportion. This corresponds to design’s inbuilt
optimism in general (70), and also follows hopeful views that technology will solve
most problems. While investigating sustainability understandings in the winning
projects and honourable mentions of the eVolo skyscraper competition would be a
research paper on its own, it can be said that sustainability is understood as a limitation

25
in these conceptual projects, and as a problem, or something to resolve. It is important
to note here that the calls of the competition frame the projects responses and that the
calls change slightly year by year, although the core focus remains on high rise
architecture, technology and sustainability.
AD and eVolo reference sustainability and technology differently: while AD is
filled with topics that have to do with technology and with fewer references to
sustainability, the opposite is true for eVolo, where sustainability is a recurring topic
throughout the years, and technology is less often referenced directly.

5.1.2. Biology
Throughout the years, biology and topics which can be associated to it frequently
appear in both AD and eVolo. This simply puts quantitative data behind Phillip
Steadman’s (71) statement that ‘as a matter of historical fact, biology, of all sciences,
has been that to which architectural and design theory have most frequently turned to.’
Recently, other studies have looked at the relationship between biology and
architecture, and similar points were made by (72), (71), (73), (74), (75). Biology comes
up as a topic strongly connected to computational architecture. Tracing the depth and
scope of biology’s influence as a model, as a metaphor, as an analogy, as a source for
novel building materials, and as a field to entangle to computational architecture into a
new paradigm, as suggested by (76) can be subject for fruitful future research. Looking
specifically at the relationship between sustainability, biology, and computation in
contemporary architecture can also make for interesting investigations. Technological
advancements help to integrate biology and architecture and revisit the idea of growing
living buildings.
Based on the topics that come up constantly surrounding computational
architecture, namely technology, sustainability, and biology (and to a lesser extent
mathematics and physics), it can be argued that the field is currently shaped according
to the following model:
(Mathematics + Physics + Biology) * Technology / Sustainability
Technology helps to explore and enhance old (but yet unexplored) or new ideas from
mathematics (as argued for example in (77)), biology, physics (as discussed for example
in (78)) in architecture, while sustainability comes as a constraint or limitation,
sometimes to avoid creating purely technological explorations.

26
5.2. Waves of influence
While the topics described above appear with a rather even distribution, there are topics
that are more popular in certain periods. In (79) Heinrich Wölfflin read the history of art
in waves, explaining that art takes turns between being fascinated with the static aspect
of life (the classical) to focusing on the dynamic aspect of life (the baroque) and returns
in an upward spiral. Looking at the topics that come up in ways of speaking about
computational architecture, similar waves of influence might be visible. Fig. 11 shows a
map of topics that come up when speaking about computational architecture. On the
upper part are the topics that come up in waves, and at the bottom are topics that come
up constantly in the corpus.

Fig. 11 – Topics that come up consistently and in waves in ways of speaking about computational
architecture

5.2.1. Computational architecture between engineering and art

In the entire corpus, art comes up as a topic more often than engineering in general.
However, there is a period roughly between 2007 and 2012 when there are more
references to engineering and words connected to it (such as ARUP). This also
corresponds to the two traditions of architecture-as-technology and architecture-as-art
established after the Enlightenment (31). In the words of Nigel Cross, designerly ways
of knowing do not fall neither in the humanities nor within the sciences (80).

27
5.2.2. The Deleuze connection might be fading
ComPara shows numerous references to philosophers Deleuze, Deleuze-Guattari and
DeLanda, who was the philosopher whose declared role was to explain Deleuze to
architects (81), and who has done so by teaching in many of the avant-garde
architectural programmes around the world. All three names appear in the keywords
associated with the Introduction article in AD. A title of one eVolo project from 2011:
‘Rhizome Tower: A Thousand Underground Plateaus’ (82), makes a direct reference to
Deleuze and Guattari. These correspond to the so-called Deleuze connection to
architecture (83), (84), (85), (86), (87). Since 2016, neither ‘Deleuze’ nor ‘DeLanda’
have appeared in the AD Introduction keywords. ‘Deleuze’ still appears in the texts of
AD articles, but with less frequency. On the other hand, object-oriented ontology
(OOO) has been gaining popularity: ‘Harman’ and ‘Morton’, together with ‘Merleau-
Ponty’ and ‘Heidegger’, were part the keywords six times since 2012. The word
‘perception’ is also much more frequent in 2020 than in previous years in AD article
texts.

5.2.3. From object to subject

Perhaps the most interesting trend in the corpus is a transition in interests from object to
subject. The frequency of the word ‘perception’ might be connected to the progression
of trends throughout the years. In the period between 2008 and 2012, the main topics
revolved around structures, engineering, and building (both as a noun and as a verb). In
2012, both AD and eVolo began including topics that involved humans. In 2012, the
word ‘human’ appeared for the first time in the issue titles of AD, while the word
‘resident’ is an influential topic in the eVolo abstracts. Since then, topics related to
humans and people have been used more frequently (see Fig. 6, 7, 8, 10) and they are
discussed in the following ways:
1. Architecture in relationship to the social (‘people’, ‘community’, social’)
The social is a traditional concern in architectural theory in general. The literature on
design for sustainability shows that design is moving from product-level approaches to
a social-technical system focus (88). Reappraising the social might also be connected to
the frequency of the word system over the years in the corpus. For computational
architecture, this has interesting implications, as computational architecture has long
been dominated by: interest in the objects that can be created by means of computation,

28
the design processes that computation can facilitate, and the development of novel
materials and new tools, rather than the social.
2. Architecture in relation to perception (‘human’)
Perception has also been a topic connected to architecture and its theory (89), although
less frequently or directly than the social. Again, this trend echoes what is happening in
other design fields. For example, in interaction design, rooted in Dewey’s Art as
Experience (90), the interest has similarly moved from investigating objects to focusing
on and studying experiences (91)
3. Human creativity and artificial intelligence (‘neuro’, ‘brain’, ‘AI’, ‘machine
learning’).
Here, the discussions run between the future role of the architect, digital authorship, and
toolmaking. Some question whether AI will render the role of the architect obsolete
(92), while others state that it will simply become a prosthesis, helping architecture
evolve and allowing architects to generate more and better solutions (88). In this way,
AI would simply be a continuation of CAAD tools. Recently, much work has been
dedicated to using computation to partly automate the generation of architectural
solutions (94), (95), (96) while others have tried to articulate the relationship between
neuroscience, artificial intelligence, and architecture (97), (98).

In the last few years, the ways of speaking about computational architecture have shown
more topics that have to do with subjects rather than objects. It can be said that
computational architecture is surrounded by a new subjectivity which has at its core
‘people’, those for whom architecture is and how they perceive space, but also the
future role and relevance of the architect herself.

29
References

1. David B, Dieter M. Postdigital Aesthetics. London: Palgrave Macmillan UK;

2015. 1–3.

2. Dassault Systems. Catia [Internet]. 2020 [cited 2020 Nov 2]. Available from:

https://www.3ds.com/products-services/catia/

3. Autodesk. Maya [Internet]. 2020 [cited 2020 Nov 2]. Available from:

https://www.autodesk.com/products/maya/overview?support=ADVANCED&plc

=MAYA&term=1-YEAR&quantity=1

4. Wortmann T, Tunçer B. Differentiating parametric design: Digital workflows in

contemporary architecture and construction. Design Studies. 2017 Sep;52.

5. Castelo-Branco R, Brás C, Leitão AM. Inside the Matrix: Immersive Live

Coding for Architectural Design. International Journal of Architectural

Computing. 2021 Jun 28;19(2).

6. Horvath A-S. Assessing Site-Geometry for Architectural Design Using Graph

Theory. In: Chiorean C, editor. Proceedings of the Second International

Conference for PhD students in Civil Engineering and Architecture: Building the

Community of Young Researchers. Cluj-Napoca: U.T. Press; 2014. p. 611–9.

7. Horvath A-S, Becus R. Cluj Geoweb. In: Proceedings of the 1st International

Edition of Cadet Inova for Young Inventors [Internet]. 2015. Available from:

https://vbn.aau.dk/en/publications/6a9cae48-e009-4341-9741-e27283236c02

8. Bjørn P, Wulff M, Petræus MS, Møller NH. Immersive Cooperative Work

Environments (CWE): Designing Human-Building Interaction in Virtual Reality.

Computer Supported Cooperative Work (CSCW). 2021 Jun 16;30(3):351–91.

9. Carpo M. The Second Digital Turn Design Beyond Intelligence. MIT Press;

2017.

30
10. Horvath A-S, Rühse V. The Chladni Wall [Internet]. Vol. 1294, Communications

in Computer and Information Science. 2020 [cited 2021 Oct 18]. Available from:

https://link.springer.com/chapter/10.1007/978-3-030-60703-6_50

11. Dahlstedt P. Between Material and Ideas: A Process-Based Spatial Model of

Artistic Creativity. In: McCormack Jon and d’Inverno M, editor. Computers and

Creativity [Internet]. Berlin, Heidelberg: Springer Berlin Heidelberg; 2012. p.

205–33. Available from: https://doi.org/10.1007/978-3-642-31727-9_8

12. Vite C, Horvath A-S, Neff G, Møller NLH. Bringing Human-Centredness to

Technologies for Buildings. In: CHItaly 2021: 14th Biannual Conference of the

Italian SIGCHI Chapter. New York, NY, USA: ACM; 2021.

13. Horvath A-S, Vite C, Holten Møller NL, Neff G. messyBIM: Augmenting a

building information model with messy talk to improve a buildings’ design

process. In: Menéndez-Blanco M, Uğur Yavuz S, Schubert J, Fogli D, Paternò F,

editors. Title of host publication CHItaly 2021 Joint Proceedings of Interactive

Experiences and Doctoral Consortium. CEUR Workshop Proceedings; 2021. p.

7–14.

14. Horvath A-S, Rühse V, Raptis D. SoundSculpt: A Design Framework for 3D

Modelling and Digitally Fabricating Sound Patterns. In: ACM International

Conference Proceeding Series. 2020.

15. Menges A, Ahlquist S. Computational Design Thinking. London: Wiley; 2011.

11–13.

16. Leach N. Digital Morphogenesis. Architectural Design [Internet]. 2009

Jan;79(1):32–7. Available from: http://doi.wiley.com/10.1002/ad.806

17. Schumacher P, Gage MF. Architects Patrik Schumacher and Mark Foster Gage

face off [Internet]. U.S.A.: Youtube: Texas A&M College of Architecture; 2017

31
 [cited 2020 Nov 2]. Available from:

https://www.youtube.com/watch?v=i1LHqssdGE8

18. Caetano I, Santos L, Leitão A. Computational design in architecture: Defining

parametric, generative, and algorithmic design. Frontiers of Architectural

Research. 2020 Jun;9(2).

19. Michale Hensel. IoA Silver Lecture Michael Hensel: Confronting the Current

Crisis of Architecture [Internet]. Austria: Youtube: IoA, Die Angewandte; 2013

[cited 2020 Nov 2]. Available from:

https://www.youtube.com/watch?v=g9G2WEcFKRE

20. Cash PJ. Developing theory-driven design research. Design Studies. 2018

May;56.

21. Forty Adrian. Words and buildings : a vocabulary of modern architecture .

Words and buildings : a vocabulary of modern architecture /. London: Thames &

Hudson; 2000.

22. Medway P. Virtual and Material Buildings. Written Communication. 1996 Oct

6;13(4).

23. Damron R, Spector T. How Architects Write. New York: Routledge; 2013.

24. Thomas A. Markus. Language Structure and Building Types. Nordic Journal of

Architectural Research. 1992;5(4):35–48.

25. Carpo M. The Digital Turn in Architecture 1992-2012. London: Wiley; 2012. 8–

10.

26. Becher T. Academic tribes and territories : intellectual enquiry and the cultures of

disciplines. Reprint. Milton Keynes : Society for Research into Higher Education

Open University Press; 1993.

32
27. Flowerdew J. Discourse in English Language Education. London: Routledge;

2013.

28. Ghassan A, Blythe M, Yee J. Designerly Ways of Speaking: Investigating How

the Design Tribe of Researchers Speak on Design Thinking [Internet]. PQDT -

Global. [Ann Arbor]: University of Northumbria at Newcastle (United

Kingdom); 2019. Available from: https://www.proquest.com/dissertations-

theses/designerly-ways-speaking-investigating-how-

design/docview/2425238336/se-2?accountid=8144

29. Trowler P. Academic Tribes and Territories: the theoretical trajectory.

Österreichische Zeitschrift für Geschichtswissenschaften : ÖZG. 2014;25(3).

30. Bearn G. The Formal Syntax of Modernism: Carnap and Le Corbusier. British

Journal of Aesthetics. 1992;32(3):227–41.

31. Cameron D, Markus TA. The Words Between the Spaces. Routledge; 2003.

32. Gausa M. The Metapolis Dictionary of Advanced Architecture: City, Technology

and Society in the Information Age. Barcelona: Actar; 2003.

33. Curl JS. The Oxford Dictionary of Architecture. Oxford: Oxford University

Press; 2015.

34. Beloso BS. A Lexical description of English for Architecture: A Corpus-based

Approach. Santiago de Compostela: Peter Lang; 2015.

35. Cabrera T. Interpreting Architecture: The ARCHINT Corpus. Tradumàtica:

traducció i tecnologies de la informació i la comunicació. 2016;(14):156–71.

36. Evolo. Evolo [Internet]. Evolo Magazine. 2020 [cited 2020 Nov 2]. Available

from: http://www.evolo.us/about/

37. Aielo C. eVolo Skyscrapers 3: Visionary Architecture and Urban Design. New

York: Evolo Press; 2018.

33
38. Neil Spiller (ed.). Architectural Design [Internet]. 2020 [cited 2020 Nov 4].

Available from: https://onlinelibrary.wiley.com/journal/15542769

39. Culpeper J, Demmen J. Keywords. In: Biber D, Reppen R, editors. The

Cambridge Handbook of English Corpus Linguistics. Cambridge: Cambridge

University Press;

40. Mackiewicz J, Thompson I. Adding Quantitative Corpus-Driven Analysis to

Qualitative Discourse Analysis: Determining the Aboutness of Writing Center

Talk. The Writing Center Journal [Internet]. 2016;35(3):187–225. Available

from: http://www.jstor.org/stable/43965694

41. Landauer TK, Foltz PW, Laham D. An introduction to latent semantic analysis.

Discourse Processes. 1998 Jan;25(2–3).

42. Hofmann T. Probabilistic latent semantic analysis. In: Proceeding of Uncertainty

in Artificial Intelligence. Stockholm; 1999. p. 289–96.

43. Blei D, Ng AY, Jordan MI. Latent Dirichlet Allocation. Journal of Machine

Learning Research. 2003;993–1022.

44. Sinclair S, Rockwell G. Voyant Tools [Internet]. 2016 [cited 2020 Nov 2].

Available from: http://voyant-tools.org/

45. Paranyushkin D. InfraNodus: Generating Insight Using Text Network Analysis.

In: The World Wide Web Conference. New York, NY, USA: ACM; 2019.

46. Atenstaedt R. Word cloud analysis of the BJGP : 5 years on. British Journal of

General Practice. 2017 May;67(658).

47. Heimerl F, Lohmann S, Lange S, Ertl T. Word Cloud Explorer: Text Analytics

Based on Word Clouds. In: 2014 47th Hawaii International Conference on

System Sciences. IEEE; 2014.

34
48. McNaught C, Lam P. Using Wordle as a Supplementary Research Tool. The

Qualitative Report. 2010 May;15(3):630–43.

49. Padmanabhan K, Hendrix W. Introduction. In: Practical graph mining with R.

Boca Raton: CRC Press; 2013. p. 1–7.

50. Corbin J, Strauss A. Basics of Qualitative Research. Techniques and Procedures

for Developing Grounded Theory. 4th ed. Los Angeles: Sage Publications; 2015.

51. Beyer H, Holtzblatt K. Contextual Design: Defining Customer-Centered

Systems. San Francisco, CA, USA: Morgan Kaufmann Publishers Inc.; 1997.

52. Horvath A-S. ComPara: A Corpus Linguistics Dataset of Computation in

Architecture. Mendeley; 2020.

53. Aielo C. Evolo Skyscrapers. New York: Evolo Press; 2012.

54. Aielo C. eVolo Skyscrapers 2: 150 New Projects Redefine Building High. New

York: Evolo Press; 2014.

55. Stoevski P. The New Tower of Babel [Internet]. Honorable Mention 2014

Skyscraper Competition. 2014 [cited 2020 Nov 3]. Available from:

https://www.evolo.us/the-new-tower-of-babel/

56. Asadov N. House of Babel: Post-Crisis Skyscraper [Internet]. Honorable Mention

2012 Skyscraper Competition. 2012 [cited 2020 Nov 3]. Available from:

https://www.evolo.us/house-of-babel-post-crisis-skyscraper/

57. Song Q, Pengfei K, Ying B, Nuoya R, Shen G. Sand Babel: Solar-Powered 3D

Printed Tower [Internet]. Honorable Mention 2014 Skyscraper Competition.

2014 [cited 2020 Nov 3]. Available from: https://www.evolo.us/sand-babel-solar-

powered-3d-printed-tower/

35
58. Hsin L te. Taiwan Babel Tower [Internet]. Honorable Mention 2016 Skyscraper

Competition. 2016 [cited 2020 Nov 3]. Available from:

https://www.evolo.us/taiwan-babel-tower/

59. Sun Y, Xu T, Zhang L, Wang D, Wang T. The Scaffold of Babel [Internet].

Honorable Mention 2017 Skyscraper Competition. 2017 [cited 2020 Nov 3].

Available from: https://www.evolo.us/the-scaffold-of-babel/

60. Quinones L. Moonscraper [Internet]. Honorable Mention 2011 Skyscraper

Competition. 2011 [cited 2020 Nov 3]. Available from:

http://www.evolo.us/moonscraper/

61. Dong M, Xiang Y, Xie A, Han X. Stratosphere Network of Skyscrapers. 2013

Mar 12 [cited 2020 Nov 3]; Available from: https://www.evolo.us/stratosphere-

network-of-skyscrapers/

62. Tognoni E, Tinti F, Mariani D. Rhizome Tower: A Thousand Underground

Plateaus [Internet]. Honorable Mention 2011 Skyscraper Competition. 2011

[cited 2020 Nov 3]. Available from: https://www.evolo.us/rhizome-tower-a-

thousand-underground-plateaus/

63. Chauvel S. Quantum City [Internet]. Special Mention 2007 Skyscraper

Competition. 2007 [cited 2020 Nov 3]. Available from:

https://www.evolo.us/quantum-city/

64. Maltsev I, Melnik A. Quantum Skyscraper: Multipurpose Research Complex

[Internet]. Honorable Mention 2013 Skyscraper Competition. 2013 [cited 2020

Nov 3]. Available from: https://www.evolo.us/quantum-skyscraper-

multipurpose-research-complex/

65. Murphy R. Sustainability: A Wicked Problem. Sociologica. 2012;(2):1–23.

36
66. Joksimovic A, Nikolic J. Noah’s Ark: Sustainable City [Internet]. Honorable

Mention 2012 Skyscraper Competition. 2012 [cited 2020 Nov 3]. Available

from: https://www.evolo.us/noah%E2%80%99s-ark-sustainable-city/

67. Chen H, Guo L. Oceanscraper. Honorable Mention 2012 Skyscraper

Competition. 2012.

68. Vlastic M, Djordjevic V, Jovanovic M, Markovic D. Moses: A Decentralized

Floating Network of Skyscraper Cities [Internet]. Honorable Mention 2013

Skyscraper Competition. 2013 [cited 2020 Nov 3]. Available from:

https://www.evolo.us/moses-a-decentralized-floating-network-of-skyscraper-

cities/

69. Yao C, Yunfeng X, Xiaodi L, Rui X, Xiaoxiang Y. Promised Land Waterscraper

[Internet]. Honorable Mention 2013 Skyscraper Competition. 2013 [cited 2020

Nov 3]. Available from: https://www.evolo.us/promised-land-waterscraper/

70. Dunne A, Raby F. Speculative Everything. Boston: MIT Press; 2013. 1–3.

71. Steadman P. The Evolution of Designs: Biological Analogy in Architecture and

the Applied Arts. 2nd ed. Abington, Oxton: Routledge; 2008.

72. Chayaamor-Heil N, Vitalis L. Biology and architecture: An ongoing

hybridization of scientific knowledge and design practice by six architectural

offices in France. Frontiers of Architectural Research [Internet]. 2021;10(2):240–

62. Available from:

https://www.sciencedirect.com/science/article/pii/S2095263520300704

73. Pohl G, Nachtigall W. Biomimetics for Architecture and Design: Nature -

Analogies - Technology. Biomimetics for Architecture and Design. Cham:

Springer International Publishing AG; 2015.

37
74. Gumuskaya G. Multimaterial bioprinting—minus the printer: Synthetic bacterial

patterning with UV-responsive genetic circuits. International Journal of

Architectural Computing [Internet]. 2021;19(2):121–41. Available from:

https://doi.org/10.1177/1478077120963373

75. Mizobuti V, Vieira Junior LCM. Bioinspired architectural design based on

structural topology optimization. Frontiers of Architectural Research [Internet].

2020;9(2):264–76. Available from:

https://www.sciencedirect.com/science/article/pii/S2095263519300949

76. Pasquero C, Poletto M. Bio-digital aesthetics as value system of post-

Anthropocene architecture. International Journal of Architectural Computing.

2020 Jun 5;18(2).

77. Picon A. Architecture and Mathematics: Between Hubris and Restraint.

Architectural Design. 2011 Jul;81(4).

78. Arida A. Quantum City. 1st ed. Architectural Press; 2002.

79. Heinrich Wölfflin. Principles of Art History: the Problem of the Development of

Style in Later Art. 7th ed. New York: Dover; 1950.

80. Cross N. Designerly Ways of Knowing. In: Designerly Ways of Knowing.

London: Springer; 2005. p. 1–13.

81. DeLanda M. Deleuze and the Use of the Genetic Algorithm in Architecture.

Architectural Design. 2001;72(1):7–12.

82. Tognoni E, Tinti F, Mariani D. Rhizome Tower: A Thousand Underground

Plateaus. eVolo Skyscraper competition. 2011.

83. Frichot H, Loo S. Deleuze and Architecture. Edinburgh: Edinburgh University

Press; 2013.

38
84. DP Architects. Rhizome House [Internet]. Archello. 2016 [cited 2020 Nov 3].

Available from: https://archello.com/project/rhizome-house

85. Erik Yek Ung Hieng. The Rhizome and The Wasted Square [Internet].

KooZArch. 2018 [cited 2020 Nov 3]. Available from:

https://www.koozarch.com/interviews/the-rhizome-and-the-wasted-square/

86. Jakobsen A. Experience in-between architecture and context: the New Acropolis

Museum, Athens. Journal of Aesthetics & Culture. 2012 Jan 25;4(1).

87. White SR. Gilles Deleuze and the project of architecture: an expressionist

design‐research methodology. PhD by Architectural Design, The Bartlett; 2014.

88. Ceschin F, Gaziulusoy I. Evolution of design for sustainability: From product

design to design for system innovations and transitions. Design Studies. 2016

Nov;47.

89. Pop D. Space Perception and Its Implication in Architectural Design. Acta

Technica Napocensis: Civil Engineering & Architecture. 2013;56(2):211–21.

90. Dewey J. Art as Experience. (orig. 1934). New York: The Berkley Publishing

Group; 2005.

91. Buxton B. Sketching User Experiences: Getting the Design Right and the Right

Design. San Francisco: Morgan Kaufmann Publishers; 2006.

92. Leach N. AI and creativity. University of Bologna, Department of Architecture.

Bologna; 2020.

93. Steinfeld K. Significant others: Machine learning as actor, material and

provocateur in art and design. In: As I, Basu P, editors. The Routledge

Companion to Artificial Intelligence in Architecture. Abington, Oxon; New

York: Routledge, 2021.: Routledge; 2021.

39
94. Nisztuk M, Myszkowski PB. Hybrid Evolutionary Algorithm applied to

Automated Floor Plan Generation. International Journal of Architectural

Computing. 2019 Sep 13;17(3).

95. del Campo M, Carlson A, Manninger S. Towards Hallucinating Machines -

Designing with Computational Vision. International Journal of Architectural

Computing. 2021 Mar 17;19(1).

96. As I, Pal S, Basu P. Artificial intelligence in architecture: Generating conceptual

design via deep learning. International Journal of Architectural Computing. 2018

Dec 28;16(4).

97. Cutellic P. Towards encoding shape features with visual event-related potential

based brain–computer interface for generative design. International Journal of

Architectural Computing. 2019 Mar 14;17(1).

98. Ghandi M, Blaisdell M, Ismail M. Embodied empathy: Using affective

computing to incarnate human emotion and cognition in architecture.

International Journal of Architectural Computing. 2021 Aug 28;

40