A systems thinking vocabulary – with ambiguities
Copyright 2019 Graham Berrisford. One of about 300 papers at http://avancier.website. Last updated 17/09/2019 15:20
We spend most of our time thinking in ways not well-called systems thinking.
They might be called “entity thinking”, “situation thinking”, “analytical thinking”, “creative thinking” or just “thinking”.
To call all of those “systems thinking” would be to overload the term until it means everything and nothing in particular.
So can we pin down what systems thinking is about?
That definition includes passive structures and taxonomies, like the Linnaean system for classifying organisms.
Here, the term “system” has the more interesting and useful meaning that emerged in the 20th century.
Generally, a system can be described as actors interacting in activities to advance the system’s state and/or transform inputs into outputs.
· The actors are structures (in space) that perform activities - in roles and processes that are describable and testable.
· The activities are behaviors (over time) that change the state of the system or something in its environment - governed by rules that are describable and testable.
· The state is describable as a set of state variables - each with a range of values.
· An open system is connected to its wider environment - by inputs and outputs that are describable and testable.
These concepts can be seen in writings of Ashby, Forrester and Checkland.
In Ashby’s cybernetics, a system is modelled as processes that maintain a set of state variables.
In Forrester’s system’s dynamics, a system is modelled as inter-stock flows that maintain a set of stocks (variable populations).
In Checkland’s soft systems method, a system is modelled as actors who perform processes that transform inputs into outputs for customers.
Here is a first attempt at defining some related terms.
· A system, actors interacting in activities to maintain the system’s state and/or transform inputs into outputs.
· Abstract system is a perspective or description of an entity as a system (aka soft system).
· Adaptive can mean system state change (as in homeostasis) or system mutation (see below).
· Behaviours are processes performed by parts, often called actors or agents, that play roles in a system.
· Complex is a term with numerous definitions (see below).
· Coupling is the relating of systems (as subsystems) in a wider system.
· Deterministic means that a system, in a given state, responds to a given input or event in a pre-determined way (be it natural or designed).
· Emergence primarily means the appearance of outputs, effects or state changes that arise from coupling subsystems in a wider system.
· Exceptions occur when actors do not complete actions expected of their roles.
· Holism means looking at a thing in terms of how its parts join up (e.g. bicycle and rider) rather than decomposing the parts.
· Linear means in a straight line or a sequence.
· Self-organisation can mean various things, including growth and self-regulation (homeostasis).
The terms above are often ambiguous or undefined in systems thinking discussion.
They can mean different things in different schools, as this table indicates
The measurable complication of an abstract system description
The un-measurable disorder or unpredictability of a real world situation
System state change – updating the values of system variables
System mutation - changing the roles and rules of the system
Actors playing roles and acting according to rules
A group of self-aware actors who inter-communicate and act as they choose, or a problematic situation
A property arising from coupling subsystems into a large system
Not seen before, new, or surprising.
For more, read on.
We do commonly abuse the term “system”.
We point to an entity (e.g. a business organisation or a biological organism) and casually call it "a system".
But with no explicit or implicit reference to a particular system description, this is to say nothing.
Idly calling an entity (or process, problem or situation) a system is meaningless, because one entity can realise countless systems.
“Different observers of the same [concrete] phenomena may conceptualise them into different [abstract] systems.” Ackoff 1971
Ashby urged us not confuse a concrete entity with an abstract system that the entity realises.
Our first impulse is to point at [some real entity or machine] and to say "the system is that thing there".
This method, however, has a fundamental disadvantage: every material object contains no less than an infinity of variables and therefore of possible systems.
Any suggestion that we should study "all" the facts is unrealistic, and actually the attempt is never made.
What is necessary is that we should pick out and study the facts that are relevant to some main interest that is already given.” (Ashby 1956)
System theorists distinguish what some call “soft systems” and Ackoff called “abstract systems” from their realizations.
An abstract system (e.g, the rules of Poker) is a theory of, or insight into, how some part of the world works.
A concrete system (e.g. a game of Poker) is a real-world application, or empirical example, of such a theory.
Science requires us to show the latter conforms (well enough) to the former.
The basis of system theory
Abstract systems (descriptions)
<create and use> <represent>
System thinkers <observe and envisage > Concrete systems (realities)
These papers take this triangular, scientific, view of system theory as axiomatic.
· An abstract system (e.g. the normal regular heart beat) is a description or model of how some part of the word behaves, or should behave.
· A concrete system (e.g. your own heart) is a realisation by a real-world entity that conforms well enough to an abstract system.
Abstract systems are realised by concrete systems; and conversely, concrete systems realise abstract systems.
An abstract system does not have to be a perfect model of what it describes; it only has to be accurate enough to be useful.
And note that systems thinking hides the full complexity of real-world entities that realise systems.
In discussion and testing of the stickleback mating ritual, no attention is paid to any individual stickleback, or its complex internal biochemistry.
A set of roles and rules (the logic or laws actors follow)
The stickleback mating ritual
Actors playing the roles and acting according to the rules
Countless pairs of stickleback.
Note that the relationship between physical entities and abstract systems is many-to-many.
· One abstract system (e.g. the game of poker) may be realised by countless physical entities (countless card schools).
· One physical entity (e.g. a card school) may realise countless abstract systems (e.g. poker, whist etc).
… there can be no such thing as the unique behavior of a very large system, apart from a given observer.
… there can be as many systems as observers… some so different as to be incompatible.
… studying only carefully selected aspects of things is simply what is always done in practice.” (Ashby 1956, 6/14)
Unfortunately, most of us use the term system for both abstract systems (types), and concrete/real world entities or social networks that instantiate them.
And so, in systems thinking discussions, we often confuse them.
But remember, an abstract system may be realised by several real-world entities - each of which might also do others things
And one entity might realise other abstract systems – each defined by taking a different perspective of the entity.
Reconciling different abstract or soft systems is a theme of Checkland’s "soft systems methodology”.
This section elaborates on the definitions above, highlighting some ambiguities.
A system is composed of actors interacting in orderly activities to maintain the system’s state and/or transform inputs into outputs.
Here, as in most system thinking schools, a system is dynamic, meaning it behaves in ways that involve and change the structural state of the system.
This is true in Ashby’s cybernetics, Forrester’s system dynamics and Checkland’s soft systems
And after Forrester, Meadows defined a system thus:
“A set of elements or parts that is coherently organized and interconnected in a pattern or structure that produces a characteristic set of behaviors."
In short, a system is characterised by what does more than what it is made of.
This principle is occasionally called the primacy of behaviour.
Behaviours are processes performed by parts, often called actors or agents, that play roles in a system.
Ashby put it that cybernetics deals with “all forms of behaviour in so far as they are regular, or determinate, or reproducible.”
In other words, the behaviours of a system are orderly processes that can be described by an observer or designer.
When performed, the processes can be tested because they produce observable outputs and/or internal effects or state changes.
Alternatively, the term behavior can refer to the external appearance of a system.
That is, to how a system is observed to change state over time - its state change trajectory.
An abstract system is a perspective or description of an entity as a system (Checkland called this a soft system).
One abstract system may be realised by several real-world entities (each of which might also do others things).
One real world entity can realise several abstract systems – each defined by taking a different perspective of the entity.
Unfortunately, people use the term system for both abstract systems (types), and concrete/real world entities that instantiate them.
And so, in systems thinking discussions, we often confuse abstract systems with the concrete entities or social networks that realise them.
Deterministic means that a system, in a given state, responds to a given input or event in a pre-determined way (be it natural or designed).
The response may be to choose one of many actions and/or to complete a complex process.
The choice between options may be made using random or probabilistic rules – which makes the response relatively unpredictable.
Some equate unpredictability with complexity – which is misleading - since even a very simple system can be unpredictable.
Chaotic means disorderly, with no regular or repeated pattern.
Some use the term non-linear as synonym for chaotic; properly speaking it has a different meaning
Linear state change: progress or change over time that is represented in a graph as a straight line.
Non-linear state change: progress or change over time that is represented in a graph as a curve or jagged.
The system of interest can be orderly, yet produce chaotic results.
The system’s behavior in response to one event may be deterministic and predictable.
Yet the long-term behavior - the trajectory of state changes over time - may be chaotically unpredictable.
Coupling is the relating of systems (as subsystems) in a wider system.
For discussion of coupling varieties, or design patterns, read this paper on system coupling varieties
Emergence primarily means the appearance of outputs, effects or state changes that arise from coupling subsystems in a wider system.
For example, the forward motion of a bicycle and its rider emerges from their interaction; neither can achieve it on their own.
However, the term emergence has been used to mean many other things.
E.g. the emergence of order from disorder, the emergence of a system from what seems chaos, or the emergence of system mutations by evolution or design.
And “The concept has been used to justify all sorts of nonsense.” Gerald Marsh.
For more on emergence, read “Emergence”.
Exceptions occur when actors do not complete actions expected of their roles.
This is common in business systems composed of processes in which human actors play roles.
(People are not well-called the “parts” of any system they play roles in, since they also act outside of their roles.)
The need to design systems with exception handling is a common source of complexity.
Complex is a term with numerous definitions.
The term complexity can refer to internal processing or external appearance.
It can mean complexity in a system’s operations – variety in its variables, roles and rules.
Or else convolutions in a system’s observable state change trajectory
A system that is complex in one way can be simple in the other.
See below for more discussion of complexity.
Holism means looking at something in terms of how its parts join up (e.g. bicycle and rider) rather than dissecting each part.
Having a holistic view of a thing does not mean you know the “whole” - all there is to know about the thing and/or its parts.
Since how you identify the parts and join them up is only one perspective out of countless possible ones.
Adaptive can mean either system state change (as in homeostasis) or inter-generational system mutation.
(Systems, being orderly, can be contrasted with disorganised or ever-changing entities in which no order or pattern can be observed.)
Self-organisation can mean various things, including growth and self-regulation (homeostasis).
In sociology, it often means something very different – the process by which actors who play roles in a system define or redefine the roles and rules of that system.
Social systems thinkers often speak of a Complex Adaptive System (CAS).
The trouble is, it isn't clear they agree:
· why they call thing they are talking about a system
· why they call it complex, or how they could measure that
· in what ways they expect it to adapt and
· when an adaptation changes one system into a different one.
In cybernetics, a system is complex if the system description is complex; the roles and rules are complex
To social systems thinkers, a system is complex if the reality is complex, the actors are complex; their roles and rules may be lightly prescribed, if at all.
In cybernetics, a system adapts to feedback from its environment by changing state – which may be called self-regulating.
To social systems thinkers, a system mutates as actors change its roles, rules or aims - which may be called self-organising.
In cybernetics and system dynamics, a system is a collection of repeated or repeatable activities.
In social systems thinking, a system is a collection of actors, who interact as they choose.
If there are complex adaptive systems, then what other kinds are there
How about a simple system that adapts by changing state: a simple adaptive system (SAS)?
E,g, a bicycle + rider system, or a motor car + driver system.
Social systems thinkers tend to dismiss such systems as "linear" or "mechanistic" or "deterministic".
How about an entity that mutates continually: a continuously evolving entity (CEE)?
E.g. IBM, or any informal human society.
A continuously evolving entity is not a system in the ordinary sense of the term.
It is rather an ever-unfolding process, whose behavior is not describable and testable.
The following sections expand on these ambiguities.
Two kinds of social group may be distinguished.
A social network is a group of actors who communicate with each other, and (here) in which what actors say and do is ad hoc; there are no rules.
A social system is a group of actors playing roles according to rules, in which what actors do is constrained by the roles and rules of the system.
Although any human society or business might be describable as mix of networks and systems, we surely need to be clear which we are talking about.
Suppose you observe a group of people playing a game of cards.
You identify the actors in that social network, then work out the roles they play and the rules they follow.
You can distinguish three things.
· The soft or abstract system - the rules of the game - a perspective of the group’s behaviour.
· The real machine or concrete system – the actors distributing and playing their cards.
· The social network - composed of actors who meet to play roles in the card game.
The actors in the social network are much more than the roles they play in the game of cards.
If your aim is to understand or motivate the actors, you need tools other than "systems thinking".
Suppose you observe a second group of people.
You identify the actors in a social network who talk to each other, and perhaps some goals they have in common.
You see they communicate and do stuff, guided by their personal and to-some-extent shared goals.
But you cannot identify any roles they play or rules they follow.
You see what may be called a social network, but to call it a system adds no useful meaning.
It seems some in the latest generation of systems thinkers are not clear what a system is.
To say every named group of people (or social network) is a system is to use the term with no useful meaning.
To say every entity composed of inter-related parts is a system is to say nothing of practical use.
To say every problem and situation we encounter is a system, is to denude the term of its value.
Some use the term complex adaptive entity to describe entities and situations that are unstable and disorderly.
That is, to describe things which are not systems in either a natural language or a general system theory sense.
If that thing is a business, problem or situation that requires an "intervention”, you may need some social network thinking as well as systems thinking.
Complex systems were introduced by Ashby thus:
“Not until… the 1920s… did it become clearly recognised that there are complex systems…
they are so dynamic and interconnected that the alteration of one factor immediately acts as cause to evoke alterations in others, perhaps in a great many others.”
Ashby defined complexity as measure of state variety – meaning the number of different states a system can take.
Others define it differently.
E.g. Snowden defines complexity in a way that is particular to his classification of problematic situations.
Others speak of complex systems with reference to at least five different situations:
1 A complicated orderly system
Consider an organism or software system whose processes are variegated, complicated or convoluted.
It seems intuitively reasonable to call that complex, or relatively complex.
2 An unpredictable situation
Surely unpredictable cannot mean complex.
A situation in which actors must respond to unforeseen inputs in ad hoc ways is not a system at all.
And unpredictable state change patterns can be produced by very simple systems (see next point).
3 Non-linear system state change
A non-linear or convoluted state change pattern does not imply a complex system.
Chaos theory showed us it can result from repeating simple processes.
4 A disorderly situation
Consider for example a war zone, or an uncoordinated, decoupled, set of silo systems.
Surely disorder is chaotic rather than complex? (Disorder is simpler than order).
5 A system composed by coupling other systems
It might seem obvious that coupling two subsystems makes a more complex system.
But only if you are obliged to describe or manage the internals of those subsystems.
Else, you can take the holistic view and ignore the internals of the subsystems.
E.g. a card game can be described as simple system, regardless of whether it is played by people or software.
A system is an abstraction from reality, and most human system designers take human abilities for granted, they are axiomatic.
Often, an entity is called a complex system where one or more of the following are true.
· No measure of complexity has been agreed.
· No level of abstraction has been agreed.
· No quantifiable properties are described, which makes any measure of complexity impossible.
· No description of the entity as a system has been agreed, or even made.
· No description is possible, because the entity changes continually, rather than generationally.
Which is more complex out of communism or capitalism? Too difficult to answer?
OK, which is more complex out of IBM, Microsoft, a chicken and a hen’s egg?
A description of IBM as a receiver of money from customers and sender of money to suppliers is simple.
A description of IBM that included every activity of every employee would be complex beyond imagination.
But then, a description of a hen’s egg that included every sub-atomic particle would be even more complex.
Read our paper on complexity for more.
Sooner or later, the environment of a system changes in a way that threatens its survival.
System state change
System state change (within a system generation) is a change to the values of a system’s state variables (e.g. body temperature).
To change Ashby’s system state is to update its variable values in response to an event or condition.
To change Forrester’s system state is to change the quantity of a stock in response to an inter-stock flow.
State change examples:
· A bicycle + rider system - changes state by accelerating, or steering to the left.
· A crystal system - changes state by growing in a liquid.
· A heater system - changes state in response to messages from thermostat.
Tools for System Dynamics show the trajectoryquantitative in a graph.
A graph may show the values of a system’s
quantitative state variables
A graph may show the values of a system’s quantitative state variables
· steadily increase or decrease over time
· dwell for a while in a stable state (an attractor), then move to another stable state
· change in a non-linear, jagged or “chaotic” way.
It turns out that the simplest of systems can change state in a non-linear or chaotic way.
Whatever the state change trajectory looks like, it is an inexorable result of actors behaving according to given rules.
While the state of a weather system may change in non-linear way, the laws of physics do not change.
Some describe a system with a non-linear state or chaotic state change trajectory as a complex system.
But while that trajectory may look complex, the system itself may be very simple and orderly.
System mutation (between system generations) is a change to the nature of a system.
To change Ashby’s system is to change its variable types or processes.
To change Forrester’s system is to add/remove stocks or change inter-stock flows.
System mutation examples:
· A bicycle is converted into monocycle.
· A motor car is converted into a boat.
· An entity is replaced by another (as a parent is replaced by a child).
This change may be observed in differences in how the system works or what it produces.
The simplest of systems can mutate, or be replaced by a new generation.
Mutation is creative in the sense that it changes the very nature of a system, from one generation to the next.
Mutation can occur in at least three ways:
· redesign by actors outside the system - as a machine or software system may evolve
· redesign by actors who also play roles inside the system - as a card game may evolve.
· self-replication with changes - as a virus evolves.
First, we must distinguish between system state changes (be they linear or chaotic) and system mutations (be they small or large).
Within a generation, by homeostasis or reconfiguration, a system can adapt to changes; let us call that an agile system.
Between generations, by evolution or redesign, a system can be adapted to changes by a higher-level process or entity.
Second, we must distinguish between continuous and discrete mutation.
A game of cards is a system in which there are regular, determinate and repeatable processes.
People can’t play a game of cards unless the players agree the rules, at least for the duration of one round.
Provided actors change the system incrementally, and all together, the classical concept of a system is upheld.
Continuous mutation undermines the very concept of a system, since it is disorganising rather than organising.
Instead of seeing an island of stability or order, we see an ever mutating entity that is never describable and testable.
Third, we must
distinguish a system from whatever entity or a process causes it to mutate.
Ashby and Maturana said the change agent must sit outside the system of interest; a machine cannot change itself.
For them, “re-organisation” requires the intervention of a higher level process or actor.
E.g. The process of biological evolution runs over and above any individual organism.
And to modify the car + driver system, you play a role in a different system, which may be called car design or psycho-therapy.
This paper defines some terms central to systems thinking and to discussion in other papers.
For further discussion of system change and change control, read “System change”.
The paper on Complex Adaptive Systems explores ambiguities in two academic definitions of that term.
The paper on System thinking ideas used in Agile explores the relevance of systems thinking ideas to agile software development and enterprise architecture.