Visualizing

Complexity
Darjan Hil
Nicole Lachenmeier

Visualizing
Complexity

Modular
Information
Design
Handbook

Birkhäuser
Basel
Contents 6 Preface
8 How the book is structured
11 The MID system

1
Data
15 dimensions
17 1A From text to
data
31 1B From data to
data cube

2
Diagrammatical
41 dimensions
43 2A Quantity
59 2B Position
67 2C Relationship

3
Visual
77 dimensions
79 3A Color
87 3B Shape
97 3C Line
107 3D Pattern
115 3E Contour
123 3F Isotype
 4
Structuring
133 dimensions
135 4A Structuring by
sorting
147 4B Structuring by
grouping

5
The variety of
­multidimensional
154 ­visualizations
156 Overview of all 80
elements
158 How the MID system
works
160 Mapping four data
dimensions — instructions
162 How sorting works
164 26 multidimensional
­visualizations

6
217 Appendix
218 The authors
220 Further reading, sources,
,

and tools
Preface

The title of this book, “Visualizing Complexity”, contains the word

complexity, a frequently used term, which quite fittingly describes
current thinking. We live in a time when, thanks to computers
and an infinite amount of information, people are doing more and
more tasks at the same time, taking on different roles and switching
very quickly between subjects. In this context, we understand
complexity to mean that an issue is impacted by many factors, all
of them interwoven, and is obscure and difficult to grasp. But
why visualize complexity? Quite simply because we firmly believe
that many a seemingly complex issue can be better processed when
it is externalized visually and can be viewed from some distance.
Initially, it is not about the quality of the visualization, but simply
about the process of getting something out of one’s head onto the
page (digital or analog) and thereby making it communicable.
In the context of this book, we understand issues to be data as
a general concept. Whether we are dealing with text or numbers,
in our opinion both are raw material, i.e. a data set to be processed.
Both can be considered in terms of quantity or quality, and can
be processed visually. In this context it is important to be aware that
any consideration of an issue assumes a certain perspective and
tells an associated story. For each set of data there are various
approaches and perspectives. Therein lies the complexity mentioned
above. The aim is not to simplify the complexity, but to enable the
reader to skilfully navigate through the different perspectives of the
topic. To facilitate this navigation, we have developed a method
called modular information design; MID for short. In this system,
all elements, the smallest units of the system, are made combinable
in modules similar to the periodic table system employed in chem­
istry. It is important to note that in order to facilitate simplified
orientation in the MID system, elements are grouped into higher-
level dimensions. Dimensions are units used to organize elements at
a certain level.

6
This book is the result of ten years of practical experience in which
we enjoyed experimenting and carried out analysis work with the
aim of making the little known — to date — but demanding interface
discipline of information design more accessible through straight­
forward principles. We are convinced that visualization as a general
problem-solving approach helps to create greater transparency
and enables wider participation and better decision-making. In addi­
tion, people tend to prefer seeing and reading attractively presented
text and figures — and understand them better — than the usual pre­
sentations with their standard diagrams. We want to demonstrate
that the world of information design has many beautiful facets. We
therefore dare — in an unconventional way — to make connections
between disciplines and issues where before there were no obvious
ones, the objective being to increase interdisciplinary understanding.
This book is intended to be a source of inspiration to all those
who work with figures so that they may see data from a new per­
spective and are able to explore design options for the communica­
tion of information. For designers, this handbook can make figures
more accessible, and thereby open up new design opportunities for
the presentation of information. Generally speaking, we want to
convey the joy of working correctly, but also creatively, with text
and figures, and we want to motivate people to start their own visual­
ization of complexity. For this it is not necessary to have special
talent, or a computer, or special training — but instead a pen and
paper, an interest, and a willingness to experiment. We hope that you
will enjoy working with this handbook and wish you every success.

Darjan Hil and Nicole Lachenmeier

7
How the book
is ­structured

This book is structured visually and Yellow section

subdivided into color-coded sections. In the yellow section, the structure of a double
page consists of a data set, an explanation,
The first four sections convey the a title, and an abstract icon as the representa­
necessary basic understanding and tion of the principle explained therein. The
form the basis for the examples of yellow section, as well as the entire book, only
uses a single small data set that describes
multidimensional visualizations found a fictitious family history.
in the fifth section.
It is important to know that the Blue, red, green sections
In the blue, red, and green sections, the ele­
book does not have to be read from ments of the modular information design are ex­
front to back, but can be skimmed for plained on a single page each. An abstract
inspiration. The structure of the ex­ icon represents each principle, which is supple­
mented by an element title and an element ID.
amples can easily be decoded at the In this way, the elements can be traced through­
end of the book, and details can be out the entire book. A simple example in the
checked in the earlier sections. bottom left-hand corner and a short explanation
gives each element a context for practical
­application.

Example section
In this section, each example of data select­
ed from the data set is combined on one page
with the elements presented, thus creating a
multi­dimensional visualization example. Each
page consists of a listing of the data dimensions
used and the used elements as icons with the
element ID. Step by step, the structure becomes
clear with the help of an accurate legend and
a descriptive title, as well as a short highlighted
text with the specific findings for telling the
story from the diagram.

8
 Explanation

Title Data set

Icon

Simple
example
Element–ID

Descriptive
title

Legend
Highlight

9
The MID system

The construction of the visualizations in this book is based on

a system. We understand the term system to mean that a number
of different components, together with certain principles, form a
whole. When applying the principle of deliberately selecting ele­
ments from the modular system and combining them with different
dimensions to create a new visualization, we speak of modular
information design.
The MID system consists of four higher-level modules: the
data dimensions (yellow), the diagrammatical dimensions (blue),
the visual dimensions (red), and the structuring dimensions
(green). Together, the last three dimensions form 80 distinctly
identifiable elements.
The idea behind the MID system is that by combining a data
set with selected elements of the three groups, a large number of
multidimensional visualizations can be designed. Using a simple
framework, it is possible to encode not only all known diagram
types but also a nearly infinite number of new visualizations. This
system can be understood as a visual algorithm. But likewise, it is
also possible to use MID to analyze and decode existing visualiza­
tions. Multi­dimensional information design does not depend on
specific tools. It is possible to use pen and paper and a wide range
of different software tools. The fifth chapter contains step-by-step
instructions.

11
The MID system

Data dimensions Structuring dimensions

The data-based dimensions describe the steps The structuring dimensions have the greatest
associated with the analysis of a data set. influence on the story of a visualization. The
Here, the primary concern is not procedures question here is: What are the data dimensions
relating to statistics but the essential conceptual used for sorting and grouping a diagram?
considerations as to which questions can be Just by opting for different ways of sorting it is
clarified with this data set, which stories can possible to tell different stories. The sorting
be told, and which perspectives can be ­adopted. also determines the arrangement of the indi­
From this, the selection of the individual data vidual data elements and therefore has a great
dimensions is derived, i.e. the columns of a influence on the overall shape. On the other
data set. hand, grouping brings harmony and order into
the format.
Diagrammatical dimensions
The diagrammatical dimensions contain all Multidimensional visualizations
­elements collected from the various areas of Through the multidimensional visualizations, all
mathematics and information technology. dimensions are applied together. New dia­
The common denominator of this collection is gram variations are created from three or more
that a data set can be translated into a simple data dimensions that are mapped onto the
diagram in a mathematically correct manner. elements of the other three dimensions. For the
The diagrams thus created describe either multidimensional visualizations, it is always
quantities, positions in a system of coordinates, necessary to have at least four elements.
or ­relationships.

Visual dimensions
The visual dimensions can be looked upon
as the creative dimensions. Here, a great many
design principles can be used to create a
dia­gram with several data dimensions. Gener­
ally speaking, the visual elements presented
here are part of the basic vocabulary of design­
ers. We have subdivided these into the fol­
lowing groups: color, shape, line, pattern, con­
tour, and the special group of isotypes,
according to Neurath.1

12
 Data dimensions

Diagrammatical
dimensions

Visual
dimensions

Structuring
dimensions

Multidimensional
visualizations

13
1
Data
dimensions

What do we need to know about a data set in order to be able to

decide which data dimensions should be used for the visualization?
Do we already at this stage need analyses or statistics methods
to answer this question? Is it possible to look neutrally at a data
set whilst observing a certain distance? Are there intermediate sta-
gesat which visual patterns can be detected in the data set or
in the different dimensions of the data set? In this chapter we illus-
trate approaches that answer these and other similar questions
and demonstrate a flowing transition from analysis to visualization.
Regarding the conceptual consideration of a data set from its
possible access points, a fundamentally open attitude is required.
This makes it possible to engage in interdisciplinary cooperation
and communication on equal terms.

15
1A
From text
to data

In this chapter we present a data set as a basis for the entire book.
It is a text about the Wagner and Brown families, which can be
used to extract and visualize data in a step-by-step process. The
names of the family members are based on the statistically most
common family and first names of the 19 th century in German-
and English-speaking countries. The objective is to illustrate the
individual steps required for visualization and to make the possible
variations visible, without having to change too many para­meters at
once. Please note that the story is fictitious and does not relate to
real events. This simple family story makes it possible to understand
data easily and establish a connection between them and you.

17
Unstructured running text as a
starting point

Running text is intended to give the

reader time to bring to mind the respec-
tive content in one’s imagination. Written
stories feature descriptive details and
inspire the readers' power of imagina-
tion, encouraging them to mentally dive
into their own world. The information
in a section is unstructured, without how-
ever having a negative effect on how
easy it is to understand.
Narrators tell their stories in the way
their thoughts flow. However, for the
purpose of searching or obtaining a
quick overview, narrations are less suit-
able. For example, when searching for
Otto Wagner’s date of birth in the text
about the entrepreneurial family, one has
to scan the text line by line until this
is found. Clearly in this case we cannot
talk about a quick overview or rapid
­navigation.

18
The story of the
entrepreneurial families

The success of the Browag AG company goes back to Hermann Wagner and
James Brown. Hermann Wagner was born in Vienna in 1871 (died 1961). During
his student years he got to know Vienna-born Emma Moser (1875), who was
four years younger (died 1960). In 1900, Hermann and Emma married in Vienna,
where the children Otto (1901, died 1924) and Paul (1914, died 2011) were also
born. With the help of the Moser family’s financial resources, Hermann was
able to set up Wagner Farben GmbH and establish it successfully in Vienna.
After finishing school, Otto decided to go to Munich to study and Paul decided
to follow in his father Hermann’s footsteps. Otto Wagner died tragically of
­tuberculosis in Munich at the age of 23.
In the course of his business, Paul traveled to other countries and, in 1934,
whilst in Paris, he met his future wife, ­Elisabeth Brown of Brown Chemicals Inc.
The company Brown Chemicals Inc. had been founded in London by James
Brown (born 1882, died 1947). He benefited greatly from his marriage to Marie
Durand (born 1879, died 1951), who came from a prosperous Parisian family.
James built up his business in London as well as in Paris, where the couple
spent most of their time. Marie and James had two daughters, Anna (born 1913,
died 1996) and Elisabeth (born 1915, died 2014). Anna was born while the family
spent some time in London; she grew up there, went to a London boarding
school, and spent the rest of her life in London. Her sister Elisabeth grew up in
Paris where she lived with her parents; from an early age, she joined papa
James’s company, where she later met Paul Wagner.
The wedding between Paul and Elisabeth not only sealed the matrimo-
nial bond, but also the business relationship, which led to the formation of the
­newly merged Browag AG. In 1935, Hermann was born in Paris (died 1987)
and seven years later, along came his sister Marie (born 1942, died 2020).
Owing to his commitment to the time-consuming management of the branch
in Vienna, Paul did not move away from Vienna.
Thus it came about that — some years later —Elisabeth and Paul separated.
Paul decided to sell his company shares to Elisabeth, and to quit Browag AG.
Their son Hermann decided to go to Vienna to study and to live with his grand-
parents, Hermann and Emma. Marie, on the other hand, wanted to study art
in London and decided not to take up her mother’s offer of joining the company
in Paris. Elisabeth Brown is considered one of the most successful female en-
trepreneurs of the 21st century and is leaving her entire fortune to charitable
organizations with a focus on design.

19
Unstructured running text
with highlights

In this version, the running text has

already been enriched with a few
support­ing symbols. Each symbol rep-
resents a certain type of information
and is marked in yellow. Although
at this stage we cannot yet talk about
a visualization, a search, e.g. for Otto
Wagner’s date of birth is made easier.
The text is still unstructured.

▶ First name
◀ Family name
● Year of birth
◖ Year of death
■ Place of birth
◆Home town

20
The story of the
entrepreneurial families

The success of the Browag AG company goes back to ▶ Hermann ◀ Wagner

and ▶ James ◀ Brown. ▶ Hermann ◀ Wagner was born in ■ Vienna in
● 1871 (died ◖ 1961). During his student years he got to know ■ Vienna-born
▶ Emma Moser (1875), who was four years younger (died ◖ 1960). In 1900,
▶ Hermann and ▶ Emma married in ◆ Vienna, where the children ▶ Otto
(● 1901, died ◖ 1924) and ▶ Paul (● 1914, died ◖ 2011) were also born. With the
help of the ◀ Moser family’s financial resources, ▶ Hermann was able to
set up Wagner Farben GmbH and establish it successfully in ◆ Vienna. After
finishing school, ▶ Otto decided to go to ◆ Munich to study and ▶ Paul
decided to follow in his father ▶ Hermann’s footsteps. ▶ Otto ◀ Wagner died
tragically of tuberculosis in ◆ Munich at the age of 23.
In the course of his business, ▶ Paul traveled to other countries and,
in 1934, whilst in Paris, he met his future wife, ▶ Elisabeth ◀ Brown of Brown
Chemicals Inc. The company Brown Chemicals Inc. had been founded in
■ London by ▶ James ◀ Brown (born ● 1882, died ◖ 1947). He benefited greatly
from his marriage to ▶ Marie Durand (born ● 1879, died ◖ 1951), who came
from a prosperous ■ Parisian family. ▶ James built up his business in ■ London
as well as in ◆ Paris, where the couple spent most of their time. ▶ Marie and
▶ James had two daughters, ▶ Anna (born ● 1913, died ◖ 1996) and ▶ Elisabeth
(born ● 1915, died ◖ 2014). ▶ Anna was born while the family spent some
time in ■ London; she grew up there, went to a London boarding school, and
spent the rest of her life in ◆ London. Her sister ▶ Elisabeth grew up in ■ Paris
where she lived with her parents; from an early age, she joined papa ▶ James’s
company, where she later met ▶ Paul ◀ Wagner.
The wedding between ▶ Paul and ▶ Elisabeth not only sealed the matri­
monial bond, but also the business relationship, which led to the formation of
the newly merged Browag AG. In ● 1935, ▶ Hermann was born in Paris (died
◖ 1987) and seven years later, along came his sister ▶ Marie (born ● 1942, died
◖ 2020). Owing to his commitment to the time-consuming management of
the branch in Vienna, ▶ Paul did not move away from ◆ Vienna.
Thus it came about that — some years later— ▶ Elisabeth and ▶ Paul sep­
arated. ▶ Paul decided to sell his company shares to ▶ Elisabeth, and to
quit Browag AG. Their son ▶ Hermann decided to go to ◆ Vienna to study and
to live with his grandparents, ▶ Hermann and ▶ Emma. ▶ Marie, on the other
hand, wanted to study art in ◆ London and decided not to take up her mother’s
offer of joining the company in ◆ Paris. ▶ Elisabeth ◀ Brown is considered
one of the most successful female entrepreneurs of the 21 st century and is leav-
ing her entire fortune to charitable organizations with a focus on design.

21
Running text as a list,
structured with highlights

A mere compilation of data in the form

of a list is not enough to structure
the under­lying text. A list guides the eye
through the listed items. The added
­value is not achieved until the sentence
structure becomes comparable through
identical arrangement. The symbols
are used to make things even clearer. It
is possible — albeit not yet very easy —
to detect a pattern in the lines.

▶ First name
◀ Family name
● Year of birth
◖ Year of death
■ Place of birth
◆Home town

22
The story of the
entrepreneurial families

• ▶ Hermann ◀ Wagner, born ● 1871 in ◆ Vienna, lived in ◆ Vienna, died

◖ 1961, married to ▶ Emma ◀ Moser, children ▶ Otto and ▶ Paul
• ▶ Emma ◀ Moser later ◀ Wagner, ● 1875 born in ■ Vienna, lived in
­◆ Vienna, died ◖ 1960, married to ▶ Hermann ◀ Wagner, children ▶ Otto
and ▶ Paul
• ▶ Otto ◀ Wagner, born ● 1901 in ■ Vienna, lived in ◆ Munich, died ◖ 1924
• ▶ Paul ◀ Wagner, born ● 1914 in ■ Vienna, lived in ◆ Vienna, died ◖ 2011,
married to ▶ Elisabeth ◀ Brown, children ▶ Hermann and ▶ Marie
• ▶ James ◀ Brown, born ● 1882 in ■ London, lived in ◆ Paris, died ◖ 1947,
married to ▶ Marie ◀ Durand, children ▶ Elisabeth and ▶ Anna
• ▶ Marie ◀ Durand later ◀ Brown, born ● 1879 in ■ Paris, lived in ◆ Paris,
died ◖ 1951, children ▶ Elisabeth and ▶ Anna
• ▶ Elisabeth ◀ Brown later ◀ Wagner, born ● 1915 in ■ Paris, lived in ◆ Paris,
died ◖ 2014, married to ▶ Paul ◀ Wagner, children ▶ Hermann and ▶ Marie
• ▶ Anna ◀ Brown, born ● 1913 in ■ London, lived in ◆ London, died ◖ 1996
• ▶ Hermann ◀ Wagner, born ● 1935 in ■ Paris, lived in ◆ Vienna, died ◖ 1987
• ▶ Marie ◀ Wagner, born ● 1942 in ■ Paris, lived in ◆ London, died ◖ 2020

23
Structure in table form

When, in a list consisting of sentences

with an identical structure, regular
­spaces are inserted between the differ-
ent parts of the sentence, a table is
created. The placement of structured
text elements in a grid generates good
readability and comparability of the
contents. The highlighting symbols
used previously are no longer required
here — the gap between the parts
of the sentence orders the elements in
accordance with their affiliation and
columns are created.

▶ First name
◀ Family name
● Year of birth
◖ Year of death
■ Place of birth
◆Home town

24
The story of the
entrepreneurial families

◀ Family ▶ First ■ Place of ● Year of ◆ Home ◖ Year of ▶ Children

name name birth birth town death
◀ Wagner ▶ Hermann ■ Vienna ● 1871 ◆ Vienna ◖ 1961 ▶ Otto
▶ Paul
◀ Wagner ▶ Emma ■ Vienna ● 1875 ◆ Vienna ◖ 1960 ▶ Otto
(◀ Moser) ▶ Paul
◀ Brown ▶ Marie ■ Paris ● 1879 ◆ Paris ◖ 1951 ▶ Anna
(◀ Durand) ▶ Elisabeth
◀ Brown ▶ James ■ London ● 1882 ◆ Paris ◖ 1947 ▶ Anna
▶ Elisabeth
◀ Wagner ▶ Otto ■ Vienna ● 1901 ◆ Munich ◖ 1924

◀ Brown ▶ Anna ■ London ● 1913 ◆ London ◖ 1996

◀ Wagner ▶ Paul ■ Vienna ● 1914 ◆ Vienna ◖ 2011 ▶ Hermann

▶ Marie
◀ Wagner ▶ Elisabeth ■ Paris ● 1915 ◆ Paris ◖ 2014 ▶ Hermann
(◀ Brown) ▶ Marie
◀ Wagner ▶ Hermann ■ Paris ● 1935 ◆ Vienna ◖ 1987

◀ Wagner ▶ Marie ■ Paris ● 1942 ◆ London ◖ 2020

25
Structure in table form,
nested

Through a new arrangement, it is

possible to organize a table such that
repeating terms can be omitted.
This has the effect of making patterns
discernible already at the table stage.
Nesting a table in this way increases
the read­ability many times over. In
this step it is always advisable to start
with the column that has the fewest
expressions. In this case it is the family
name, with the two expressions
Wagner and Brown.

▶ First name
◀ Family name
● Year of birth
◖ Year of death
■ Place of birth
◆Home town

26
The story of the
entrepreneurial families

◀ Family ■ Place of ◆ Home ▶ First ● Year of ◖ Year of ▶ Children

name birth town name birth death
◀ Wagner ■ Vienna ◆ Vienna ▶ Hermann ● 1871 ◖ 1961 ▶ Otto
▶ Paul
▶ Emma ● 1875 ◖ 1960 ▶ Otto
▶ Paul
▶ Paul ● 1914 ◖ 2011 ▶ Hermann
▶ Marie
◆ Munich ▶ Otto ● 1901 ◖ 1924

■ Paris ◆ Vienna ▶ Hermann ● 1935 ◖ 1987

■ Paris ◆ London ▶ Marie ● 1942 ◖ 2020

◀ Brown ■ Paris ◆ Paris ▶ Marie ● 1879 ◖ 1951 ▶ Anna

▶ Elisabeth
▶ Elisabeth ● 1915 ◖ 2014 ▶ Hermann
▶ Marie
■ London ◆ Paris ▶ James ● 1882 ◖ 1947 ▶ Anna
▶ Elisabeth
◆ London ▶ Anna ● 1913 ◖ 1996

27
Hierarchical structure

The construction of a family tree is

founded on two basic elements: firstly,
on a clear structure, as in a table,
and secondly, on the further information
about the connections between the
various people mentioned in the text.
The newly created order makes it
possible to discern patterns even faster.
Family trees are often used in visual­
ization to represent hierarchies in groups.
In this concrete example, the family
tree shows the sequence of relationships
between the generations of the two
families. Usually, a box represents
ap­ erson, and the nodes and linear con-
nections represent the relationships ▶ First name
­between the members. ◀ Family name
● Year of birth
◖ Year of death
■ Place of birth
◆Home town

28
The story of the
entrepreneurial families

◀ Wagner ◀ Wagner (◀ Moser) ◀ Brown ◀ Brown (◀ Durand)

▶ Hermann ▶ Emma ▶ James ▶ Marie
■ Vienna ■ Vienna ■ London ■ Paris
● 1871 ◖ 1961 ● 1875 ◖ 1960 ● 1882 ◖ 1947 ● 1879 ◖ 1951
◆ Vienna ◆ Vienna ◆ Paris ◆ Paris

◀ Wagner ◀ Wagner ◀ Wagner (◀ Brown) ◀ Brown

▶ Otto ▶ Paul ▶ Elisabeth ▶ Anna
■ Vienna ■ Vienna ■ Paris ■ London
● 1901 ◖ 1924 ● 1914 ◖ 2011 ● 1915 ◖ 2014 ● 1913 ◖ 1996
◆ Munich ◆ Vienna ◆ Paris ◆ London

◀ Wagner ◀ Wagner
▶ Hermann Jr. ▶ Marie Jr.
■ Paris ■ Paris
● 1935 ◖ 1987 ● 1942 ◖ 2020
◆ Vienna ◆ London

29
1B
From data to
data cube

In order to make it easier to orient the phases of data analysis and

help with understanding the different ways a data set can be
viewed, we have developed the data cube as a metaphor. We are not
so much referring to the fact that each cube has exactly six equal
sides, but instead are using this term in the metaphorical sense that
a cube can be viewed from different sides. For this reason, we
could also refer to the cube as a meta-cube or hyper-cube, but for
simplicity’s sake we’ll just stick with data cube. For example, each
column of a table, i.e. each data dimension, can be understood
as one side of the cube. The steps proposed here for data analysis
belong to the field of elementary exploratory statistics. We only
want to introduce as much data science as is necessary for the next
chapters to show different perspectives of the entrepreneurial
families’ data set used as an example.

31
 Data dimensions as perspectives
on a data cube

Family name First name Place of birth Year of birth Home town Year of death

Wagner Hermann Vienna 1871 Vienna 1961

Wagner (Moser) Emma Vienna 1875 Vienna 1960

Brown (Durand) Marie Paris 1879 Paris 1951

Brown James London 1882 Paris 1947

Wagner Otto Vienna 1901 Munich 1924

Brown Anna London 1913 London 1996

Wagner Paul Vienna 1914 Vienna 2011

Wagner (Brown) Elisabeth Paris 1915 Paris 2014

Wagner Hermann Jr. Paris 1935 Vienna 1987

Wagner Marie Jr. Paris 1942 London 2020

32
 Three of the eleven data dimensions as The metaphor of a data cube helps us view
perspectives on three sides of the a data set from a number of different an-
data cube of the entrepreneurial families. gles. We refer to these perspectives as data
­dimensions (here, as columns of the table).
The entrepreneurial family can be ­considered
from the aspect of its family name, its home
towns, or from any other perspective. It is
important that the core always contains the
same data set. Each viewing angle results
in a different approach, analysis, statement,
■ Place of birth and story.
The data from the first part of this chap-
ter is listed here in the form of a table on
the left-hand side. However, it is also possi-
● Year of birth ble to use the existing data to derive new
data, groups, or categorizations: the first
names can be used to infer the gender, the
date of birth to infer the generation, the
◀ Family name home town and the place of birth to infer
relocations, the years of birth and death to
infer the age at death, and from this, the
age group too.This new data can be found in
the table on the right-hand side.

Gender Generation Relocation Age at death Age group

M 1 N 90 > 85

F 1 N 85 70 – 85

F 1 N 72 70 – 85

M 1 Y 65 < 70

M 2 Y 23 < 70

F 2 N 83 70 – 85

M 2 N 97 > 85

F 2 N 99 > 85

M 3 Y 52 < 70

F 3 Y 78 70 – 85

33
 Data dimensions with the
data types and expressions

Alphabet
Location

Hierarchy Time

Category

Family name First name Place of birth Year of birth Home town Year of death

Wagner Hermann Vienna 1871 Vienna 1961

Wagner (Moser) Emma Vienna 1875 Vienna 1960

Brown (Durand) Marie Paris 1879 Paris 1951

Brown James London 1882 Paris 1947

Wagner Otto Vienna 1901 Munich 1924

Brown Anna London 1913 London 1996

Wagner Paul Vienna 1914 Vienna 2011

Wagner (Brown) Elisabeth Paris 1915 Paris 2014

Wagner Hermann Jr. Paris 1935 Vienna 1987

Wagner Marie Jr. Paris 1942 London 2020

Alphabet Alphabet Location Time Location Time

2 8 3 10 4 10
Expressions

34
 The data cube of the entrepreneurial Each data dimension, or each perspective on
families with three data dimensions, the data cube, can be assigned to a ­certain
differentiated by data type and the type of data. In general, information can
number of expressions. be organized in accordance with five types
of data, which are geolocation, text, time,
category, or figures. This ordering ­principle
“Place of birth”—
was defined by Richard Saul ­Wurman2 under
data dimension of the
Location type as
the name L.A.T.C.H.: Location, Alphabet
­c ategory with three (to be understood as text), Time, Category,
expressions — Vienna, Hierarchy (to be understood as numbers).
Paris, and London Each data type can have a finite or in-
finite number of expressions — values or text.
“Year of birth” — Although in our example on the left the
data dimension of data type “family name” is text, the number
the Time type, of expressions can be limited to two:
always with different
Wagner and Brown. When the expressions
expressions
are finite, each L.A.T.C.H. data type can
be used to form a category. “Family name” is
“Family name”—
a category with two expressions.
data dimension of
the Alphabet type as
category with two
expressions — Wagner
and Brown

Gender Generation Relocation Age at death Age group

M 1 N 90 > 85

F 1 N 85 70 – 85

F 1 N 72 70 – 85

M 1 Y 65 < 70

M 2 Y 23 < 70

F 2 N 83 70 – 85

M 2 N 97 > 85

F 2 N 99 > 85

M 3 Y 52 < 70

F 3 Y 78 70 – 85

Category Category Category Hierarchy Category

2 3 2 10 3

35
 Simple evaluation of the
­expressions of a data
dimension

Family name First name Place of birth Year of birth Home town Year of death

Wagner Hermann Vienna 1871 Vienna 1961

Wagner (Moser) Emma Vienna 1875 Vienna 1960

Brown (Durand) Marie Paris 1879 Paris 1951

Brown James London 1882 Paris 1947

Wagner Otto Vienna 1901 Munich 1924

Brown Anna London 1913 London 1996

Wagner Paul Vienna 1914 Vienna 2011

Wagner (Brown) Elisabeth Paris 1915 Paris 2014

Wagner Hermann Jr. Paris 1935 Vienna 1987

Wagner Marie Jr. Paris 1942 London 2020

Alphabet Alphabet Location Time Location Time

Wagner: 7 Hermann: 2 Vienna: 4 First: 1871 Vienna: 4 First: 1924

Brown: 3 Marie: 2 Paris: 4 Last: 1942 Paris: 3 Last: 2020
Other: 6 London: 2 Years: 71 London: 2 Years: 96
Middle: 1906 Munich: 1 Middle: 1972

36
 The data cube of the entrepreneurial Once the data type of a data dimension has
families with two data dimensions, been identified, it is possible to establish
evaluated by data type and expression. the number of expressions. For example: the
“family name” data dimension is of the
Alphabet data type. Since we only have two
different names, we can define “family
name” as a category with two expressions:
Vienna Paris London Wagner and Brown. Seven people are from
the Wagner family and three from the Brown
family. In the “Category” data type, the
Place of birth evaluation consists of a simple addition of
the different expressions.
With the “Time” and “Figures” (in
L.A.T.C.H.: Hierarchy) data types it is possi-
ble to carry out elementary statistical evalu-
ations, such as sum, average, maximum.
“Location” and “Text” (in L.A.T.C.H.:
Alphabet) are the two data types in which
Family name
sorting or adding the expressions, as evalua-
tion, is most effective.
Brown Wagner

Gender Generation Relocation Age at death Age group

M 1 N 90 > 85

F 1 N 85 70 – 85

F 1 N 72 70 – 85

M 1 Y 65 < 70

M 2 Y 23 < 70

F 2 N 83 70 – 85

M 2 N 97 > 85

F 2 N 99 > 85

M 3 Y 52 < 70

F 3 Y 78 70 – 85

Category Category Category Hierarchy Category

M: 5 Generation 1: 4 N: 6 Min: 23 <70: 3

F: 5 Generation 2: 4 Y: 4 Max: 99 70 – 85: 4
Generation 3: 2 Avg: 74 > 85: 3
Median: 80.5

37
 Nesting of data dimensions

Family name First name Place of birth Year of birth Home town Year of death

Wagner Wagner Wagner Wagner Wagner Wagner

Wagner: 5 Emma: 1 Paris: 3 1871 London: 1 1924

Wagner (Moser): 1 Elisabeth: 1 Vienna: 4 1875 Munich: 1 1961

Wagner (Brown): 1 Hermann: 2 Brown 1901 Paris: 1 1960

Brown Marie: 1 Paris: 1 1914 Vienna: 4 1987

Brown: 2 Otto: 1 London: 2 1915 Brown 2011

Brown (Durand): 1 Paul: 1 1935 Paris: 2 2014

Brown 1942 London: 1 2020

Anna: 1 Brown Brown

James: 1 1879 1947

Marie: 1 1882 1951

1913 1996

38
 Two-dimensional evaluation of the data By determining the sequence of the data
cube of the entrepreneurial family using dimensions, a nesting of its characteristic
nesting of data dimensions. expressions is created. In our table, the
sequence is the family name followed by
the place of birth.
Following a two-dimensional evaluation
with nesting of the data dimensions it is
possible to draw the first conclusions and
discern new patterns in the data set. With
Family name
each new data nesting, a new perspective
is created from which a data story can
be told. With each change of perspective,
Brown Wagner
understanding of the selected data set
grows. This type of approach is known in
statistics as exploratory data analysis or
Place of birth Place of birth exploratory statistics.
Having now evaluated all dimensions of
our data set and carried out all possible
London Paris Vienna
nesting, it is now the perfect point in time
to work with the 80 elements of the infor­
mation design.

Gender Generation Relocation Age at death Age group

Wagner Wagner Wagner Wagner Wagner

Female: 3 Generation 1: 2 Yes: 3 23 < 70: 2

Male: 4 Generation 2: 3 No: 4 52 70 – 85: 2

Brown Generation 3: 2 Brown 78 > 85: 3

Female: 2 Brown Yes: 1 85 Brown

Male: 1 Generation 1: 2 No: 2 90 < 70: 1

Generation 2: 1 97 70 – 85: 2

99

Brown

65

72

83

39
 2C
Relationship

2B
Position
2A
Quantity
2
Diagrammatical ­
dimensions

The 25 elements of the diagrammatical dimensions are a compila-

tion of the most frequent types of diagram from the fields of geometry
and set theory that can be distinguished at the visual level. The
construction of some diagram types requires some mathematical
understanding, whereas others are quite easy to grasp. In the context
of this chapter we do not focus on the mathematical formulae,
but concentrate exclusively on the visual translation of content into
a diagram.
The diagrammatical dimensions can be subdivided into three
groups: the representation of quantities, in our case the number of
people or a statement of age; of positions, points in a coordinate
system or on a geographical map that are non-negotiable; and finally
of relationships, in our data set those between the different family
members. In principle, representations of quantities, positions, and
relationships can also be combined with each other. For examples,
see chapter five. Each individual element in this chapter is ­symbolized
by an icon; it is visualized using a basic example from the sample
data set and supported by a short explanation.

41
2A
Quantity

Quantities are specific values that

need to be translated visually. As a
rule they are represented by numbers,
for example five women. A single
number can be represented as a line,
as a countable unit, or as an area.
In detail, and when combining several
numbers, further distinctions and
subtleties exist. Here, we have defined
13 possible diagrammatical elements
of the quantity-related ­representation.

43
Line length 2A.1
Quantity

Place of birth The line length is simply produced with

a ruler and gives a quick overview of
the different number values. It is there-
Paris fore not surprising that the bar dia-
gram is the most frequently used type
of diagram. Line lengths can be very
Vienna
easily distinguished and the diagrams
are constructed with lines of equal
thickness.
London

10 mm = 1 person

44
Line length 2A.2
Countable Quantity

Place of birth Countable line lengths are our favor-

ite when it comes to representing
quantities in an easily readable way.
Paris
Not only is it easy to produce these
manually on paper, but they also make
it possible to compare when read-
Vienna
ing without a measuring instrument.
And best of all, a dot can also assume
an identity and tell a story.
London
Tip: It is quicker to count in groups
of five.

1 dot = 1 person

45
Line length 2A.3
with line break Quantity

Place of birth Line lengths with a line break are

­always used when there is only limited
space available in the direction of
Paris the line.
This principle can often be found
in old atlases, when a value is multiple
Vienna
times greater than the other values
and the line needs a kink or a line
break. It is important to set the break
London
point at the same place for all lines.

10 mm = 1 person

46
Line length 2A.4
Countable with line break Quantity

Place of birth Countable line lengths with a line

break are a combination of the two pre-
vious element types. Compared to
the line length with a line break, the line
Paris
break used with countable dots can
follow a deliberate strategy in order to
make it easier and quicker to read
Vienna
the quantities. In this case it is import-
ant that the groups of dots have enough
distance between them.
London

1 dot = 1 person

47
Area size 2A.5
Quantity

Place of birth The area size is a compact device

and a flexible option for distinguishing
quantities, whilst allowing for free
design in space. However, the size is
Paris (4)
only roughly perceived in relation
Vienna (4)
to others. For this reason, the precise
differences should also be annotated
in numbers.
Tip: Circular areas should be calculated
London (2)
using the mathematical formula
(r 2 ), otherwise drastic errors can
result from visual perception.

48
Area size 2A.6
Countable Quantity

Place of birth The countable area size is based

on the principle that the eye is particu-
larly good at judging the proximity
of objects. Proximity and distance can
be used for design purposes and
­distri­butions can be precisely repre­
Paris sented. A unit may consist of a
geometric shape (such as a circle) or
Vienna
an organic shape.

London

1 dot = 1 person

49
Area size 100% 2A.7
as a circle from the center out Quantity

Age group In this diagram the segments of the

circle are divided from the center
out using an identical angle. The sum
of all angles is 360 degrees. Radii
of different sizes result in different-
< 70 (3) 70–85 (4) sized circle segments.

> 85 (3)

50
Area size 100% 2A.8
as a circle from the edge in Quantity

Age group The principle of this diagram is the

same as that on the left-hand side
(2A.7). Whilst the diameter is the
same, the only difference is that the
2 areas point inwards from the edge
< 70 (3) 4
6
70–85 (4) rather than growing from the center
8 outwards. The largest value must
be known from the beginning to en-
sure that the areas do not overlap
at the center.

> 85 (3)

51
Area size 100% 2A.9
in an angle as a circle Quantity

Age group When the total of all quantities adds

up to 100 percent and each value
is calculated as a percentage of a 360-­
degree circle, we obtain the well-
known pie chart. The fact that these
< 70 (3) 70–85 (4) are so widely used is due to com­
puter software, which makes it easy to
design them. Constructing such a
chart with a compass and ruler would
be too labo­rious in everyday analog life.
> 85 (3)

52
Area size 100% 2A.10
in an angle as a circle segment Quantity

Age group When there is not enough space for a

full circle as shown on the left-hand
side (2A.9) or it makes more sense to
show the area as a segment, it is
70–85 (4)
pos­sible to divide the 100 percent of
values over 180 degrees or 90 de-
grees. Whether to choose a full circle
< 70 (3) > 85 (3)
or segment is more of a design
decision than a mathematical one.

53
Area size 100% 2A.11
in an angle as a ring Quantity

Age group Basically, this diagram is to be used

and calculated in the same way as the
pie chart 2A.9. This type of diagram is
prefera­ble to a pie chart when addi-
tional information is to be placed in the
< 70 (3) 70–85 (4) center, e.g. the total quantity as a
­number. This diagram is often referred
to as a donut diagram.

> 85 (3)

54
Area size 100% 2A.12
as a rectangle Quantity

Age group The 100 percent does not have to be

subdivided into segments of a cir-
cle, but can also be represented as a
rectangle divided into smaller rect­
angles. This type of diagram is called
< 70 (3) a treemap. Calculating the different
rectangle sizes is somewhat more
70–85 (4)
complicated than calculating the size
> 85 (3) of slices in pie diagrams with a 360-
degree distribution.

55
Area size 100% 2A.13
as a countable rectangle Quantity

Age group The area of a rectangle can consist

of 100 countable dots, thus represent-
ing 100 percent of a quantity distri­
bution. We have discovered this type
of diagram as a combination of 2A.12
< 70 (3) and 2A.6 and feel certain that its po-
tential has not yet been fully exploited.
70–85 (4)
Without the existence of computers,
> 85 (3) this type would probably be more pop-
ular because it can be manually pro-
duced, quickly and intuitively.

56
2B
Position

Points in a coordinate system are not

negotiable; they have to be exactly
right. Since these points represent
positions in space, they are essential
to the main message in these forms
of presentation: each point has a
very specific significance. We distin-
guish five position-related diagram-
matical elements. Commonly, the
most frequent forms of presentation
are coordinate systems with two
axes and geographical maps.

59
Points in space 2B.1
One coordinate axis Position

Home town Points on a coordinate axis with a scale

are a very precise method of repre­
senting data, because every value has
Persons
a clearly defined position. One disad-
5 vantage of this is that the quantity
4 Vienna cannot immediately be perceived as a
3 Paris value, but has to be deduced using
2 London its position. In addition, difficulties can
1 Munich
arise when values overlap.
0

60
Points in space 2B.2
Two coordinate axes Position

Home town A typical coordinate system consists

of two axes that open up space for
placing values with two dimensions at
an exact position. Compared to 2B.1,
there is less potential for overlapping
London because the expressions (e.g. towns
Munich and number of people) are distributed
Paris on their own axis. Two-dimensional
Vienna axes extending in both directions can
also open up a negative number space,
which creates a cross shape with four
0 1 2 3 4 5 Persons
fields.

61
Points in space 2B.3
Three coordinate axes Position

Home town Points in a three-dimensional space

Relocation require three axes, which often im­
pinges on the readability; sometimes it
is difficult to ascertain on which axis
Persons a point is located. In this case, the
4 interactive possibilities of a turnable
3 virtual space can provide a solution.
no
2

1 yes

0
London

Munich

Paris

Vienna

62
Points in space 2B.4
Polar axes Position

Home town In a polar coordinate system, each

point has a distinct position on a scale
in connection with a direction. In this
Vienna
way, it is possible — as in 2B.2 — to avoid
Persons overlapping and to make further state-
ments about a certain group
of data.
Munich Paris

London

63
Points in space 2B.5
Geographical coordinates Position

Home town Maps can be understood as points in

space with geographical coordinates.
The respective points of the coordi-
London nates do not always have to be depicted
in the form of an exact map, but can
Munich Vienna also be shown in a creative, mental grid
that indicates directions and relation-
Paris
ships. Countries do not always need to
be shown with complex borders, but
can also be depicted in a reduced form,
depending on the intended statement.

1 dot = 1 person

64
2C
Relationship

The representation of relation-

ships shows the connections
between several points following
a certain logic. Relationships
can be represented as networks,
hierarchically, and also nested.
In all cases, the specific content
determines the form of
representation.

67
Network 2C.1
Non-hierarchical Relationship

Family The network without a hierarchical

structure (direction) is the basic
diagram used to represent relation-
ships. It is based on the principle that
elements (nodes) can be connected
to each other by edges (connections).
A quantity can be represented using
the number of connections. Further-
more, the length of the connecting
line can be significant. Networks are
used in a wide range of fields; how­
ever, sometimes — beyond a certain
quantity of data — they are no longer
clearly readable.
1 dot = 1 person

68
Network 2C.2
Hierarchical Relationship

Family A hierarchical network is either nested

or features a relationship sequence.
The principle is very similar to that of
2C.1; the difference is, that here, there
is a higher-level structure, possibly
also a specific direction of reading.
A family tree is an example of a type of
diagram with a non-directional hier­
archy. When the lines indicate a direc-
tion (with an arrow), the hierarchy is
shown to also indicate a process.

1 dot = 1 person

69
Network 2C.3
Linear Relationship

Family A network presented in linear form

contains all points on an axis, similar to
a scale (2B.1). This opens up further
possibilities for using the space differ-
ently and reordering the connections
(nodes). Here too, it is possible to
create applications with a fixed hierar-
chy or sequence. Overlaying lines
indicate clusters.

1 dot = 1 person

70
Network 2C.4
Circular Relationship

Family A network represented in a circle shows

the relationships of a closed group. In
this case the points (nodes) are placed
on a circular line. Similar to the linear
network (2C.3), clusters of connecting
lines become apparent. In the case
of a directional representation, it is also
possible to represent two-way move-
ments in the system.

1 dot = 1 person

71
Nesting 2C.5
Intersection sets Relationship

Family It is also possible to represent relation-

ships without connecting lines in a
defined space. In accordance with the
Wagner Brown
basic rules of mathematical set theory,
it is possible to form intersection sets,
sub sets, and connection sets. In the
example, the different levels and over-
laps show specific information.

1 dot = 1 person

72
Nesting 2C.6
Stacking Relationship

Place of birth When relationships seem to indicate

Family a hierarchical structure, but have
no connecting lines (2C.2), stacking is
Paris Vienna London a suitable form of representation;
structures become clearly visible and
W W
have more detailed information at
B each level of this nesting.

1 dot = 1 person

73
Nesting 2C.7
Matrix Relationship

Place of birth A matrix is a table that shows two-

Family dimensional relationships. In this case,
the focus is on the correlation between
Paris Vienna London a certain column and a certain row.
It is also possible to work with quanti-
Wagner ties at the intersection.

Brown

1 dot = 1 person

74
 3C
Line

3B
Shape

3A
Color

3F
Isotype

3E
Contour
3D
Pattern
3
Visual
dimensions

The 40 elements of the visual dimensions resolve the question as

to how contents can be distinguished visually. We aim to demonstrate
that, in addition to color distinction, there are many other options
for translating a number or text in a visually attractive way that can
be clearly understood. This chapter uses findings from the fields
of graphic design, the psychology of perception, cartography, and
pictorial statistics.
The elements presented here are subdivided into the following
six groups: color, shape, line, pattern, contour, and the special group of
isotypes, according to Otto Neurath. We have analyzed these using
several publications on pictorial statistics by Neurath; they are visual­
ized here for the first time in an abstract shape that is generally
­applicable (3F).
The elements of the visual dimensions are particularly suitable
for distinguishing the expressions of the “Category” data type.
All those items of information that can be grouped together can be
categorized, for example, family names as a category with the two
expressions Wagner and Brown. Each individual element in this chap­
ter is symbolized by an icon; it is visualized using a basic example
from our data set and supported by a short explanation.

77
3A
Color

Color is one of the most frequently

used means for expressing distinc­
tions in diagrams. Every software
tool for data visualization uses color
as standard for distinguishing data
dimensions or value ranges. We are
listing here six possible visual ele­
ments from the group of colors. Some
principles are suitable for manual
sketching whereas others are only
useful in a digital context.

79
Basic colors 3A.1
Color

Family Basic colors are easy to use and

allow good differentiation. Having said
that, there are only nine colors that
allow good differentiation: yellow, or­
ange, red, pink, violet, blue, and green,
as well as brown and black. Color
provides a good option for different­
iating categories when sketching
visualizations.
Brown Wagner

1 dot = 1 person

80
Saturation 3A.2
with white Color

Family An alternative to using different colors

is to mix a main color with white. In
this way it is possible to create several
different gradations from one color.
It can also be aesthetically very pleas­
Wagner ing to work with fewer colors. This type
of color gradation is particularly suit­
able for gradations in time sequences,
for example light for newer and dark
for older values.

Brown

1 dot = 1 person

81
Saturation 3A.3
with black Color

Family Saturation with black refers to the

process of adding black to a color. Ex­
actly as in 3A.2, this saturation is
well suited to representing sequences.
Dark gradations are more likely to
be seen as dirty or as obsolete points
Wagner
in time. Colors particularly suitable
to gradation are shades of blue, green,
and gray, because the human eye
Brown is trained to perceive the sky, a forest,
and dusk.

1 dot = 1 person

82
Color gradient 3A.4
Color

Family and birth name Color gradients can be used to distin­

guish data expressions. In this case,
Durand– Wagner–
two or more colors are used to create a
Brown Moser gradual change of color. The effect
depends heavily on the contrast of the
selected colors. This becomes partic­
ularly interesting when the color mix is
Wagner
Brown
associated with a meaning.

Brown–
Wagner

1 dot = 1 person

83
Filled and empty 3A.5
Color

Family A filled shape makes a strong con­­trast

to an empty one. This can be per­
ceived as either full / empty or as
much / little. The differentiation effect
of this element can be multiplied by
com­bining with different shapes and
colors.
Brown Wagner

1 dot = 1 person

84
Transparency 3A.6
Color

Family The transparency of a shape does not

become visible until several shapes
are overlapped. It is primarily suitable
Durand– Wagner– for two-dimensional representations.
Brown Moser
Overlaying the shapes can denote
additional meaning.

Brown–
Wagner

1 dot = 1 person

85
3B
Shape

Besides color, form is one of the

most important visual differentia­tors.
Using forms in countable types of
diagrams is particularly useful be­
cause they can be comprehended very
quickly. Certain forms are also very
suitable for a quick analysis using
manual sketches. We have defined
eight possible visual elements in the
group of forms.

87
Geometrical 3B.1
Shape

Relocation There is a limited number of easily

distinguishable shapes that also work
visually in small formats. Shape as
a distinguishing characteristic is also
a very good tool for a quick hand
sketch.
yes no

1 shape = 1 person

88
Organic 3B.2
Shape

Relocation Different organic shapes can be

used to distinguish data expressions.
When using two or more organic
shapes, it is important to ensure that
there is a clear enough distinction
between them. Sometimes, abstract
yes no
organic shapes can also be perceived
as objects, for example, a flower.
A deliberate selection of shape is
therefore important.

1 shape = 1 person

89
Rotation 3B.3
Shape

Relocation Rotation is particularly suitable for

two expressions with the same shape.
The expression can then be multi­
plied by combining it with color. The
element of rotation can also be used
to turn a graphic element, for example
by 90 degrees, or to turn individual
no yes
graphic elements around a point.

1 shape = 1 person

90
Direction 3B.4
Shape

Relocation When the characteristic of a shape is

to point in a certain direction, it is
possible to work deliberately with this
element. For example, the triangle is
a sufficiently dominant shape that can
point in all four directions.

no yes

1 shape = 1 person

91
Stacking 3B.5
Shape

Relocation By stacking shapes it is possible to

visually distinguish two data expres­
sions from each other: differences
in quantity and differences in shape.
In the first case, the advantage is that
the compact representation makes
yes
it possible to count quantities quickly
whereas in the second case, a different
no
visual form is created, as shown in
the example on the left.

1 dot = 1 person

92
Blurring 3B.6
Shape

Relocation Blurring a geometric shape works

particularly well as a means of distin­
guishing when the same shape can
also be seen with sharp edges. Blurring
can be a good visual tool to indicate
uncertainty in data.

no yes

1 dot = 1 person

93
Transformation 3B.7
Shape

Relocation The transformation of a basic shape

creates one, or several, new shapes.
Transformation can be used to make a
shape appear three-dimen­sionally in
the space or to show ­movement.

no yes

1 shape = 1 person

94
Volume 3B.8
Shape

Relocation Each two-dimensional shape can also

be represented three-dimensionally
in the space. This makes it possible to
represent volumes or quantities. It is
important to keep an eye on the calcu­
lation to convert two-dimensional areas
to three-dimensional bodies in order to
yes no
avoid strong distortion in perception.

1 shape = 1 person

95
3C
Line

Lines in their various forms are often

underestimated as differentiators.
Besides their use in diagrams that are
based on the length of lines (2A),
such as bar charts, different types of
lines have great potential for rep­
resenting different relationships (2C).
A line that connects two items can
easily be enriched with additional
information. It is possible to distin­
guish three to a maximum of five
different types of lines. We are listing
here eight different types of lines
as visual dimensions.

97
Color 3C.1
Line

Age group The line color is a frequently used

and well-functioning distinguishing
characteristic of categories in line
lengths (2A) or relationship represen­
tations (2C). See also color (3A).

< 70 70–85 > 85

1 line = 1 person

98
Thickness 3C.2
Line

Age group The thickness of a line is a clear dis­

tinguishing characteristic provided that
there are not too many different thick­
nesses. Line thickness should not be
used for quantities, but for categories,
because different line lengths may
easily lead to confusion.

> 85

70–85
< 70

1 line = 1 person

99
Interruption 3C.3
Line

Age group When dashed lines are used to make a

distinction, it is important to ensure
that the individual segments and line
70–85 (4)
< 70 (3)
lengths are different enough. Dashed
lines can be used for category data
(different dashing) as well as for num­
> 85 (3) ber data (dash length /gaps in relation
to numbers).

7.5 mm = 1 person

100
Roughing 3C.4
Line

Age group A line can be modified to have a

rounded or jagged wavy shape. Each
wave frequency and shape can be
assigned to a different characteristic
expression of a category. It is also
possible to count the number of peaks
and troughs representing a unit.

< 70 (3) > 85 (3)

70–85 (4)

7.5 mm = 1 person

101
Details 3C.5
Line

Age group An often-used method in carto­

graphy is to enrich lines with various
details. These may indicate roads,
pathways, and railway tracks. When
< 70 (3)
creating one’s own lines with details
on maps, it is important to ensure
that no wrong associations are created
(for example, a state boundary). But
70–85 (4)
> 85 (3)
it is also possible to use lines with
details as a means of visual distinction,
independent of maps.

7.5 mm = 1 person

102
Organic 3C.6
Line

Age group By bending lines, numerous organic

shapes can be created. This type of line
can be used in support of a theme.
In comparison with roughing (3C.4), the
frequency of the wave is not ­relevant
< 70 (3)
> 85 (3) to the bending.

70–85 (4)

7.5 mm = 1 person

103
Detail 3C.7
Beginning or end Line

Age group Different details at the beginning or

end of a line offer a clear and simple
way to establish distinguishability
between data expressions. Here, it
is important that these details are
< 70 (3) big enough to be easily distinguished.

70–85 (4)

> 85 (3)

7.5 mm = 1 person

104
Arrow 3C.8
Line

Age group Nowadays, it is no longer possible

to imagine everyday life without arrows
at the end of a line; everybody knows
that this triangle at the end of a
line indicates a direction. Arrows are
< 70 (3) important for processes and other
directional diagrams and can also be
70–85 (4)
used effectively in combination with
> 85 (3)
non-directional lines.

7.5 mm = 1 person

105
3D
Pattern

Filling an area with a pattern can

save time when filling an area manu­
ally and it can also save a little bit
of ink in the printing process. Shapes
filled with patterns should be large
enough to allow clear differentiation.
It is important to note that, depend­
ing on their size and density, patterns
are perceived differently. On maps,
patterns are often used to represent
thematic levels. In our system we
define five different elements of this
group of patterns.

107
Texture 3D.1
Pattern

Generation When using patterns to fill an area, it

is best to choose a structure that is
easily distinguishable. In general, this
means that the elements have to
be a certain size.
1st (4)

3rd (2)

2nd (4)

108
Broken lines 3D.2
Pattern

Generation An area can be filled with lines of

equal thickness and equal spacing as
ap­ attern. The gaps in the line can
be used here as a distinguishing char­
3rd (2)
acteristic. This principle can also
1st (4)
be used to r­ epresent uncertainties in
a data set.

2nd (4)

109
Line thickness 3D.3
Pattern

Generation Line thickness in a pattern only works

well as an element when there are
not too many data expressions. Similar
to 3D.4, this element can be used
3rd (2)
for the visualization of density. Color
1st (4)
can be used for additional support.

2nd (4)

110
Density 3D.4
Pattern

Generation In contrast to the line pattern used to

fill an area, the pattern here does
not consist of a line but of dots or other
shapes. It is important that the basic
3rd (2)
pattern element is the same for all lev­
1st (4)
els of density so that it is possible to
perceive the difference between them.

2nd (4)

111
Direction 3D.5
Pattern

Generation The direction of a pattern is a relatively

weak differentiation. The element
can be created by mirroring the pattern
without changing the shape in any
other way. For example, this principle
1st (4)
3rd (2)
can be successfully used for large
areas in maps.

2nd (4)

112
3E
Contour

Contours can be used to add addi­

tional meaning to forms or to hold to­
gether groups of forms. Contours are
also sometimes referred to as peri­
meter lines. As is the case with differ­
ent types of lines, it is important to
note that not more than three to five
different types of contours can be
distinguished. We have identified five
different elements.

115
Color 3E.1
Contour

Place of birth The contour of a shape combined with

color is a very good distinguishing
characteristic. Colored contours are
a very quick method of producing
small visualizations by hand. Likewise,
filled two-dimensional shapes can
be given additional meaning by using
different contour colors.

London (2) Paris (4) Vienna (4)

116
Interruption 3E.2
Contour

Place of birth When using interruption in a contour it

is important to ensure that there is
an adequate difference in the gaps and
line lengths, just as with the interrup­
tion of lines in 3C.3. This particularly
applies to small sizes.

London (2) Paris (4) Vienna (4)

117
Thickness 3E.3
Contour

Place of birth The thickness of the contour refers to

the thickness of the line. Depending on
the size of the basic shape, not too
London (2)
many line thicknesses should be used
because it is hard to distinguish them
without magnification when the dis­
Paris (4)
tance between the shapes is too small.

Vienna (4)

118
Shape 3E.4
Contour

Place of birth The contour can have its own shape

and stand alone or form the edge of an
existing shape. Nesting a shape with a
London (2)
contour creates a new shape that can
easily be distinguished.

Paris (4)

Vienna (4)

119
Details 3E.5
Contour

Place of birth Details can be combined well with a

contour. This makes it possible to
create new and easily distinguishable
Vienna (4) contours. See also 3C.5.
Paris (4)

London (2)

120
3F
Isotype

At the beginning of the 20 th century, Otto Neurath was an important

Austrian scientist and pioneer of the so-called “Vienna Method of
Pictorial Statistics”. Together with graphic artist Gerd Arntz, he devel­
oped the International System of Typographic Picture Education
(Isotype), some of which is still in use today.
Because Neurath has been a big source of inspiration for our data
visualizations over the last 15 years, we want to dedicate the follow­
ing chapter to Isotype. Following our visual analysis of numerous
publications, particularly, the Gesellschafts- und Wirtschaftsatlas from
1931, we have defined eight principles in eight elements.
In contrast to the very abstract shapes in the previous chapters,
the more figurative shapes used in the Isotype system are its special
characteristic. We have, however, also worked out a few principles
in our analysis that will also work for abstract shapes.
We have already pointed out that figurative shapes trigger direct
associations in the viewer and that, sometimes, additional and in­
appropriate meanings are read into them. There is a big risk of manip­
ulation or that misunderstandings arise in the context of an unintend­
ed cultural context. We therefore advise not to work with figurative
shapes when the correct icons are not available or when an incorrect
interpretation could, and is likely to, occur.
The icons on the following pages have been taken from the visual
Isotype vocabulary used in the nineteen-thirties (except for one icon
representing women). We are conscious of the fact that these historic
images do not meet current requirements ­regarding gender equality
and inclusivity.

123
Figurative shape 3F.1
Isotype

Gender Stylized figures in the Isotype system

are a pictorial translation of concrete
terms (e.g. worker, chair). Nowadays,
the terms “icon” and “pictogram”
are more commonly used. With figura­
female tive shapes it is important that they
are shown to be of equal size and that
they are countable. When one icon is
male larger than the other one, there is a risk
of misinterpretation; in this case, the
larger figure could be interpreted as an
adult and the smaller one as a child.

124
Figurative shape 3F.2
Detail Isotype

Gender When using figurative shapes with

details, the basic shape should be
large enough to make it easy to distin­
guish the variations. The smaller the
icons are, the harder it is to recognize
female the details. The detail does not have
to be within the shape but can also be
placed next to the shape as a distin­
male guishing characteristic. It is important
to think clearly about how the new icon
combinations will be interpreted.

125
Direction 3F.3
Isotype

Relocation The direction in figurative shapes is

suitable for terms that describe a
movement. Here, it is important to en­
sure that the direction can be inferred,
for example, from the position of the
yes legs. Furthermore, when showing half
of a figure or figurative shape, it is
common sense to only make the split
no vertical.

126
Background 3F.4
Shaded Isotype

Relocation Shadows are suitable for both figura­

tive and abstract shapes. In the Iso­
type pictorial statistics they are used to
visually translate multiples of a quanti­
ty. Nevertheless, the effect of depth
yes created by shadow can also be used for
other subjects.

no

127
Background / Foreground 3F.5
Isotype

Relocation Simple abstract details behind or in

front of a shape are simple graphic
devices for emphasizing or highlighting
a certain statement. This element can
easily be combined with figurative
yes shapes as well as with abstract shapes.

no

128
Background 3F.6
Area Isotype

Place of birth A popular Isotype style device is to

hold several, often countable, shapes
together by means of a background
area. This can be an abstract design, in
one color, or with color gradients, or
yes even the flag of a country. It is also
possible to reverse the colors from the
foreground shape to the background
no area.

129
Frame or 3F.7
separating line Isotype

Relocation A closed line can be used as a frame

that holds a group of elements to­
gether. When the line is not closed but
only separates an area it is called
a separating line. In contrast to 3E,
Vienna here, the line is not used to delineate
the shape.
London

Paris

130
Background 3F.8
Figurative Isotype

Place of birth Besides a simple area (3F.6) and a

simple frame (3F.7), an illustrative
background is the last element of the
Isotype principles for distinguish­
Vienna ing groups. Illustrations can provide a
context for the information shown ­in
the foreground. However, it is important
Paris to ensure that the background of the
overall composition is not more visually
dominant than the foreground.

London

131
 4B
Grouping

4A
Sorting
4
Structuring
dimensions

In this chapter the diagrammatical and visual elements are combined

in a new graphic form with 15 structuring elements. This is where the
magic of design really happens. The arrangement in the object (4A
Sorting) and in the space (4B Grouping) defines the formative shape of
the visualization. Sorting means defining the data dimension used for
ordering a visualization. At the same time, this also determines the
shape of the diagram.
In order to avoid one single diagram containing too much data,
the diagram can be subdivided into several diagrams, a process called
grouping. Which type of sorting is best for readability and which
serves best for the purpose of visualization can only be found out by
trying and comparing. Each individual element in this chapter is
symbolized by an icon; it is visualized using a basic example from our
data set and supported by a short explanation.

133
4A
Structuring
by sorting

Sorting means making a conscious decision as to which data dimen-

sion should be used for ordering a visualization. This order can
then appear in linear or other shapes. In this way, different visual
representation shapes are created with the same data. Sorting
the data places a certain focus on the data set and thus controls the
story it is meant to tell. For example, in our data set the age, the
gender, but also the place of birth, can move to the foreground. All
data that does not have a unique position can be sorted (see the
position-related elements 2B).
Texts are always sorted alphabetically and numbers hierarchically.
The bar chart in the following icons are representative of all dia­
grammatical elements and can be replaced by them. (See also multidi-
mensional visualization examples.) We have defined nine possible
ways of sorting, which are explained and visualized on the next pages.
In chapter five we explain how sorting functions in practice.

135
Linear axis 4A.1
Sorting

Age Depending on the intended statement

and the space available, it is also
Otto (23)
possible to arrange the lines turned
Hermann Jr. (52)
by 90 degrees.
James (65)

Marie (72)

Marie Jr. (78)

Anna (83)

Emma (85)
Example: all ten persons are sorted by
age (ascending) on a linear axis (turned
Hermann (90)
by 90 degrees).
Paul (97)

Elisabeth (99)

1 dot = 10 years

136
Linear axis 4A.2
with line break Sorting

Age Sorting characteristics on a linear axis

)
52

with a line break may be used when

.(
Jr

5)

2)
n
3)

(6

only limited space is available. It is

(7
an
(2

es

ie
m
o

ar
er
tt

important to continue the sorting after

Ja

M
O

the line break along the line.

0)

)
78
(9

5)

3)

.(
n

(8
an

(8

Jr
a

ie
m

a
m

nn

ar
er

Em

M
A
H

Example: all ten persons are sorted

by age (ascending) along a line with a
line break.
9)
(9
h
7)

et
(9

ab
ul

is
Pa

El

137
Circular 4A.3
Parallel Sorting

Age group Using parallel circular lines as a means

of sorting is better suited to the com-
< 70 (3)
parison of data in percent than for the
70–85 (4)
representation of absolute numbers.
> 85 (3) This method is only suitable for quanti-
ties, not for two-dimensional areas.

Example: all ten persons are sorted by

three age categories (ascending) using
parallel circular lines.

60° = 1 person

138
Radially 4A.4
at an angle Sorting

9)

Age With radial sorting always using the

(9
th

same angle it is possible to produce

be
isa

7) star-shaped diagrams. When two-

El

l (9
P au dimensional areas are used instead of
)
(90
rm
ann lines there may be overlaps.
He

5)
E m m a (8

A n n a (8 3
)
Ma
Example: all ten persons sorted by age
O tto (2 3

M
rie
Jr. (
78)
(descending) in radial formation with
ar
ie the same angle.
He

(7
Ja

2)
rm

m
)

es
ann

(6
5)
Jr. (
52)

2.5 mm = 10 years

139
Radially 4A.5
Evenly distributed Sorting

Age Here, the difference to 4A.4 is that

Elisabeth (99)

the values are distributed using the

entire 360 degrees of the circle. To
7)
(9

make a precise comparison of the

ul
Pa

possible values, it may be helpful to

Ot

Her
to

man
n J r. show a scale in the background.
(2

(52) )
3)

(90
m ann
Her

es (
65) Emm Example: all ten persons sorted
Jam a (8
5) radially by age (descending) with even
Marie Jr. (78)

­distribution.
An
2)

na
(7

(8
ie

3)
ar
M

2.5 mm = 10 years

140
Radially 4A.6
Evenly distributed on shape Sorting

Age Starting the evenly distributed sorting

Eli

lines from a circle or other geomet­

sab

)
rical shapes creates space. In this way
eth

52
.(
Jr
it would also be possible to represent
(99

Pa n
3)

ul an
(97
)

O tto (2

) er
m two-dimensional areas or even, within
H
the circle, negative values.
5)
Ja m e s (6
n (9 0 )
H e rm a n
Ma Example: all ten persons sorted
rie
(72
) radially by age (ascending) with even
5)
(8 distribution, starting from a circle.
Ma

a
m
rie

Em
Anna

Jr.
(78
(8 3 )

1 dot = 10 years

141
Axis 4A.7
Free shape Sorting

(52)

Age The line of an axis does not always

n J r.

have to be straight. Sorted values

O tto (2 3)

man

can also be arranged along a curved

)
65
s(
Her

line.
me

2)
(7
Ja

8)
ie

7
r. (
ar
M

J
a rie
M
)
(83
na
An

Emma
(8 5 ) Example: all ten persons sorted by
age (ascending) along a curved line.
Her ma nn (90 )

Pa u l
(97)
Eli
sa
be
th
(99
2.5 mm = 10 years )

142
Linear axis 4A.8
Stacked Sorting

Home town Stacked sorting requires two data

Gender dimensions. In stacked diagrams
(e.g. stacked bar charts) the values
must be sorted and grouped at
Munich the same time.

London

Paris
Example: all ten persons sorted by
Vienna home town and stacked by gender.

1 dot = 1 person / filled dot = female

143
Free 4A.9
without axis or point Sorting

Age There are visualizations without an

3)
(8

axis or any sorting. Here, the percep-

a
nn
A

tion principle of proximity and dis-

H

Pa
er

tance applies, i.e. close elements are

m

ul
an

(9
n

7)
considered to be related.
(9
O

0)

)
tt

78
o

9)

.(
(2

(9

Jr
3)

ie
h
et

ar
ab

M
is
El

Example: all ten persons arranged

er
Ja

m
m

an

randomly in the space.

es

n
(6

Jr
5)

.(
52
2)

)
(7
5)
(8

ie
ar
a
m

M
Em

144
4B
Structuring
by grouping

Grouping is the second structure-giving dimension. In this process,

a multidimensional diagram is subdivided in accordance with a
data dimension and arranged in the space.
The two principles of grouping are repetition and juxtaposi-
tion. The advantage of grouping is based on the way humans recog-
nize patterns: our eyes are trained to visually perceive even very
small differences. Groups can be placed in the available space in a
grid or using some other principle.
In the following representations of elements all groups are
re­presented with identical sorting based on a linear axis (4A.1). The
elements of 4A.1 can be replaced by all elements in group 4A. We
have defined six possible principles of grouping.

147
Linear 4B.1
Grouping

Home town Linear grouping is generally only suit-

Age able when there are not many groups
Munich because the space in a row is often
very limited. It is also possible to turn
London the arrangement by 90 degrees.

Paris

Example: all ten persons sorted by

age (ascending) and grouped by their
four home towns along a line.
Vienna

1 dot = 10 years

148
Grid 4B.2
Grouping

Home town When groups are arranged in a grid,

Age the terms “Small multiples” or “Facet-
ting” are also used. With this princi-
ple it is possible to use the space in a
very efficient and structured way.
Munich London

Example: all ten persons sorted by

age (ascending) and grouped by their
four home towns in the grid.

Paris Vienna

2.5 mm = 10 years

149
On shape or angle 4B.3
Grouping

Home town When groups are arranged around

Age a circle or a different geometric shape
one needs to note that the radius
­increases as the number of groups
Lo
h
ic

nd
un

increases.
on
M

Example: all ten persons sorted by

age (ascending) and grouped by their
a

four home towns around a geometric

nn

Pa
ie

ris
V

shape.

150
Mirroring 4B.4
Grouping

Gender When using mirroring for grouping

Age it is important to ensure that the data
points opposite each other are com­
Otto (23) parable. This principle only works when
Marie (72) there are opposites, i.e. between two
Hermann Jr. (52) groups.
Marie Jr. (78)
James (65)
Anna (83)
Hermann (90)
Example: all ten persons sorted by
Emma (85) age and grouped mirrored by gender.
Paul (97)
Elisabeth (99)

2.5 mm = 10 years

151
Free in space 4B.5
Grouping

Home town When groups are placed freely in

Age space, the available format can be fully
utilized. The arrangement of groups
can be random or follow certain design
criteria.
Lond
ich
Mun on

Example: all ten persons sorted by

age (ascending) and grouped by their
four home towns freely in space.

Pa r i na
s V ie n

2.5 mm = 10 years

152
On levels 4B.6
Grouping

Gender When groups are placed on levels in

He Ot
Age rm
an
to
(23 space it is important to note that
)
nJ
r. (
5
overlaps may make it difficult to read
Jam 2)
He es
(65
the diagram.
rm
an )
n(
90
Pa )
ul
(97
)

Ma Example: all ten persons sorted by

rie
Ma
rie
(72
)
age (descending) and grouped by their
Jr.
An
(78 gender on two levels.
na )
Em (83
ma )
Eli (85
sa
be )
th
(99
) 1 dot = 10 years

153
5
The variety of
­multidimensional
­visualizations
Using 26 innovative diagram variations, this chapter shows the con-
nection and concrete application of all the elements and data
described in the previous chapters. This is just a small selection of
the numerous combinations that are possible with the ­Modular
Information Design (MID) system for the visualization of complexity.
All 80 elements are used at least once. In each visualization, a listing
of the elements and data dimensions used indicates, in a step-by-
step fashion, how this is structured.
The examples show that even with just a few data dimensions
and different elements, it is possible to achieve great design variety.
Hopefully this will motivate the reader to try to design new forms
of representation because the necessary attention relies only on
attractive and easy-to-understand visualizations.
To design is to think. That means trying different combinations
in order to form variations. Mapping (see instructions on p. 160)
selected elements on the selected data dimensions makes it easier
to understand the data set and it is possible to discover stories
that could not be detected from a simple table. Creating variations
repeatedly not only helps to view the existing data from a differ-
ent perspective, but also enables us to “think outside of the box”
when trying to solve a problem.

154
By varying selected data dimensions and elements in a deliberate
way, it is possible to create many different and new visualizations in
a quick and focused manner. We need to design and compare sever-
al variations in order to be able to assess which is the most suitable
for the selected data dimensions and which is the most readable.
Sometimes a small subject-related question helps to make a deci-
sion. For example: In year X, which country had the largest propor-
tion Y of women in group Z?
Multidimensional visualizations cannot be created using just
one correct software tool. It is always worth starting with pen
and paper. The next step may involve graphics software, an online
tool, or just a code. A selection of our favorites is included in the
Appendix. However, there is no such thing as the perfect tool.

155
Overview of all 80 elements

Diagrammatical 2A
dimensions Quantity

2B
Position

2C
Relation-
ship

Visual
dimensions

3A
Color

3B
Shape

156
3C
Line

3D
Pattern

3E
Contour

3F
Isotype

Structuring
dimensions

4A
Sorting

4B
Grouping

157
How the MID system
works

The MID system maps the elements On the one hand, MID can be used to
of the diagrammatical, visual, and encode new forms of representa-
structuring dimensions onto the data tion, i.e. diagram types; on the other
dimensions of the data cube. By hand, MID can also be used to de-
mapping the elements on the data code the structure of existing visual-
dimensions, a multidimensional izations. The example on the right-
visualization is created. This always hand side shows the mapping of five
requires, as a minimum, a diagram- elements on four data dimensions.
matical (blue), a visual (red), and Detailed instructions can be found
a structuring (green) element. On on the following page.
principle, only one element is mapped
on to the data cube for each perspec-
tive, i.e. the respective data dimen-
sion. An exception exists regarding
the elements of the structuring di-
mension of sorting (4A). These are
always tied to a blue or red element.

158
Mapping
(Encoding)

Element of
Element of visual dimensions
diagrammatic
dimensions

Element of
structuring
dimensions

Data dimensions

Decoding Multidimensional
variants

159
Mapping four data
­dimensions — instructions

Age

Gender

Family

Generation

E
The instructions show these
selected elements for a concrete
visual­ization. By replacing ­
elements it is possible to create
a large diversity of ­variations.
Here, four data dimensions
(yellow) and five elements of
the diagrammatical, visual, and
structuring dimensions are
combined. (A–E, blue, red, green).
Sorting (D) never has its own
data dimension, but always in­-
volves a connection with another
element.

160
A B 2 mm = 10 years

Wagner
Brown

Female
Male

Age → Family →
10 expressions 2 expressions

C D

Gender → Sorting by age,

2 expressions ascending

E Every visualization can be built up

step by step. The sequence in which
this is carried out is less import-
ant than actually carrying out the
steps. (The example only shows
one of many possible allocations of
the elements to the data dimen-
sions.) During the process of sketch-
ing, several steps are often carried
out simultaneously. However, it is
worth retaining individual steps, for
example the line length, to be used
at a later date.

Grouping by generation →
3 expressions

161
How sorting
works

For these visualizations, the same All ten family members are radially
three data dimensions (age, family, sorted with even distribution
and gender), as well as the same ­according to the data dimension
elements of the system (2A.1, 3A.2, selected.
3C.1, 4A.5), were used four times.
Therefore, the form of sorting A (Age)
(element 4A.5) is the same, only The sorting by age clearly indicates that the
the organization of the data differs: three people who died at the youngest age are
men.
in each case a different data dimen-
sion is in the foreground and this B (Family)
determines the sequence in which The sorting by family highlights the fact that the
three members of the Brown family are closer
the ten data points are shown. This together in age than the members of the Wag-
means that the star always looks ner family.
a little dif­ferent, and the story always
C (Gender, Age)
unfolds a little differently. In one The sorting by gender indicates that the differ-
case the age is in focus, in another ence in the age of the women is much smaller
case the family, and in two other than that of the men.

cases, the gender. D (Gender, Family, Age)

This sorting helps to compare the age, not only
within the gender, but also — at the same
time— within the families.

When sorting is by quantity (here, age), the

orga­nization is fixed. When sorting a data
dimension, for example, the category of gender,
it is mandatory to define whether to sort by
family or by age first.

162
A B 2 mm = 10 years

Wagner
Brown

Female
Male

Sorted by Sorted by
1. Age 1. Family
2. Age

C D

Sorted by Sorted by
1. Gender 1. Gender, 2. Family
2. Age 3. Age

163
Rays
Age as line length, sorted
radially at an angle and
grouped freely in space by
generation.

Marie Emma
James (72)
(65) (85)

Hermann
(90)

Anna
Paul
First (83)
(97)
generation

Elisabeth
Marie Jr. (99)
(78)

Hermann Jr.
(52)
Otto
(23)

Second
generation

Third
generation

In each generation the

number of women and men
is in balance.

164
 Example 1
Age

2A.1

6.25 mm = 10 years

Family Gender Home town

3C.1 3C.3 3C.7

Wagner Female London

Brown Male Munich
Paris Vienna

Generation

4A.4 4B.5

Sorted by Grouped by
age generation

Elements used
2A.1 Quantity. Line length
3C.1 Line. Color
3C.3 Line. Interruption
3C.7 Line. Detail. Beginning or end
4A.4 Sorting. Radially at an angle
4B.5 Grouping. Free in space

165
Semi-circles
Age by line length, sorted
parallel circular by age
and grouped mirrored by
relocation.

0 0

25 25

o
tt
O

Not
50 Relocated 50
relocated
He rm a
nn J
r.

Ja
m

Elisabeth
es

M
ar Pa u l
ie
J r.
Her m a n n
rie

a na a
M
75 Emm An 75

100 100

In the Wagner and Brown

families, the men are
more likely to relocate than
the women. People live
longer when they do not
relocate.

166
 Example 2
Age

2A.1

180° = 100 years

Family Gender Home town

3C.1 3C.3 3C.7

Wagner Female London

Brown Male Munich
Paris Vienna

Relocation

4A.3 4B.4

Sorted by Grouped by
age relocation

Elements used
2A.1 Quantity. Line length
3C.1 Line. Color
3C.3 Line. Interruption
3C.7 Line. Detail. Beginning or end
4A.3 Sorting. Circular. Parallel
4B.4 Grouping. Mirroring

167
Sunrise
Age as countable line length,
arranged in a radial layout
with equal spacing, based on
a free shape.

(99 abeth
n

Pa u l
an

(9 7 )
An na
Herm

Elis
)
(83 )
(90)
ma
Em )
(85
rie

J r.
Ma )
7 2

nn
(

(52 rma
)
He
Ot to
(23 )

Jr.
(7 arie
es
Ja 5)
m
(6

8)
M

On average, the people

from the first and second
ge­n­erations have lived
longer than those from the
third ­generation.

168
 Example 3
Age

2A.2

1 dot = 10 years
½ dot = 5 years

Family Gender Generation

3A.1 3A.2 3C.6

Wagner Female 1st

Brown Male 2nd
3rd

4A.7

Sorted by
generation and
age

Elements used
2A.2 Quantity. Line length. Countable
3A.1 Color. Basic colors
3A.2 Color. Saturation with white
3C.6 Line. Organic
4A.7 Sorting. Axis. Free shape

169
Tower
Age as line length with
line break, sorted by age,
stacked on linear axis
and grouped mirrored by
relocation.

Otto (23)
Hermann Jr. (52)

James (65)

Marie Jr. (78)

Relocated

Marie (72) Not

relocated

Anna (83)

Emma (85)

Hermann (90)

Paul (97)

Elisabeth (99)

On average, the people

who relocated lived to an
older age than those who
did not relocate.

170
 Example 4
Age

2A.3

20 mm = 10 years
1 line = 25 years

Family Gender Home town

3C.1 3C.3 3C.7

Wagner Female London

Brown Male Munich
Paris Vienna

Relocation

4A.8 4B.4

Sorted by Grouped by
age relocation

Elements used
2A.3 Quantity. Line length with line break
3C.1 Line. Color
3C.3 Line. Interruption
3C.7 Line. Detail. Beginning or end
4A.8 Sorting. Linear axis. Stacked
4B.4 Grouping. Mirroring

171
Domino
Age as countable line length
with line break, sorted by age
on linear axis with line break.

Otto
(23)
Hermann Jr.
(52)
James
(65)

Marie
(72)

Marie Jr.
(78)
Anna
(83)
Emma
(85)

Hermann
(90)

Paul
(97)
Elisabeth
(99)

The three people who

died at the youngest ages
were all men.

172
 Example 5
Age

2A.4

1 dot = 5 years
½ dot = 2 ½ years

Family Gender Relocation

3A.1 3E.2 3E.3

Wagner Female Yes

Brown Male No

4A.2

Sorted by
age

Elements used
2A.4 Quantity. Line length. Countable with
line break
3A.1 Color. Basic colors
3E.2 Contour. Interruption
3E.3 Contour. Thickness
4A.2 Sorting. Linear axis with line break

173
Firework
Age as a countable area,
sorted radially by age evenly
distributed and arranged
freely without axis or
point. Generation as area.

Otto
(23)
James
(65)

Marie
Hermann Jr. (72)
(52)

Marie Jr. Emma

(78) (85)

Anna
(83)

Hermann
(90)

Elisabeth
(99)
Paul
(97)

There are more Wagners

than Browns.

174
 Example 6
Generation Age

2A.5 2A.6

S = 1st 1 line = 5 years

M = 2nd
L = 3rd

Family Gender Relocation

3C.1 3C.7 3A.2

3F.5

Wagner Female Yes

Brown Male No

4A.4 4A.9

Age Sorted
free without axis
or point

Elements used
2A.5 Quantity. Area size
2A.6 Quantity. Area size. Countable
3C.1 Line. Color
3C.7 Line. Detail. Beginning or end
3F.5 Isotype. Background / Foreground
4A.4 Sorting. Radially at an angle
4A.9 Sorting. Free without axis or point

175
Swarm
Age as countable area,
sorted radially by age evenly
­distributed on a circle.

Otto
(23)

Elisabeth Hermann Jr.

(99) (52)

Paul James
(97) (65)

Hermann Marie
(90) (72)

Emma
(85) Marie Jr.
(78)
Anna
(83)

There are three pairs of

people of a similar age,
each belonging to the same
generation, but of different
gender.

176
 Example 7
Age

2A.6

1 shape = 1 year

Generation Family Gender

3B.3 3C.1 3C.4

0° = 1st Wagner Female

120° = 2nd Brown Male
240° = 3rd

4A.6

Sorted by
age

Elements used
2A.6 Quantity. Area size. Countable
3B.3 Shape. Rotation
3C.1 Line. Color
3C.4 Line. Roughing
4A.6 Sorting. Radially. Evenly distributed
on shape

177
Windmill
Age group as area,
100 percent as circle from
the center, sorted radi­
ally, evenly distributed by
age and grouped freely in
space by home town.

Paris Vienna

James Hermann Jr.

(65) (52)

Marie Paul Emma

(72) (97) (85)

Elisabeth
(99)

Hermann
(90)

Munich London

Otto
(23)

Marie Jr.
(78)

Anna
(83)

At each of the home towns

there is one member of the
Wagner family.

178
 Example 8
Age group

2A.7

S < 70
M = 70 – 85
L > 85

Family Relocation Gender

3A.1 3A.4 3D.5

Wagner Yes Female

Brown No Male

Home town

4A.5 4B.2

Sorted by Grouped by
age home town

Elements used
2A.7 Quantity. Area size 100% as a circle
from the center out
3A.1 Color. Basic colors
3A.4 Color. Color gradient
3D.5 Pattern. Direction
4A.5 Sorting. Radially. Evenly distributed
4B.2 Grouping. Grid

179
Origami
Age as area, 100 percent as
a circle from the edge, sorted
radially and evenly distrib­uted
on a circle by home town and
age, and grouped mirrored by
relocation.

London Munich
Anna Otto
(83) (23)
100

75

50

Marie Jr. 25
(78) Hermann Jr.
0 (52)
25

50

75

100

Elisabeth Emma
(99) (85)

Relocated

Marie Hermann
(72) (90)
Not
relocated

James Paul
Paris (65) (97) Vienna

Members of the Wagner

family only live in Munich
and Vienna.

180
 Example 9
Age

2A.8

36° = 1 person
Angle = Age

Family Gender Home town

3A.1 3D.1 3F.6

3F.7

Wagner Female Separated by

Brown Male home town

Relocation

4A.6 4B.4

Sorted by Grouped by
home town and age relocation

Elements used
2A.8 Quantity. Area size 100% as a circle
from the edge in
3A.1 Color. Basic colors
3D.1 Pattern. Texture
3F.6 Isotype. Background. Area
4A.6 Sorting. Radially. Evenly distributed
on shape
4B.4 Grouping. Mirroring

181
Pie
Age as area 100 percent in
the angle as circle, sorted
radially, evenly distributed by
age, and grouped linearly
by place of birth.

London Otto Vienna

Anna (23)
(83)
Emma
(85)

James
(65)

Hermann
(90)

Elisabeth
(99)

Paul
(97)

Marie Jr.
(78)

Hermann Jr.
(52)
Marie
Paris (72)

The sum of the age of all

people born in Vienna is
very similar to that of the
people born in Paris.

182
 Example 10
Age

2A.9

4.8° = 10 years

Family Generation Relocation Gender

3A.1 3A.3 3A.4 3D.4

Wagner 1st Yes Female

Brown 2nd No Male
3rd

Place of birth

4A.4 4B.1

Sorted by Grouped by
age place of birth

Elements used
2A.9 Quantity. Area size 100% in an angle
as a circle
3A.1 Color. Basic colors
3A.3 Color. Saturation with black
3A.4 Color. Color gradient
3D.4 Pattern. Density
4A.4 Sorting. Radially at an angle
4B.1 Grouping. Linear

183
Fan
Age as two-dimensional area
100 percent in the angle as
a circle segment, sorted radially
in the angle by family and
gener­ation, grouped by place
of birth.

Emma
(85)
Hermann
(90)

Otto
(23)

Paul
(97)

Anna
(83)

Vienna

James London
(65)

Paris

Elisabeth
(99)

Hermann Jr.
Marie (52)
(72) Marie Jr.
(78)

The first two generations of

the Wagner family ­account
for half of all the people and
most of them belong to the
oldest group.

184
 Example 11
Age

2A.10

2.4° = 10 years

Family Generation Gender

3A.1 3A.3 3D.4

Wagner 1st Female

Brown 2nd Male
3rd

Place of birth

4A.4 4B.3

Sorted by Grouped by
family and place of birth
generation

Elements used
2A.10 Quantity. Area size 100% in an angle
as a circle segment
3A.1 Color. Basic colors
3A.3 Color. Saturation with black
3D.4 Pattern. Density
4A.4 Sorting. Radially at an angle
4B.3 Grouping. On shape or angle

185
Ring
Age as area 100 percent in
the angle as ring, sorted by
age.

Otto
(23)
Elisabeth Hermann Jr.
(99) (52)

James
(65)

Paul
(97)

Marie
(72)

Hermann
(90) Marie Jr.
(78)

Emma Anna
(85) (83)

The Browns only feature in

the first two generations.

186
 Example 12
Age Age

2A.11 2A.13

4.8° = 10 Jahre 1 line = 3 years

Family Generation Gender

3E.1 3E.4 3F.2

Wagner 1st Female

Brown 2nd Male
3rd

4A.6

Sorted by
age

Elements used
2A.11 Quantity. Area size 100% in an angle
as a ring
2A.13 Quantity. Area size 100% as a countable
rectangle
3E.1 Contour. Color
3E.4 Contour. Shape
3F.2 Isotype. Figurative Shape. Detail
4A.6 Sorting. Radially. Evenly distributed on shape

187
Parcels
Age as area 100 percent
distributed by percentage.

Otto James
Hermann Paul
(23) (65)
(90) (97)

Hermann Jr.
(52)

Marie
(72)

Marie Jr.
Emma Elisabeth
(78)
(85) (99)

Anna
(83)

Those who relocated had

a shorter life than those who
did not relocate.

188
 Example 13
Age

2A.12

Total area =
110 × 110 mm =
744 years

Family Relocation Gender First name

3A.1 3B.8 3D.1 3F.7

Wagner Yes Female Separated by

Brown No Male person

4A.1

Sorted by
family

Elements used
2A.12 Quantity. Area size 100% as a rectangle
3A.1 Color. Basic colors
3B.8 Shape. Volume
3D.1 Pattern. Texture
3F.7 Isotype. Frame or separating line
4A.1 Sorting. Linear axis

189
Display
Age as countable area
100 percent, sorted linearly
by family.

Otto
(23)

James
(65)

Hermann Jr. Hermann

(52) (90)

Marie Paul
(72) (97)

Marie Jr.
(78)

Anna Elisabeth
(83) (99)

Emma
(85)

The number of people in

the Brown family account
for 30 percent and their
total age for 40 percent.

190
 Example 14
Age

2A.13

1 dot = 7.5 years

Family Generation Gender First name

3A.1 3A.4 3B.2 3F.7

Wagner 1st Female Separated by

Brown 2nd Male person
3rd

4A.1

Sorted by
family

Elements used
2A.13 Quantity. Area size 100% as a
countable rectangle
3A.1 Color. Basic colors
3A.4 Color. Color gradient
3B.2 Shape. Organic
3F.7 Isotype. Frame or separating line
4A.1 Sorting. Linear axis

191
Pyramid
Positioned by birth year on
a coordinate axis. Age as line
length.

1950

Marie Jr. (78)

1940
Third
generation
Hermann Jr. (52)

1930

1920

Elisabeth (99)
Paul (97)
Anna (83)
1910

Second
generation

Otto (23)
1900

1890

James (65)
1880
Marie (72) First
generation
Emma (85)

Hermann (90) 1870

The people in the first

generation are closest to
each other in age (of death).

192
 Example 15
Age Year of birth

2A.1 2B.1

6.25 mm = 10 years Y = Decades

Family Gender Generation

3C.1 3C.3 3F.6

Wagner Female Separated by

Brown Male generation

Elements used
2A.1 Quantity. Line length
2B.1 Position. Points in space. One coordinate axis
3C.1 Line. Color
3C.3 Line. Interruption
3F.6 Isotype. Background. Area

193
Layers
Positioned by age and year of
birth /death between
two coordinate axes. Age
as line length.

100 Elisabeth (99)

Paul (97)

90 Hermann (90)
Emma (85)
Anna (83)
80
Marie Jr. (78)

Marie (72)
70

James (65)

60
First
generation

Hermann Jr. (52)

50

Third
generation
40

30

Otto (23)
20

Second
generation
10

1870 1880 1890 1900 1910 1920 1930 1940 1950 1960 1970 1980 1990 2000 2010 2020

The person who died first

and the two people who
lived the longest all belong
to the second generation.

194
 Example 16
Year of birth
Year of death
2B.2

X = Decades
Y = Age

Family Gender Generation

3C.1 3C.3 3F.7

Wagner Female Separated by

Brown Male generation

Elements used
2B.2 Position. Points in space.
Two coordinate axes
3C.1 Line. Color
3C.3 Line. Interruption
3F.7 Isotype. Frame or separating line

195
Flags
Age as line length, year
of birth /death and relocation
positioned between three
coordinate axes and sorted
by relocation and age.

2020

2000

1980

1960

1940

1920

1900

1880

Relocated
1860

Marie Anna Emma Hermann Paul Elisabeth Otto Hermann Jr. James Marie Jr.
(72) (83) (85) (90) (97) (99) (23) (52) (65) (78)

Not
relocated

Nobody from the Brown

family has lived in the
21 st century.

196
Year of birth Example 17
Year of death
Relocation
2B.3

X = Person
Y = Birth / death
Z = Relocation

Family Relocation Gender

3A.1 3A.4 3D.1

Wagner Yes Female

Brown No Male

4A.1

Sorted by
relocation and age

Elements used
2B.3 Position. Points in space. Three
coordinate axes
3A.1 Color. Basic colors
3A.4 Color. Color gradient
3D.1 Pattern. Texture
4A.1 Sorting. Linear axis

197
Target
Age as line length, ­p ositioned
on polar axes
by year of birth /death and
sorted by age.

Elisabeth Otto
(99) 2020 (23)

2000

1980

Paul Hermann Jr.

(97) 1960 (52)

1940

1920

1900

1880

Hermann James
(90) (65)

Emma Marie
(85) (72)

Anna Marie Jr.

(83) (78)

The four people with the

longest life are members of
the Wagner family.

198
 Example 18
Year of birth
Year of death
2B.4

Axis = Decades
Sector = Person

Family Gender Relocation

3A.1 3D.1 3E.3

Wagner Female Yes

Brown Male No

4A.5

Sorted by
age

Elements used
2B.4 Position. Points in space. Polar axes
3A.1 Color. Basic colors
3D.1 Pattern. Texture
3E.3 Contour. Thickness
4A.5 Sorting. Radially. Evenly distributed

199
Movement chart
Place of birth and home town
positioned as geographic
coordinates and set in rela-
tion to each other as a net-
work without hierarchy.

Marie Jr.
Anna

Elisabeth
James
Marie

London

Hermann Jr.

Hermann
Otto

Emma

Paul
Munich

Paris

Vienna

Paris has the most

relocations.

200
 Example 19
Home town Relocation
Place of birth
2B.5 2C.1

Family Home town Relocation Gender

Place of birth
3A.1 3A.5 3C.8 3F.2

Wagner Home town Yes Female

Brown Place of birth Male

Elements used
2B.5 Position. Points in space. Geographical
coordinates
2C.1 Relationship. Network. Non-hierarchical
3A.1 Color. Basic colors
3A.5 Color. Filled and empty
3C.8 Line. Arrow
3F.2 Isotype. Figurative Shape. Detail

201
Lollipop
Home towns and family
members set in relation to
each other as a network
without hierarchy.

Emma Hermann
Hermann Jr.
Paul

Elisabeth

Otto Marie

James

Anna
Marie Jr.

Overall, four people

relo­cated; the majority of
these were male members
of the Wagner family.

202
 Example 20
Home town

2C.1

Family Home town Relocation Gender

3A.1 3B.1 3B.7 3D.2

Wagner London Yes Female

Brown Munich No Male
Paris Vienna

Home town

4A.4 4B.3

Sorted by Grouped by
family home town

Elements used
2C.1 Relationship. Network. Non-hierarchical
3A.1 Color. Basic colors
3B.1 Shape. Geometrical
3B.7 Shape. Transformation
3D.2 Pattern. Broken lines
4A.4 Sorting. Radially at an angle
4B.3 Grouping. On shape or angle

203
Family tree
Family, home town, and place of
birth in relation to each other as
a hierarchical network.

Home town

Place
of birth
Hermann Jr.

Emma

Hermann

Paul

Otto

Marie Jr.

Elisabeth

James

Marie

Anna

Elisabeth was born as a

member of the Brown family
but adopted the Wagner
name through her marriage.

204
 Example 21
Place of birth
Home town
2C.2

Family Place of birth Gender Relocation

Home town
3A.1 3B.1 3F.1 3F.8

Wagner London Female Yes

Brown Munich Male
Paris Vienna

Family
Home Town
4A.1 4B.1

Sorted by Grouped by
age family and
home town

Elements used
2C.2 Relationship. Network. Hierarchical
3A.1 Color. Basic colors
3B.1 Shape. Geometrical
3F.1 Isotype. Figurative shape
3F.8 Isotype. Background. Figurative
4A.1 Sorting. Linear axis
4B.1 Grouping. Linear

205
Chain
Family set in relation as linear
network and sorted linearly
by year of birth. Age group as
area.

Hermann

Emma

Marie

James

Otto

Anna

Paul

Elisabeth

Hermann Jr.

Marie Jr.

In the Wagner family,

the birth of a female member
was always preceded by
the birth of a male member.

206
 Example 22
Age group Family

2A.5 2C.3

S < 70
M = 70 – 85
L > 85

Family Gender Relocation

3A.1 3B.6 3F.5

Wagner Female Yes

Brown Male No

4A.1

Sorted by
year of birth

Elements used
2A.5 Quantity. Area size
2C.3 Relationship. Network. Linear
3A.1 Color. Basic colors
3B.6 Shape. Blurring
3F.5 Isotype. Background / Foreground
4A.1 Sorting. Linear axis

207
Skyline
Relocation and home town
set in relation to age as a
circular network. Age as line
length.

Herm
Emma a
(85) (90) nn Pa
a
Ann (97 ul
(83) )

El
Marie Jr.

is (99
ab )
(78)

et
h
Marie
(72)

James
(65)

Hermann Jr.
(52)

Otto
(23)

Only Wagners lived in

Vienna and Munich.

208
 Example 23
Age Home town

2A.1 2C.4

6.25 mm = 10 years

Family Home town Gender Relocation

3A.1 3B.1 3C.2 3C.5

Wagner London Female Yes

Brown Munich Male No
Paris Vienna

4A.6

Sorted by
age

Elements used
2A.1 Quantity. Line length
2C.4 Relationship. Network. Circular
3A.1 Color. Basic colors
3B.1 Shape. Geometrical
3C.2 Line. Thickness
3C.5 Line. Details
4A.6 Sorting. Radially. Evenly distributed on form

209
Intersection
Place of birth and home
town set in relation to each
other as a nested section.
Age as area.

Vienna Paris

Emma Elisabeth
(85) (99)
Marie
(72)

Hermann Jr.
Hermann (52)
(90)
Paul
(97)
James
(65)

Marie Jr.
Otto (78)
(23)

Anna
(83)
Munich

London

Two Wagners relocated

from Paris, but one Brown
relocated to Paris.

210
 Example 24
Age group Place of birth
Home town
2A.5 2C.5

S < 70 Overlap =
M = 70 – 85 relocation
L > 85

Family Relocation Generation Gender

3A.1 3A.6 3B.4 3D.3

Wagner Overlap = 1st Female

Brown relocation 2nd Male
3rd

4A.9

Sorted
free in space

Elements used
2A.5 Quantity. Area size
2C.5 Relationship. Nesting. Intersection sets
3A.1 Color. Basic colors
3A.6 Color. Transparency
3B.4 Shape. Direction
3D.3 Pattern. Line thickness
4A.9 Sorting. Free without axis or point

211
Cells
Place of birth and home
town set in relation to each
other as nested stacks
and grouped on levels by
place of birth.

Hermann
Paul
Emma

Otto

Vienna Elisabeth
Marie

Hermann Jr.

Marie Jr.

Paris
Anna

James

London

Half of all people living in

Paris and London relocated,
whereas only a quarter of
those living in Vienna did so.

212
 Example 25
Place of birth
Home town
2C.6

Outside: Place of
birth
Inside: Home town

Family Gender Age group Place of birth

Home town
3A.1 3A.2 3B.5 3E.5

Wagner Female 2: < 70 London

Brown Male 4: 70 – 85 Munich
6: > 85 Paris Vienna

Place of birth

4A.9 4B.6

Sorted Grouped by
free in space place of birth

Elements used
2C.6 Relationship. Nesting. Stacking
3A.1 Color. Basic colors
3A.2 Color. Saturation with white
3B.5 Shape. Stacking
3E.5 Contour. Details
4A.9 Sorting. Free without axis or point
4B.6 Grouping. On levels

213
Cluster
Generation and home
town set in relation to each
other as a nested matrix
and sorted by generation and
home town.

Munich London Paris Vienna

First
generation

James Marie Hermann Emma

Second
generation

Otto Anna Elisabeth Paul

Third
generation

Marie Jr. Hermann Jr.

Mobility increases with

each generation.

214
Home town Example 26
Generation

2C.7

Family Relocation Gender

3A.1 3F.3 3F.4

Wagner Left = Yes Female

Brown Right = No Male

Elements used
2C.7 Relationship. Nesting. Matrix
3A.1 Color. Basic colors
3F.3 Isotype. Direction
3F.4 Isotype. Background. Shaded

215
6
Appendix

217
The authors
Lucerne School of
Art and Design

FHNW Academy of
Pioneer of DX — Data Art and Design Basel
Experience Design
Bern Academy of
the Arts
Over 100 projects for
customers such as Roche, University of
Schweizer Post, non- Initiator of the event Applied Sciences of
governmental organizati- series “On Data the Grisons
ons such as UNICEF, And Design” Basel —
the UN, and government Berlin — Vienna Teaching and
institutions such as the mentoring
Swiss Statistics Office
and the Swiss National
Science Foundation.

Superdot.studio

Information is Beautiful
Award

DPA Infographics Award

Out of Balance
Kritik der Gegenwart
Information design
winner

YAAY.ch

Training Career Highlight

218
Darjan Nicole
Hil Lachenmeier

FHNW Academy of
University of Vienna
Art and Design Basel

BSc Business
Informatics BA Visual
Communication
MSc Business (i.a. with Wolfgang
Informatics Weingart)

The most beautiful

Oracle / IBM
Swiss books

Certified Systemic
Coach
Research at FHNW
Academy of Art and
Credit Suisse Design Basel

Studio Niclac
MA Visual
Communication and
Iconic Research
Basel
Research at FHNW
Academy of Art and
Design Basel

Research at FHNW
School of Business
Basel

219
Further reading, sources,
and tools
Books that have accompanied us
along the way and that we recom-
mend as further complementary
reading to this book.

Swiss graphic Information Design 2 Wurman, R. S. (2017). ‘Under-

design Journal standing Understanding’.
Richard Saul Wurman & Jack.
Maier, M. (1986). ‘Elementare R. Klanten, N. Bourquin, et al. New York City.
Entwurfs- und Gestaltungs­ (2008). ‘Data Flow: Visualiz-
prozesse: Die Grundkurse an ing Information in Graphic
der Kunstgewerbeschule Design’ (1st Ed.). Berlin. 1 Otto Neurath
Basel’. Bern. and Jacques Bertin
Schwochow, J. (2020).
Müller-Brockmann, J. (2009). ‘Die Welt verstehen mit 264 Bertin, J. (2011). ‘Semiology of
‘Grid systems in Graphic Infografiken: aus Geschichte, Graphics’ (1st Ed.). Madison.
Design: A Visual Commu­ Wirtschaft, Gesellschaft,
nication Manual for Graphic Sport, Kunst und Kultur, Wis- Brian, S., Neurath, M. &
Designers, Typographers senschaft und Technologie’. Kinross, R. (2016). ‘Die Trans-
and Three Dimensional De- Munich. formierer: Entstehung
signers’ (4th Rev. Ed.). und Prinzipien von Isotype’
Salenstein. Wiedemann, J. & Rendgen, S. (1st Ed.). Salenstein.
(2018). ‘Understanding the
Osterer, H. & Stamm, P. World. The Atlas of Infograph- Hartmann, F. & Bauer, E. K.
(2008). ‘Adrian Frutiger — ics’ (Multilingual ed.). (2006). ‘Bildersprache: Otto
Typefaces: The Complete Cologne. Neurath Visualisierungen’
Works’ (1st Ed.). Basel. (2nd ext. Ed.). Vienna.
Wiedemann, J. & Rendgen, S.
(2019). ‘History of Information Vossoughian, N. (2011). ‘Otto
Visual perception Graphics: HISTORY OF INFO- Neurath: The Language
GRAPHICS’ (Multilingual ed.). of the Global Polis’ (1st Ed.).
Dondis, D. A. (1974). Cologne. Rotterdam.
‘A Primer of Visual Literacy’
(New Ed.). Cambridge,
Massachusetts. Information Theory of information
architecture: the design
Leborg, C. (2006). L.A.T.C.H theory
‘Visual Grammar (Design Tufte, E. R. (1990). ‘Envi­sioning
Briefs)’ (1st Ed.). New York Bradford, P. & 2 Wurman, Information’ (1st Ed.).
City. R. S. (1996). ‘Information Cheshire, Connecticut.
Architects’. New York City.
Zwimpfer, M. (1994). ‘2d — Tufte, E. R. (1997). ‘Visual
visuelle Wahrnehmung’. 2 Wurman, R. S. (1989). Explanations: Images and
Salenstein. ‘Information Anxiety’ (1st Ed.). Quantities, Evidence and
Doubleday. New York City. Narrative’. Cheshire,
­C onnecticut.

220
Tufte, E. R. (2001). ‘The Recommended tools
Visual Display of Quantitative for the visualization
Information’ (2nd Ed.). process
Cheshire, Connecticut.
For the analysis of data:
Tufte, E. R. (2006). ‘Beautiful Excel and Tableau; for the
Evidence’. Cheshire, implementation: Illustrator,
Connecticut. InDesign, RAW Graphs,
Processing / p5.js, SVG with
Vanilla JS, D3.js
Information design
manuals
Kirk, A. (2019). ‘Data Visuali-
sation: A Handbook for
Data Driven Design’ (Reprint
Ed.). New York City.

Koponen, J. & Hildén, J. (2018).

‘Data Visualization Handbook’.
Helsinki.

Lima, M. (2013). ‘Visual

Complexity: Mapping Patterns
of Information (History of
Information and Data
Visualization and Guide to
Today’s Innovative Appli­
cations)’ (Reprint Ed.).
New York City.

Lupi, G. & Posavec, S. (2018).

‘Observe, Collect, Draw!:
A Visual Journal’ (1st Ed.).
New York City.

Schwabish, J. (2021). ‘Better

Data Visualizations: A Guide
for Scholars, Researchers, and
Wonks’. New York City.

221
Acknowledgments

No book is ever written overnight, and

it is never the work of just one person.
Unfortunately, there are not enough
lines to mention all of our friends and
family members who were directly
or indirectly involved in this book, even
if only through motivating words or
unconscious influence. However, we
want to express our particular thanks
to two people from our Superdot team
who have given us many ideas and
much support — Sarah Heitz and Aaron
Ritschard; without your trust, your
patience, and your contribution, we
would not have managed. Many thanks
also to Birkhäuser Publishers for
­trusting in our work.

# visualizingcomplexity
# modularinformationdesign

222
Publication details

Concept and content right. All rights are reserved,

Darjan Hil, Nicole whether the whole or part
Lachenmeier of the material is concerned,
specifically the rights of
Layout and design translation, reprinting, re-use
Nicole Lachenmeier of illustrations, recitation,
broadcasting, reproduction on
Typesetting microfilms or in other ways,
Superdot • visualizing and storage in databases. For
complexity any kind of use, permission
of the copyright owner must
Translation from be obtained.
German into English
Hartwin Busch ISBN
978-3-0356-2504-2
Copyediting e-ISBN (PDF)
Patricia Kot 978-3-0356-2506-6
German Print-ISBN
Project management 978-3-0356-2505-9
Baharak Tajbakhsh,
Freya Mohr © 2022 Birkhäuser Verlag
GmbH, Basel
Production P.O. Box 44, 4009 Basel,
Amelie Solbrig, Switzerland
Heike Strempel Part of Walter de Gruyter
GmbH, Berlin / Boston
Paper
Juwel Offset 9 8 7 6 5 4 3 2 1
www.birkhauser.com
Printing
Eberl & Kœsel GmbH &
Co. KG
­
Library of Congress Control
Number: 2021953442

Bibliographic information pub-

lished 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 copy-

223