OMG Technical Conference: Full Day Tutorial
The Physics† of Notations: Designing
Diagramming Notations That Work
Presenter: Dr Daniel Moody
Director, Ozemantics Pty Ltd, Sydney, Australia
D
iagramming notations form an integral
part of the “language” of IT practice
and have done since the earliest beginnings of the field. They are used at all levels
of IT practice, from strategic planning down to
integrated circuit design. It is hard to think of
any area of IT practice where diagramming
notations don’t play a central role: for example, UML in software engineering, BPMN in
business process management, PERT charts in
project management. They play a particularly
critical role in communicating with business
stakeholders (end users and customers) because of their ability to present complex ideas
in a simple way.
The first IT diagramming notation and ancestor of
all modern IT diagramming notations (c. 1947)
The diagrammatic representation of a modelling notation (its visual syntax) is arguably its
most important characteristic, as this is what
notation users and their customers see and directly interact with. It performs a similar role
to a graphical user interface for a software system and has a profound effect on the usability
and effectiveness of notations. Human information processing is highly sensitive to the
exact form information is presented: apparently minor changes in visual appearance,
such as the use of colour, can have dramatic
effects on understanding and problem solving
performance. For this reason, decisions about
the visual representation of notations should
be treated with as much care as decisions
about their content (i.e. choice of concepts).
Currently, IT diagramming notations are designed in a way that has more in common with
black magic than reasoned thought. Symbols
are defined without any explanation as to why
they were chosen or the alternatives considered: the reasons for choosing particular symbols are generally shrouded in mystery. There
is also a lack of explicit principles for designing diagramming notations, with the result
that notation designers have to fall back on intuition and common sense, which is unreliable: the effects of graphic design choices are
often counterintuitive and our instincts can
lead us wildly astray.
As a result, most IT diagramming notations
violate some of the most basic principles about
how the human visual system works and often
act as a barrier rather than an aid to communication, especially with business stakeholders. Also, by using only a limited repertoire of graphical techniques, they fail to exploit the potential power of diagrams. Some of
the most powerful graphical techniques (e.g.
spatial location, colour) are rarely, if ever,
used in IT diagramming notations.
Diagramming notations play a central role in all
engineering and design disciplines, but currently
we lack sound principles for designing them
The goal of this tutorial is to establish a scientific basis for diagramming notation design: to
help it progress from an art into a science. It
defines a set of principles for designing cogni-
† It is called the “physics” of notations because it focuses on the physical (perceptual) properties of notations rather than their logical (semantic) properties.
Page 1
tively effective diagramming notations: notations that are optimised for processing by the
human mind. Importantly, the design principles are evidence-based: they are not based on
common sense, experience or opinion but on
theory and empirical evidence about how our
visual systems work. Together they provide a
scientific basis for constructing diagramming
notations, which has previously been lacking
in the IT field.
Learning objectives
At the end of the tutorial, participants will be
able to:
 Design diagramming notations in a systematic, evidence-based manner.
 Justify choice of symbols with reference to
known principles about how our visual
systems work.
 Conduct studies to evaluate the usability
and effectiveness of notations.
Presenter background
Perceptual
Discriminability
Graphic
Economy
Cognitive
Fit
Semiotic
Clarity
Cognitive
Integration
Semantic
Transparency
Visual
Expressiveness
Manageable
Complexity
Dual
Coding
The Physics of Notations: evidence-based
principles for designing cognitively effective
diagramming notations
The principles have been successfully used to
evaluate and improve several modelling notations as well as design notations from first
principles. They have recently been proposed
as an international standard for designing diagramming notations across engineering disciplines, so could have implications beyond the
IT field.
The tutorial challenges some longstanding
assumptions about how diagramming notations should be designed and how they have
been since the earliest beginnings of the IT
field. It identifies serious design flaws in some
of the leading notations used in IT practice
(e.g. UML, BPMN), together with some simple
and practical ways of improving them. It also
defines a way of measuring the effectiveness of
diagramming notations and for testing them
prior to their release (analogous to user acceptance testing for software systems).
Daniel Moody is a Director of Ozemantics, a
Sydney-based information management consultancy firm. He is recognised as one of Australia’s leading experts in data modelling and
information management and has an international reputation in these fields. He holds a
doctorate in Information Systems from the
University of Melbourne and has held senior
positions in some of Australia’s leading corporations and consultancy firms. He has conducted consulting assignments in 12 different
countries, covering a broad range of industries. He has also published over 100 scientific
papers, been a keynote speaker 9 times and
chaired several international conferences.
He was the inaugural President of the Australian Data Management Association (DAMA),
former Vice-President on the
DAMA International Board
and is listed in Who's Who in
Science and Engineering. He
has lived in 8 different countries, speaks fluent English
and can say “hello”, “thank
you” and “cheers” in at least
10 different languages.
Structure and content
The structure of the tutorial is summarised in
the mind map below:
Delivery format
The tutorial will be delivered in an interactive
manner, with an emphasis on learning by doing. A range of practical exercises and examples are used to illustrate the principles and
give participants practice in applying them.
Examples and exercises feature some of the
leading diagramming notations used in IT
practice (e.g. UML, BPMN) as well as visual
notations from other disciplines.
Mindmap of tutorial content
Page 2
1. What is a “good” diagramming
notation?
This looks at how diagramming notations are
used in IT practice and why. It defines what a
“good” diagramming notation is (the design
goal) and how to measure this.
2. The Art of Diagramming Notation
Design: Current Practice
This looks at current practice in designing IT
diagramming notations. It analyses some of
the leading diagramming notations used in the
IT field and practices of the leading notation
designers (i.e. what the experts do). The conclusion is that radical change is needed to current design practices to produce effective diagramming notations.
3. The Theory of Diagramming Notation
Design: How Diagramming Notations
Communicate
This explains how and why diagramming notations communicate, with reference to theories of communication, graphic design, visual
perception and cognition. Only by understanding how diagramming notations communicate can we improve their ability to
communicate. The theory also enables us to
explain and predict why some diagramming
notations are more effective than others.
4. The Science of Diagramming Notation
Design: Principles for Effective
Diagramming Notations
This is the main practical content of the tutorial, and describes 9 principles for producing
effective diagramming notations:
 Principle of Semiotic Clarity: there should
be a one to one correspondence between
concepts and graphical symbols
 Principle of Perceptual Discriminability:
symbols should be clearly distinguishable
from one another
 Principle of Semantic Transparency: use
symbols whose appearance suggests their
meaning
 Principle of Complexity Management: include explicit mechanisms for dealing with
complexity
 Principle of Cognitive Integration: include
explicit mechanisms for integrating separate diagrams together
 Principle of Visual Expressiveness: use the
full range of visual variables (fully utilise
the graphic design space)
 Principle of Dual Coding: use text to rein-
force and complement graphics
 Principle of Graphic Economy: the num-
ber of graphical symbols should be cognitively manageable
 Principle of Cognitive Fit: use different
visual representations for different tasks
and audiences (visual horses for cognitive
courses)
 Trade-offs and synergies: understanding
interactions among principles
5. Conclusion: A Manifesto for
Diagramming Notation Design
This reviews and summarises all the material
covered and concludes with a “manifesto” for
designing effective diagramming notations.
Intended audience
The tutorial is aimed at:
 Notation designers (e.g. members of OMG
taskforces involved in designing or revising diagramming notations): it defines
practical guidelines for constructing effective diagramming notations and improving existing ones.
 Tool vendors: it provides the basis for
providing enhanced tool support for diagramming notations, incorporating advanced graphical capabilities.
Previous presentations
This tutorial has previously been presented at
some of the most prestigious conferences in
the IT field and has received rave reviews from
participants (90-100% ratings for both content and presentation quality). These include:
 International Conference on Model Driven
Engineering Languages & Systems
(MODELS: formerly called the UML conference)
 IEEE International Conferences on Requirements Engineering (RE)
 International Conference on Business
Process Management (BPM)
 International Conference on Software Engineering (ICSE)
 IEEE Symposium on Visual Languages
and Human Centric Computing (VL/HCC)
 International Conference on Advanced Information Systems Engineering (CAiSE).
 International Conference on Conceptual
Modelling (ER)
 Australian Software Engineering Conference (ASWEC)
Page 3
Sample Slides
The “Physics” of Notations
There
There must
must be
be another
another way...
way...
“Here is Winnie-the-Pooh
coming downstairs, bump,
bump, bump on the back of his
head. It is, as far as he knows,
the only way of coming
downstairs, but sometimes he
feels that there really is another
way, if only he could stop
bumping for a moment to think
of it...”
Towards a Scientific Basis for Designing Visual
Notations in Requirements Engineering
 2/6: Current Practice
Visual
Visual language
language is
is one
one of
of the
the oldest
oldest forms
forms
of
knowledge
representation
of knowledge representation
Language for the eye
The
The Design
Design Space
Space (encoding
(encoding side):
side):
The
The Symbol
Symbol System
System of
of Graphics
Graphics
PLANAR
VARIABLES
RETINAL VARIABLES
Horizontal
Position
Shape
Size
Colour
Large
Red
Green
Blue
Medium
Small
Vertical
Position
Brightness
Orientation
Texture
o
45
Low
Medium
High
90o
o
0
 1/6: The Nature of Visual Notations
Visual
Visual notations
notations form
form an
an integral
integral part
part
of
of the
the language
language of
of IT
IT practice
practice
Source: Bertin, J. Semiology of Graphics: Diagrams, Networks, Maps.
University of Wisconsin Press, Madison, Wisconsin, USA, 1983.
Perceptual
Perceptual Distortion
Distortion
Black (0)
3
Infinite range
of physical
variations
Finite number
of perceptible
steps (length)
5
7.5
9
Source: Goldstine, H. H., & von Neumann, J. (1947). Planning and coding
of problems for an electronic computing instrument. Report prepared for
the US Army Ordnance Department.
Unselfconscious
Unselfconscious design
design culture
culture
 Instinct, imitation, tradition
 Inability to explain designs
 Lack of variety
Source: Alexander, C.W., Notes On The Synthesis Of Form. 1970, Boston,
USA: Harvard University Press. 224.
White (10)
 4/6: How Visual Notations Communicate
The
The Physics
Physics of
of Notations:
Notations:
A
A Theory
Theory for
for Visual
Visual Notation
Notation Design
Design
Scientific basis for
evaluating,
comparing,
improving, and
designing visual
notations
Perceptual
Discriminability
Graphic
Economy
Cognitive
Fit
Semiotic
Clarity
Cognitive
Integration
Semantic
Transparency
Visual
Expressiveness
Complexity
Management
Dual
Coding
Source: Moody, D.L. (2009): The “Physics” of Notations: Towards a Scientific
Basis for Constructing Visual Notations in Software Engineering. IEEE
Transactions on Software Engineering, December.
Page 4
Onomatopoeia:
Onomatopoeia: form
form 
 content
content
1.
1. Principle
Principle of
of Semiotic
Semiotic Clarity
Clarity
symbolisation mapping (encoding)
symbol
deficit
?
C1
symbol
redundancy
C2
Metamodel
(semantic
constructs)
C3
Visual
vocabulary
(graphical
symbols)
symbol
overload
?
C4
symbol
excess
denotation mapping (decoding)
Source: Goodman, N. Languages of Art: An Approach to a Theory of
Symbols. Bobbs-Merrill Co, Indianapolis, 1968.
 5/6: Principles for visual notation design
55
4
14
12
52
28
16
6
46
7
15
13
33
16
21
10
Sy
m
bo
lr
ed
un
Sy
m
da
bo
nc
lo
y
ve
Sy
rlo
m
ad
bo
le
xc
es
Sy
s
mb
ol
de
Co
fic
i
dif ntex t
fer tu
en al
t
T ia
dif extu tion
fer al
en
tia
tio
n
Diagram Type
Classes
Components
Composite
Deployments
Activities
Interactions
State Machines
Use Cases
Average
Co
ns
tru
cts
Sy
m
bo
ls
Semiotic
Semiotic Clarity
Clarity Analysis
Analysis Summary
Summary
(UML)
(UML)
15
4
3
4
10
0
3
3
26%
25
4
8
11
14
4
7
1
46%
3
0
1
0
1
0
2
1
5%
29
0
7
3
14
8
4
0
35%
7
3
5
4
7
4
2
1
20%
19
3
6
6
4
4
6
1
30%
 5/6: Principles for Visual Notation Design
 5/6: Principles for Visual Notation Design
Perceptual
Perceptual discriminability
discriminability in
in action
action
“The
“The Magical
Magical Number
Number Seven,
Seven,
Plus
Plus or
or Minus
Minus Two”
Two”
Human channel
capacity or “cognitive
bandwidth” limited by
working memory
capacity
Source: Miller, G. A. (1956). The Magical Number Seven, Plus or Minus
Two: Some Limits on Our Capacity for Processing Information. The
Psychological Review, 63, 81-97.
 5/6: Principles for Visual Notation Design
Visual
Visual Distance
Distance
Cognitive
Cognitive Integration
Integration Theory
Theory
Number of visual variables on which symbols
differ + size of differences (#perceptible steps)
o verall
cognitive
m ap
concep tual
integratio n
D iagram 1
D iagram 2
percep tual
integration
Visual
distance
 5/6: Principles for Visual Notation Design
navigatio n
b etw een
d iagram s
Source: Kim, J., Hahn, J., & Hahn, H. (2000). How Do We Understand a
System with (So) Many Diagrams? Cognitive Integration Processes in
Diagrammatic Reasoning. Information Systems Research, 11(3), 284-303.
Page 5
Dual
Dual Coding
Coding
Visual
Visual expressiveness
expressiveness
Graphics and text should not be enemies
Information
carrying
variables
4
3
2
1
5
6
Visual
expressiveness
Free variables
(degrees of
visual freedom)
7
8
Visual Saturation
0
Non-graphical
 5/6: Principles for Visual Notation Design
Representation
Representation medium
medium (or
(or production
production
method)
method)
Visual
Visual Expressiveness
Expressiveness of
of UML
UML
3
2
1
Non-graphical
4
 5/6: Principles for Visual Notation Design
5
Visual
expressiveness
6
7
8
0
Saturation
 5/6: Principles for Visual Notation Design
Notational
Notational Darwinism
Darwinism
The
The Graphic
Graphic Design
Design Space
Space
Horizontal
Position(x)
Shape
Orientation
 5/6: Principles for Visual Notation Design
Vertical
Position(y)
Size
Texture
Brightness
Colour
Colourising
Colourising diagrams
diagrams
 5/6: Principles for visual notation design
 6/6: Conclusion
A
A Manifesto
Manifesto for
for Designing
Designing Effective
Effective
Visual
Visual Notations
Notations
 6/6: Conclusion
Page 6