Complexity

Theory
Si Guide Series
 What we Cover in This Guide

Probably the number one reason we fail to be successful in our endeavors to tackle complex issues is
due to our failing to understand and work with complexity. Successful interventions in to complex
systems requires a basic understanding of complexity. This is given to us by complexity theory, which
helps us to understand the basic features and dynamics of complex systems such as self-organization,
nonlinearity, networks, adaptation and evolution.

This guide is designed as a synthesis of the many different perspectives and approaches to complexity
theory. The most important thing is a conceptual and intuitive understanding for complex systems. The
aim here is primarily to communicate the central concepts in an intuitive fashion. This should enable
readers to be able to understand and communicate in a few words what complexity is and the key
features to complex systems.

After much time spent reviewing the literature on the subject, we identify four basic interpretations to
the theory of complexity that forms the structure to this guide seen on the following page.
 Systems Thinking Systems Awareness Systems Theory Complexity Theory Adaptive Systems

Where am I?
This paper forms part of
our set of 20 guides guide
covering all things systems
innovation. System Inquiry Systems Modeling Mapping Actor Mapping Leverage Points

Transition Design Transition Models Furtures & Naratives System Gardening Co-Design

+++
Ecosystem Building Networks Value Network Scaling Change Systemic Evaluation
 Guide Content

Self-Organization Phase Transitions Nonlinearity Networks

Self-organization deals with how We see how interdependence We talk about the difference We explore network theory
many small distributed parts between elements creates between linear and nonlinear which helps us to capture and
interact and self- organize to create feedback loops that can lead to systems and chaos theory describe the structure and
global patterns of organization. exponential change and phase architecture of complex
transitions systems of all kind
 What is Complexity Theory?
Complexity theory is a set of theoretical frameworks used for
modelling and analyzing complex systems within a variety of
domains. Complexity theory draws upon models developed in
many different areas such as computer science, ecology,
physics, mathematics, and engineering.

Out of these different models has emerged a core set of

commonalities that over the past few decades has come to be
recognized as a generic framework for studying complex
systems in the abstract.

Complexity theory is used to model and interpret the dynamics

and behavior of systems that exhibit complexity. To do this it
encompasses a very broad and diverse set of models and
methods, such as network theory, self-organization theory,
systems theory, nonlinear systems dynamics, evolutionary
theory, game theory.
 What is Complexity Science?
Complexity science is a new approach or method to science
that has arisen over the past few decades based on the new
theoretical framework of complexity theory and computational
methods in order to make an empirical investigation of the
complex systems that make up our world; such as weather
patterns, ecosystems, the internet or cities.

On a methodological level complexity science uses various

forms of computational methods such as agent-based
modeling, cellular automaton, network analysis among other
computational and data-driven methods. Complexity science
is a science fundamentally based on computation. Because
complex systems typically involve a great many parts, a vast
array of connections, emergence and highly dynamic behavior
it is typically not possible to model them without computer
models and simulation.
 Why Complexity Theory?
Although the world has always been complex with recent advances in
economics, information technology, globalization and urbanization,
among other developments, we find ourselves living in an increasingly
complex world. If we look at where our money, food, clothes, car or
media comes from we will note that they were the product a vast
networks that span around the planet; what we call complex systems.

Although we are now starting to recognize these complex systems we

are far from understanding them. Our more traditional reductionist
approach to science leaves us ill-prepared to understand complexity.
This lack of understanding means a lack of capacity to manage or
design them. The result of this is a lack of agency when it comes to
shaping the outcomes of the systems that affect our lives collectively.
Thus gaining a better, theoretical and empirical, understanding of
complex systems is central to responding to the most complex
challenges of our time.
 What is a Complex System?
A complex system is a special class of system that has evolved to
exhibit complexity, meaning that it is characterized as having many
different parts that are highly interconnected and interdependent.

Systems start simple and go through a process of evolution to

become more complex. The evolution of complexity involves the
process of differentiation - meaning the system comes to have
more parts, with those parts being more specialized - and
integration - meaning the parts become more interconnected and
interdependent.

Through this process of integration and differentiation during the

system's development, it comes to have a large number of diverse
elements that are highly interconnected and interdependent, what
we call complexity.
 Examples of Complex Systems

Global Biosphere Cities Financial Markets Social Networks

Many different subsystems, Millions of people, social Many different agents Individuals forming large
from ocean currents to and economic institutions, interacting and reacting to networks with complex
weather patterns, from technology and environment each others behavior to structures that determine
microorganisms to the all interdependent create the emergent state information flows
geosphere. of a market
 What is a System?
A complex system is a special kind of system. A system is simply a set
of parts called elements and a set of connections between these
parts called relations. These parts can be ordered or unordered, an
unordered system is simply a set of things. Because there is no
specific structure or order we can describe a set by simply listing all of
its elements and their properties.

If in contrast, through the relations these parts are ordered in a

specific way then they can function together as an entirety and out of
these parts working together we get the emergence of a global
pattern of organization that is capable of functioning as a coherent
whole; this whole we call the system. The basic model of a system
consists of elements and relations, that are interconnected and
interdependent in forming a whole system that operates in some
environment.
 What is Complexity?
Complexity is a property of a system. Complex systems have
many interconnected and interdependent parts. They consist of
many parts, many elements interacting on many different levels.

Complexity arises when these parts come to have many

interconnections between them. Likewise interdependence
between the parts is also a central characteristic of complexity.
Parts cannot be fully separated or isolated but form synergies with
feedback dynamics.

While the parts of a complex system may be interdependent, they

typically also have high degrees of autonomy and/or adaptive
capacity. Elements have a degree of autonomy often through
their capacity to adapt to their local environment according to
their own set of instructions. This means the system may organize
itself in different ways and is not determined by some top-down
global pattern, but instead may exhibits self-organizing behavior
and evolutionary dynamics.
 Open Dynamic
Complex systems are open, Complex systems are dynamic,
meaning they shape, and are adaptive and change over time
influenced by, many other things through evolution and
in their environment. transformation, often in
unpredictable ways

Emergent
Patterns in complex adaptive
systems can emerge from the
interaction between the parts
without order having to be pre-
specified and imposed.
Interconnected
They are more defined by their networks
of connections and the relationships
Characteristics of between these, rather than the properties
of their parts
Complex Systems
 Self-
Organization
 Self-Organization Overview
Self­-organization is one of the major themes within complex systems and a pervasive phenomenon in our
world, complex organizations like schools of fish, ant colonies, or car traffic manage to organize
themselves into emergent patterns without any form of global coordination.
 Self-Organization - How it Works

Initial Disorder Peer Interaction Feedback Attractors

Self-organization can only take Component parts interact & As coordination results in more Positive feedback creates
place in the absence of a fixed some elements come to efficient outcomes it creates a explosive growth bringing
global pattern synchronize their states positive feedback dynamic parts into an emergent pattern
 Organization
Complex systems are composed of many small parts without
centralized control. Without centralized control, global
organization is an emergent feature of the local interactions
between the parts. Self-organization presents a concrete model for
understanding how this process takes place.

The theory of self-organization has come to explore a new

approach to the age-old question about the emergence of order in
the universe. We tend to assume when we see an ordered system
that the order was somehow externally imposed on the system.

Self-organization describes a process whereby a system transitions

from an initially disordered state to an ordered system through
local internal interactions only, without the need for a global
design pattern to be imposed. A key aspect of this is the need for
an initial state of disorder that creates the space for emergence to
happen.
 Peer Interaction
The second essential element is dense peer-­to-­peer interactions
between elements of the organization. Interactions and
information need to be flowing horizontally not vertically. Dense
interactions are important because all the members may start
out with divergent opinions, activities or agendas, but the more
they interact the greater the requirement to coordinate their
states.

For example, as long as we build a big wall separating societies

we can all go on doing our different things without much need
for coordination or emergent organization. However, when we
take down that wall, we may well all come into initial conflict but
it will be more difficult for us in the long run to maintain our
divergent activities and there will be a much higher reward for
coordinating activities because of the increase in peer
interaction.
 Autonomous Elements
Whereas traditional centralized forms of organization are
relatively static, self­-organization is a dynamic process, that
requires certain conditions, most importantly it requires
autonomous or adaptive elements densely interacting locally.

If we look at the process of self­-organization within a flock of

birds or school of fish, we see emergent self­-organizing
patterns form through each fish or bird operating under their
own set of simple instructions to follow their local neighbors
while also maintaining a certain distance from them.

Because all the elements are responding locally and can act
immediately to what they experience self-organizing systems
are often highly robust being able to adapt rapidly and
reform after an external alteration.
 Feedback Dynamics
Feedback loops are the mechanisms through which some
local small event, which may have been random at the start,
can get amplified into a new macro level pattern of
organization. A positive feedback loop is one that is self-
reinforcing: more begets more. The more products a business
sells the more it can invest in its business, meaning it can
produce better, cheaper products, meaning it will sell more,
etc. This is a nonlinear process of change.

An example of this might be a wave at a football match, this is

an emergent self-organizing phenomenon as no one is
coordinating it. Some small - one or two people raising their
arms - initial event takes hold and gets amplified into a large
macro phenomenon; it emerges out of the synchronized
states of the members. The more people that join in the wave
the stronger the signal for others to do like wise, thus more
begets more through positive feedback which amplifies the
change process.
 Attractors
As this positive feedback process of change continues we
reach a point where there are more elements in this new
configuration than not. At this point the new pattern becomes
the default - that is to say, the easiest option to take - and we
can call this an attractor. An attractor is a set of values or
states toward which a system tends to evolve for a wide variety
of starting conditions to the system.

Those elements in the system that interact more often and

come to some form of consensus or coordination then form an
attractor. Because they are now working together they are
more effective than the other members in the organization,
thus generating better results which makes it more attractive
for others to join that organization because they will get a
greater reward from it. We now get a positive feedback loop
as the more people that join the organization the more
valuable it becomes for future prospective members to do
likewise.
Nonlinear
Systems
 Overview
Although it is often said that nonlinear systems describe the
vast majority of phenomena in our world. They have
unfortunately been designated as alternatives, being defined
by what they are not. For centuries science and mathematics
have been focused upon simpler linear interactions and orderly
geometric forms that can be described in beautifully compact
equations. It is only in the past few decades with the rise of
chaos theory that scientists have begun to approach the world
of systems that are not linear.

Nonlinearity describes how when two things interact the

output is more or less than the sum of their parts in isolation,
thus it is non-additive and nonlinear. This nonlinearity arises out
of the interdependency between elements within a system and
interdependence over time through feedback loops.
 What is a Linear System?
Linear systems are characterized by what is called the superposition
principles. There are just two principles: the additivity principle and
the homogeneity principle.

Additivity states that when we put two or more components together,

the resulting combined system will be nothing more than a simple
addition of each component’s properties in isolation. If we take two
apples and weigh them, each will have a given weight; now, if we
combine them, the combined weight will be nothing more than the
sum of each taken separately.

The principle of homogeneity states that the output of the system is

always directly proportional to the input. Twice as much into the
system, twice as much out, four times as much in, four times as much
out, and so on. The direct implication of this homogeneity principle is
that things scale in a linear fashion, which clearly fails to account for
the effect that the output of the previous state of the system will have
on its current or future states.
 What is Nonlinearity?
Nonlinear systems can be defined as those that defy the
superposition principles. The additivity principle breaks down in
nonlinear systems because the way we put things together and
the type of things we put together affects the interactions that
make the overall product of the components combination more
or less than a simple additive function; thus defying the additivity
principle. There are many examples of this such as bees and
flowers creating a synergistic interaction.

The homogeneity principle implies that there is no feedback loop

over time, that things exist in something of a vacuum. However
as soon as we put a system into its environment where it
operates within both space and time, there will inevitably be
feedback loops, as the actions it takes affect its environment and
other systems with those effects feeding back to affect the future
state to the system. If our economic activity pollutes our
environment now this will over time feedback to affect future
economic activity.
 Chaos Theory
Chaos describes how a system's dynamics over time may be
sensitive to initial conditions, resulting in the potential for
widely divergent outcomes given only very small differences
in the system's input values, thus making the future state of
the system very difficult to predict. This chaotic and
unpredictable behavior happens even though these systems
are deterministic, meaning that their future behavior is fully
determined by their initial conditions, with no random
elements involved.

A double pendulum is a classic example of a chaotic system,

consisting of only two interacting components, each limb,
with these limbs being both strictly deterministic when taken
in isolation, but when we join them this very simple system
can, and does, exhibit unpredictable chaotic behavior.
Starting the pendulum from a slightly different initial
condition would result in a completely different trajectory.
 What is Chaos?
Chaos theory is the study of nonlinear dynamical systems that are highly sensitive to initial conditions - eg
weather patterns - popularly referred to as the butterfly effect. Small differences in initial conditions yield
widely diverging outcomes for such systems, rendering long-term prediction generally impossible.

Leads to widely
divergent possible
outcomes

Small change in
system input value
Amplified by
positive feedback
 Phase
Transitions
 Feedback Loops
Feedback is at the heart of nonlinear phenomena of almost all kinds,
feedback in space or over time is what generates nonlinear behavior.
A feedback loop can be defined as a channel or pathway formed by
an 'effect' returning to its 'cause,' and generating either more or less
of the same effect. An example of this might be a dialogue between
two people, what one person says now will affect what the other
person will say and that will in turn feed back as the input to what the
first person will say in the future.

Feedback loops are divided into two qualitatively different types,

what are called positive and negative feedback. A negative feedback
loop represents a relationship of constraint and balance between two
or more variables when one variable in the system changes in a
positive direction the other changes in the opposite. A positive
feedback loop in contrast, is a self-reinforcing process, where more
begets more, the increase in the values associated with one element
in the relation affects the other to also increase in value, with this
then, feeding back to increase the value of the first. The result is a
compounding effect and exponential change.
 Equilibrium
The concepts of equilibrium and non-equilibrium are an extension of the
idea of linearity and nonlinearity. Equilibrium is a point of stasis where a
system has equal forces acting on it; thus equilibriums systems states are
created by negative feedback. Equilibrium systems states and the
negative feedback that creates them can be seen to be “normal” as
they equate to a stable and sustainable process of development. For
example, we could identify this equilibrium state in someone who is
working to pay their living. The person creates an income which is then
spent again on supporting themselves, creating a financially sustainable
situation as the inflows and outflows balance each other.

A system moves towards non-equilibrium when an action is taken

without there being a counteraction to balance this. When each action
induces the system to go off in the same direction - positive feedback -
the system can rapidly become destabilized and move towards a non-
equilibrium state of change, which is unsustainable. For example, a
financial bubble is a non-equilibrium state of development because
there is no counterbalancing force on the price of the assets, so it just
stays increasing.
 Attractor
For any given system we can create a state
space representing all the possible states that the system
might take. For example, during a week a person may have
a number of different activities, e.g. working, resting,
entertainment, exercising, etc. each of these is a state that
would be represented in the space of different states to the
person's activities.

Over time what we see is that systems typically settle in to

routine set of states that represent the different counter
balancing forces in the system. The attractor is then this
subset of states that corresponds to the system’s typical
behavior.

An attractor is a set of states towards which a system will

naturally gravitate and remain cycling through until altered.
When a system is in a stable single attractor it has a single
equilibrium representing the primary counterbalancing
forces acting on it.
 Bifurcation
A bifurcation means a branching out. In the case of
nonlinear systems theory, the term refers to a point where
the future trajectory of an element in a system divides or
branches out, as new attractor states emerge. From this
critical point it can go in two different trajectories which
are the product of these attractors. Each branch represents
a trajectory into a new basin of attraction with a new
regime and equilibrium.

A bifurcation is a qualitative topological transformation in

the state space resulting in a spitting of this attractor into
two distinct stable attractors. For example, this could be a
political system, where during a revolution what was
previously a dictatorship with a single basin of attraction
now becomes a system with two basins of attraction; the
old regime and the counter movement.
 Regime Shift
A regime defines a macro-level pattern of organization within a
system and an ordered way of operating. When a system
experiences high levels of positive feedback it can be pushed out
of its stable basin of attraction, far from equilibrium and enters into
a phase transition period characterized by rapid change and
instability. During a phase transition, a bifurcation forms and the
system moves into a new regime, under a new set of feedback
loops that create a new basin of attraction with a new equilibrium,
or regime.

Every time a negative feedback loop is broken the system moves

farther away from its stable equilibrium attractor. As it moves away
it moves towards a critical phase transition area far from its
equilibrium; an unstable regime governed by positive feedback
where some small event can get amplified rapidly driving the
system through the phase transition into another basin of
attraction. During a regime shift, the system comes to have two or
more basins of attraction and can flip between them; this is called
"bistability."
 Equilibrium State Bistable System New Quilibirum
With a single dominant attractor Emergence of a new attractor New defualt attractor

Regime A Regime A Regime B Regime B

For example Dictatorial political system Dictatorship Democatatic Democatatic
 Phase Transition
A phase transition may be defined as some smooth, small
change in a quantitative input variable that results in an
abrupt qualitative change in the system’s overall state.

The transition of ice to steam is one example of a phase

transition. At some critical temperature, a small change in
the system's input temperature value results in a systemic
change in the substance after which it is governed by a
new set of parameters and properties. For example, we
can talk about cracking ice but not water, or we can talk
about the viscosity of a liquid but not a gas as these are in
different phases under different physical regimes and thus
we describe them with respect to different parameters.
 Punctuated Equilibrium
As opposed to linear systems that may develop in an overall
incremental fashion, the exponential growth that nonlinear systems
are capable of leads to a different overall pattern to their
development, what we might call punctuated equilibrium.

Within this model of punctuated equilibrium, the development of a

nonlinear system is marked by a dynamic between positive and
negative feedback, with negative feedback holding the system
within a basin of attraction that represents periods of stable
development. These stable periods are then punctuated by periods
of positive feedback which take the system far from its equilibrium
and into a phase transition as the fundamental topology of its
attractor states change and bifurcate.

Examples of this punctuated equilibrium might be the

development of economies that go through periods of stable
growth then rapid change through an economic crisis and recovery.
Network
Theory
 Overview
Network theory can be understood as the study of
connectivity. Network theory is before anything a way of
looking at systems that focuses our attention on connectivity.
What network theory can contribute to our understanding of
complexity is an analysis of the structure to complex systems; it
tells us about their overall makeup and the topology of how
they are interconnected. Network theory explicitly focuses our
attention on the connections between parts and how that
shapes the part and the overall system.

This perspective will become relevant when the system reaches

a certain degree of connectivity, at which point it becomes
more defined by the network of connections rather than the
specific behavior of the parts. It now becomes relevant to
switch our focus to looking at the nature of connectivity within
the system and we use network theory to do this.
 The Nature of Networks

Relational Topology Emergence Nonlinear

Networks are based on a Networks create a very Networks are not formal, Networks are inherently
relational paradigm, they different kind of space to they typically emerge in an complex and nonlinear as
focus our attention on the one we are used to, organic fashion, from the linkages can grow exponentially
connectivity and structure we call this their topology bottom-up relative to nodes
 What’s Covered Here

Graph Theory Network Structure Types of Networks Network Dynamics

Graph theory is the basic Networks are defined by Different network types Network dynamics looks at
formal mathematical their structure as this have been identified as the conditions under which
language for describing shapes how things flow reoccurring across a wide networks form, grow or
networks across them variety of systems disintegrate
 Graph Theory
Network theory comes from graph theory, the area of mathematics
that studies networks. Graph theory is a formal language that
provides us with the basic vocabulary to start to talk about networks.
Graph theory helps us in starting to understand the basic elements
2 of graphs, the basic parameters to defining nodes and connections.
The theory of graphs helps us to define important and basic features
to a network such as degree of connectivity, node centrality,
1 weighted graphs etc.

Graphs can be relatively simple having only one kind of relation

between the parts or more complex multi-level graphs. Graphs that
2 have multiple types of edges and nodes are called multiplex
networks which allow us to capture how different networks
interrelate, overlap and affect each other. For example, a city may
be modelled as a multiplex network consisting of many intersecting
networks, e.g. water systems, transport networks, social networks
etc. It is important to be aware that real world complex systems are
the product of many overlapping networks interacting dynamically.
 Node Centrality
A primary question people wish to answer when
analyzing a network is that of how influential or
significant a given node is within the overall network,
this is called the node centrality. For example, we might
want to know, who is the most influential person in a
social network, or the most important business within
an industry.

Centrality indices are answers to the question “What

characterizes an important node?” However this
question is not simple to answer as there are many
factors that can give a node an influential position
within a network and it often depends on the question
being asked.
 Metrics for Quantifying Node Centrality

Degree of Closeness Betweenness Prestige

Connectivity Centrality Centrality Metrics
A measure of the direct How close a node is to any Captures the node’s role How significant a node is
number of connections the other node, i.e. how easily as a connector or bridge based upon how significant
node has the node can reach other between other groups of the nodes it is connected to
nodes nodes are
 Network Structure
The structure of a network may be called its topology. Topology is the overall pattern or structure to a
network; the way in which the constituent parts are interrelated or arranged. Within the context of
network theory, it defines the way different nodes are placed and interconnected with each other and has
a significant effect on how something will spread through the system. Examples of different network
topologies include star, tree or ring structures.

Many parameters can be identified as affecting this structure such as the overall degree of connectivity,
the number of nodes, the average path length, the systems diameter etc. Primary among these
parameters is network density, how connected is the overall system. Key Concepts here include network
topology, connectivity, diameter, clustering and more.
 Types of Networks
The degree to which a network is centralized or
decentralized is a key determinant of its overall workings.
This feature to a network is captured by the network’s
degree distribution. Degree distribution tries to capture the
difference in the degree of connectivity between nodes in a
graph. It is really asking the question, do all the nodes have
roughly the same amount of connections or do some have
very many while others have very few connections?

By answering this, we will get an idea of how centralized or

distributed it is, which is a defining factor to networks telling
us how something will flow through it, which nodes have
influence, or how quickly can we affect the entire network.
Key concepts here include degree distribution, Small World
networks, Scale Free networks, distributed graphs and
centralize vs decentralized networks.
 Centralized Networks
Centralized networks represent networks with a very high degree
distribution, meaning they are very unequal in terms of how
connected and influential the different nodes in the network are. In
this type of network structure, there will be very many nodes with a
very low level of connectivity and a very few with an exceptionally
high degree of connectivity.

A good example of a centralized network might be global banking

activity, as a very few core nodes dominate the global financial
system. This type of centralized structure to a network is
surprisingly prevalent in our world seen in social networks where a
very few people may have millions of followers while the vast
majority have few.

These kinds of networks are often the product of preferential

attachment which describes how a resource is distributed among a
number of nodes according to how much they already have, so
that those who already have a lot receive more than those who
have little. In more familiar terms this is called “the rich get richer.”
 Decentralized Networks
Distributed networks are defined by a low level of degree
distribution. This means that most of the nodes have a similar
degree of connectivity. As there are no dominant nodes to
provide global functions for the entire network, each node
must contribute to the network’s maintenance. In a
distributed network there is limited centralized global
coordination, nodes are largely self-sufficient and are not
dependent upon nodes outside of their neighborhood.

An example of a distributed network might be a community

alert group, where each member of the community has equal
responsibility and authority to act when there is an event that
others should know about. There is no hierarchy, and in this
example, the network is only actualized when needed, thus
placing very limited constraints on its members. Within the
world of computing distributed networks are also called mesh
networks.
 Characteristics of Decentralion
Distributed networks have a number of advantages and
disadvantages. On the positive side, they may be very robust to
failure. As there is no critical or strategic nodes in the network, any
node can theoretically be replaced by any other. As noted
elements may have a high degree of autonomy, with limited
network maintenance taxes placed upon them. But also this type of
network can be less efficient in many circumstances.

Without centralized nodes, there is typically limited capacity for

centralized batch processing to leverage economies of scale, and
diffusion across the network can be slow as there are no central
hubs with which to reach many nodes in a single hop. This can also
create problems in terms of coordinating the network as a whole.
In many ways, a distributed network represents a system in a fine
balance and relatively stable state and this is often not what we see
when we look at real-world networks - which are typically
centralized in some way to take advantage of the efficiencies
created.
 Network Dynamics
The study of how networks form, grow and change over time is a
relatively new area of research, but it is critical to understanding how
to foster the development of some types of networks and reduce the
development of others. Likewise, the rules under which a network
was created and developed will play a large role in how something
will spread across it and ultimately how robust it is to failure. For
example, researchers have studied innovation as a process of
diffusion across a network.

Because networks are inherently nonlinear their growth is typically

nonlinear also, meaning there will likely be sub-linear growth up to a
certain tipping point, and then positive feedback will kick-in to give
us super-linear exponential growth. If enough nodes join the
network, then we may reach the critical mass and get a tipping point.
The tipping point is the critical point in the system’s development as
it defines where positive feedback will gain traction leading to rapid
and irreversible state change.
 Network Diffusion
Network diffusion describes how something spreads out or
diffuses along the connections of a network as a function of the
properties of its nodes and the structure of connectivity within
the system. A classic example of this would be the diffusion of
a disease through some population; equally, we might be
dealing with how the loss of one species in an ecosystem has
an effect on others; the spread of financial contagion from one
institution to another; or the spread of some information within
a group of people. This spreading on a network is called
network propagation or diffusion.

The network’s topology is a key consideration in understanding

how something is likely to spread across it, the primary factor
here being simply the overall degree of connectivity of the
network. We also need to ask whether this dissemination is
random or strategic. That is, whether there is some logic
behind the promotion and dissemination aimed at strategically
affecting nodes that have a high degree of connectivity, and
thus enabling a more rapid diffusion.
 Network Robustness
Network robustness is the capacity for a network to
maintain functionality given some perturbation. Network
robustness may be modeled in terms of some external
perturbation that propagates through the system,
destroying links and nodes on its way. In such a case we
want to ask how easily does it spread and what is the
resistance to its spreading within the network.

When talking about network robustness & resilience, we

are often asking what will happen to the network’s overall
connectivity and integration if we remove some
components or connections, equally how will this failure
then spread within the network system. A network's
robustness is a function of a number of different
parameters such as its degree of overall connectivity, the
nature of the connections in the system and the overall
structure of the network - e.g. is it centralized or
distributed - among other factors.
Conclusion
 Takeaways
Complexity theory is a powerful set of models to help us better understand complex systems. By
understanding complexity theory we can truly start to work with complexity rather than against it - which
is a huge advantage given the world of complex systems we find ourselves inhabiting in the 21st C.

Many of the findings from complexity science run counter to our intuition and traditional linear ways of
thinking. For this reason it is important that we are aware of these features to complex systems before
making interventions; otherwise our best-made plans may well fail by simply operating under the wrong
assumptions. The number one reason we fail in tackling complex issues is because we operate based on
simple linear assumptions and fail to recognize the nature of complexity.

Complexity theory includes a vast and sophisticated set of models. Going in-depth on these is far
beyond the scope of this short guide, thus this guide should certainly not be thought of as exhaustive.
Here we have simply touched upon some of the key aspects to complex systems that you will need in
order to get a basic understanding for their dynamics and characteristics. We encourage you to continue
your reading in this area so as to broadened your understanding of complexity and your capacity to work
with it.
 CC BY-NC 4.0 DEED

Version 2.0
A Systems Innovation Publication
www.systemsinnovation.network
Reference and Resources
Turner, J.R. and Baker, R.M. (2019). Complexity Theory: An Overview with Potential Applications for the Social Sciences. Systems, [online] 7(1), p.4. Available at: https://www.mdpi.com/2079-8954/7/1/4/htm [Accessed 1 Sep.
2020].

Wikiwand. (2018). Complex system | Wikiwand. [online] Available at: https://www.wikiwand.com/en/Complex_system [Accessed 1 Sep. 2020].

Render, J. (2019). What is Complexity Theory and How Can it Improve the Workplace? [online] Agile-Mercurial. Available at: https://agile-mercurial.com/2019/05/13/what-is-complexity-theory-and-how-can-it-improve-the-
workplace/#:~:text=Complexity%20Theory%20is%20primarily%20made,as%20the%20organization%20of%20businesses. [Accessed 1 Sep. 2020].

Amagoh, F. (2016). Systems and Complexity Theories of Organizations. Global Encyclopedia of Public Administration, Public Policy, and Governance, [online] pp.1–7. Available at:
https://link.springer.com/referenceworkentry/10.1007%2F978-3-319-31816-5_73-1#:~:text=Systems%20theory%20focuses%20on%20structures,and%20outcome%20of%20the%20system. [Accessed 1 Sep. 2020].

Cybernetics | Britannica. (2020). In: Encyclopædia Britannica. [online] Available at: https://www.britannica.com/science/cybernetics [Accessed 1 Sep. 2020].

Complexityexplorer.org. (2020). Complexity Explorer. [online] Available at: https://www.complexityexplorer.org/courses/55-introduction-to-information-theory [Accessed 1 Sep. 2020].

Willy, C., Neugebauer, E.A.M. and Gerngro, H. (2003). The Concept of Nonlinearity in Complex Systems. European Journal of Trauma, [online] 29(1), pp.11–22. Available at: https://link.springer.com/article/10.1007/s00068-
003-1248-x [Accessed 1 Sep. 2020].

Jaeger, Heinrich M and Liu, A.J. (2010). Far-From-Equilibrium Physics: An Overview. [online] arXiv.org. Available at: https://arxiv.org/abs/1009.4874 [Accessed 1 Sep. 2020].

Introduction to Network Introduction to Network Theory Theory. (n.d.). [online] Available at: https://www.cl.cam.ac.uk/teaching/1011/PrincComm/slides/graph_theory_1-11.pdf [Accessed 1 Sep. 2020].

Flying Magazine (2019). Europe 24: Air Traffic Data Visualization. YouTube. Available at: https://www.youtube.com/watch?v=paG20mOqLEQ [Accessed 1 Sep. 2020].

Wikiwand. (2013). Graph theory | Wikiwand. [online] Available at: https://www.wikiwand.com/en/Graph_theory#:~:text=In%20mathematics%2C%20graph%20theory%20is,also%20called%20links%20or%20lines). [Accessed 1
Sep. 2020].

Chan, S. (2001). Complex Adaptive Systems. [online] Available at: http://web.mit.edu/esd.83/www/notebook/Complex%20Adaptive%20Systems.pdf.

Haken, H. (2008). Self-organization. Scholarpedia, [online] 3(8), p.1401. Available at: http://www.scholarpedia.org/article/Self-organization [Accessed 1 Sep. 2020].

Baskin, K. (2013). The Complexity of Evolution: History as a Post-Newtonian Social Science The implications of Quantum Mechanics were so violently shock- ing to its pioneers that for Werner Heisenberg ’the foundations.
Social Evolution & History, [online] 12(1). Available at: https://www.sociostudies.org/journal/files/seh/2013_1/003-027.pdf [Accessed 1 Sep. 2020].

Schueler, G.J. (1996). The Unpredictability of Complex Systems. Journal of the Washington Academy of Sciences, [online] 84(1), pp.3–12. Available at: https://www.jstor.org/stable/24530991?seq=1 [Accessed 1 Sep. 2020].

Wikiwand. (2020). Systems design | Wikiwand. [online] Available at: https://www.wikiwand.com/en/Systems_design [Accessed 1 Sep. 2020].

Rzevski, G. (2015). Complexity as the defining feature of the 21st century. [online] ResearchGate. Available at: https://www.researchgate.net/publication/279668439_Complexity_as_the_defining_feature_of_the_21st_century
[Accessed 1 Sep. 2020].
Meaning it is composed of many diverse and autonomous components that interact in a nonlinear, networked fashion with the whole system evolving over time.

Petty, M. (n.d.). Modeling and Validation Challenges for Complex Systems. [online] Available at: https://www.sisostds.org/DesktopModules/Bring2mind/DMX/API/Entries/Download?
Command=Core_Download&EntryId=35445&PortalId=0&TabId=105 [Accessed 1 Sep. 2020].

Wikiwand. (2018). Complex system | Wikiwand. [online] Available at: https://www.wikiwand.com/en/Complex_system [Accessed 1 Sep. 2020].

Pelánek, R. (n.d.). PV205: Complex Systems, Seminar Themes. [online] Available at: https://www.fi.muni.cz/~xpelanek/PV205/slidy/intro.pdf [Accessed 1 Sep. 2020].

Systems Innovation (2015). Systems & Sets. YouTube. Available at: https://www.youtube.com/watch?v=U7RzKZQjAK8 [Accessed 1 Sep. 2020].

Psu.edu. (2020). Download Limit Exceeded. [online] Available at: http://citeseerx.ist.psu.edu/viewdoc/download?doi=10.1.1.59.6293&rep=rep1&type=pdf [Accessed 1 Sep. 2020].

Siyari, P., Dilkina, B. and Dovrolis, C. (2018). Emergence and Evolution of Hierarchical Structure in Complex Systems. [online] arXiv.org. Available at: https://bit.ly/3lDdReQ. [Accessed 1 Sep. 2020].

Willy, C., Neugebauer, E.A.M. and Gerngro, H. (2003). The Concept of Nonlinearity in Complex Systems. European Journal of Trauma, [online] 29(1), pp.11–22. Available at:
https://link.springer.com/article/10.1007/s00068-003-1248-x [Accessed 1 Sep. 2020].

Barlas, Y. (n.d.). 5.12 SYSTEM DYNAMICS: SYSTEMIC FEEDBACK MODELING FOR POLICY ANALYSIS. [online] Available at: http://web.boun.edu.tr/ali.saysel/ESc59M/BarlasEOLSS.pdf [Accessed 1 Sep. 2020].

Turnbull, L., Hütt, M.-T., Ioannides, A.A., Kininmonth, S., Poeppl, R., Tockner, K., Bracken, L.J., Keesstra, S., Liu, L., Masselink, R. and Parsons, A.J. (2018). Connectivity and complex systems: learning from a multi-
disciplinary perspective. Applied Network Science, [online] 3(1). Available at: https://appliednetsci.springeropen.com/articles/10.1007/s41109-018-0067-2 [Accessed 1 Sep. 2020].

Turnbull, L., Hütt, M.-T., Ioannides, A.A., Kininmonth, S., Poeppl, R., Tockner, K., Bracken, L.J., Keesstra, S., Liu, L., Masselink, R. and Parsons, A.J. (2018). Connectivity and complex systems: learning from a multi-
disciplinary perspective. Applied Network Science, [online] 3(1). Available at: https://appliednetsci.springeropen.com/articles/10.1007/s41109-018-0067-2 [Accessed 1 Sep. 2020].

Wikiwand. (2020). Complex network | Wikiwand. [online] Available at:

https://www.wikiwand.com/en/Complex_network#:~:text=In%20the%20context%20of%20network,graphs%20modelling%20of%20real%20systems. [Accessed 1 Sep. 2020].

Petty, M. (n.d.). Modeling and Validation Challenges for Complex Systems. [online] Available at: https://www.sisostds.org/DesktopModules/Bring2mind/DMX/API/Entries/Download?
Command=Core_Download&EntryId=35445&PortalId=0&TabId=105.

Comfort, L. (1985). INTEGRATING THEORY AND PRACTICE IN DYNAMIC SYSTEMS Self-Organization in Complex Systems. [online] Armenia. Costa Rica. Available at:
http://theisrm.org/documents/Comfort%20%281994%29%20Self-Organization%20in%20Complex%20Systems.pdf [Accessed 1 Sep. 2020].

Rainer Feistel and Ebeling, W. (1989). Evolution of Complex Systems: Self Organization, Entropy and Development. [online] ResearchGate. Available at:
https://www.researchgate.net/publication/237005512_Evolution_of_Complex_Systems_Self_Organization_Entropy_and_Development [Accessed 1 Sep. 2020].

OpenLearn. (2019). Management: perspective and practice. [online] Available at: https://www.open.edu/openlearn/money-business/leadership-management/management-perspective-and-practice/content-
section-2.3.1 [Accessed 1 Sep. 2020].
What Is Self-Organization? (n.d.). [online] Available at: http://assets.press.princeton.edu/chapters/s7104.pdf.

Sethna, J. (n.d.). Statistical Mechanics Entropy, Order Parameters, and Complexity. [online] Available at: http://pages.physics.cornell.edu/~sethna/StatMech/EntropyOrderParametersComplexity.pdf [Accessed 1 Sep.
2020].

Heylighen, F. (1970). The Science Of Self-Organization And Adaptivity. [online] ResearchGate. Available at: https://www.researchgate.net/publication/2817974_The_Science_Of_Self-Organization_And_Adaptivity
[Accessed 1 Sep. 2020].

Nonlinear Systems

Wikiwand. (2020). Nonlinear system | Wikiwand. [online] Available at: https://www.wikiwand.com/en/Nonlinear_system [Accessed 1 Sep. 2020].

Smith, M. and Karam, E. (2018). Linear Causality in Family Systems Theory. Encyclopedia of Couple and Family Therapy, [online] pp.1–2. Available at: https://link.springer.com/referenceworkentry/10.1007%2F978-3-319-
15877-8_285-1#:~:text=Linear%20causality%20is%20a%20framework,all%20behaviors%20in%20a%20system. [Accessed 1 Sep. 2020].

Wikiwand. (2020). Nonlinear system | Wikiwand. [online] Available at:

https://www.wikiwand.com/en/Nonlinear_system#:~:text=In%20mathematics%20and%20science%2C%20a,are%20inherently%20nonlinear%20in%20nature. [Accessed 1 Sep. 2020].

Fractalfoundation.org. (2018). What are Fractals? – Fractal Foundation. [online] Available at: https://fractalfoundation.org/resources/what-are-
fractals/#:~:text=Fractals%20are%20infinitely%20complex%20patterns,systems%20%E2%80%93%20the%20pictures%20of%20Chaos. [Accessed 1 Sep. 2020].

Network Theory

Wikiwand. (2012). Network theory | Wikiwand. [online] Available at: https://www.wikiwand.com/en/Network_theory#:~:text=Network%20theory%20is%20the%20study,have%20attributes%20(e.g.%20names). [Accessed 1
Sep. 2020].

Wikiwand. (2013). Graph theory | Wikiwand. [online] Available at: https://www.wikiwand.com/en/Graph_theory#:~:text=In%20mathematics%2C%20graph%20theory%20is,also%20called%20links%20or%20lines). [Accessed
1 Sep. 2020].

Wikiwand. (2020). Degree distribution | Wikiwand. [online] Available at:

https://www.wikiwand.com/en/Degree_distribution#:~:text=The%20degree%20of%20a%20node%20in%20a%20network%20(sometimes%20referred,node%20has%20to%20other%20nodes. [Accessed 1 Sep. 2020].

Sciencedirect.com. (2020). Eigenvector Centrality - an overview | ScienceDirect Topics. [online] Available at: https://www.sciencedirect.com/topics/computer-science/eigenvector-centrality [Accessed 1 Sep. 2020].

Wikiwand. (2020). Network topology | Wikiwand. [online] Available at: https://www.wikiwand.com/en/Network_topology [Accessed 1 Sep. 2020].

Porter, M. (2012). Small-world network. Scholarpedia, [online] 7(2), p.1739. Available at: http://www.scholarpedia.org/article/Small-world_network [Accessed 1 Sep. 2020].

Lenaerts, T. (2011). Scale Free Networks. Encyclopedia of Astrobiology, [online] pp.1492–1493. Available at: https://link.springer.com/referenceworkentry/10.1007%2F978-3-642-11274-4_1405 [Accessed 1 Sep. 2020].

Wikiwand. (2020). Scale-free network | Wikiwand. [online] Available at: https://www.wikiwand.com/en/Scale-free_network [Accessed 1 Sep. 2020].

Wikiwand. (2020). Robustness of complex networks | Wikiwand. [online] Available at:

https://www.wikiwand.com/en/Robustness_of_complex_networks#:~:text=Robustness%2C%20the%20ability%20to%20withstand,is%20important%20for%20many%20fields. [Accessed 1 Sep. 2020].