History

A brief history systems thinking

The roots of systems-thinking go back a very long way. Aristotle (384-322BC) wrote ‘the whole is something besides the parts’ [Aristotle 2016] – in modern terms ‘the whole is more than the sum of its parts’. The even older Chinese book ‘I-Ching’ written around 1000BC -700 BC means ‘everything is changing’. The essence of an object derives from the changing functions of its components. The movement of sun and moon and stars is dynamic circularity (feedback loops) of dynamic balance. Followers of Confucius (551-479BC) desire a harmonious society rather than a relationship with the supernatural – nature and humans are one whole. Central to Buddhism is the idea of constant flux – things come into being and cease to be – with no inherent or fixed nature. Phenomena arise together in a mutually interdependent web of cause and effect.

By contrast in the west, since Aristotle, we have regarded substance as a fundamental notion. Substance is the thing itself and separate from its properties and the processes. Matter of a substance is the stuff of which it is composed and form is the way that stuff is put together so that the whole can function. Matter has the potential to become an actual form – for example wood has the potential to be formed into a chair. Change is not perpetual, as in Chinese thinking, rather it is a process in which matter either takes on or loses form.

A few influential western thinkers have seen substance as too static. Before looking at this let’s just review briefly systems-thinking in the 20th century. We will rely largely on a book by Ramage and Shipp [2009]. Unfortunately they like many others, describe systems-thinking as making sense of the complexity of the world by looking at it ‘in terms of wholes and relationships, rather than splitting it down into its parts and looking at each in isolation’.

We take a different approach. We value wholes and relationships but we do not reject the modelling of parts in isolation – that would be to reject most of engineering science. Our decisions will be the result of judgements about the dependabilities of, and the interactions between, all of our isolated models. In other words our lower level models need to be successful together for the success of the higher level model.

Ramage and Shipp group the people who have contributed to the development of systems-thinking into seven categories – early cybernetics, general systems theory, systems dynamics, soft and critical systems, later cybernetics, complexity theory and learning systems. In all of these movements people were willing to cross disciplinary boundaries and not be constrained by subject silos. But Ramage and Shipp did not include systems engineering nor risk studies. The former includes INCOSE  which is dominated by aeronautical, space, defence, nuclear and IT industries with, until recently, little input from civil engineering. Civil engineering systems has its own rather separate community.

In a very brief history we cannot hope to include all those who contributed to the development of these disciplines. The term cybernetics was coined by Norbert Wiener (1894-1964) to capture the study of control and communication in animals and machines and in particular the formalisation of feedback. He defined information as a measure of the degree of organisation of a system. He saw the idea of a message, as a sequence of measurable events distributed in time, as important – whether transmitted by electrical, mechanical or nervous means. He influenced Gregory Bateson (1904-1980) who extended the ideas to the social sciences. Stafford Beer (1926-2002) applied cybernetics to management of large organisations. He believed that organisations needed to become more flexible and adaptive. He developed the Viable Systems Model (VSM) to explain and facilitate how systems are capable of independent existence. It contains five interacting subsystems; 1) function & implementation, 2) communication & coordination, 3) structures & control, 4) looking outwards, monitoring & planning and 5) policy making & steering the organisation as whole.

General systems theory is about systems as a whole with properties that cannot be reduced to the properties of their components. It is concerned with open systems emergence, boundaries and various levels of a hierarchy of organisation. The founders were von Bertalanffy and Kenneth Boulding (1910-1983). Von Bertalanffy suggested that both matter and energy could cross the boundary of an open system. For example living organisms maintain themselves by exchanging materials with their environment with a continuous building up and breaking down of their components. They can produce entropy (disorder) due to irreversible processes but also import negative entropy. Boulding wrote about levels of subsystems embedded in higher level systems or holism first suggested by Jan Smuts (1870-1950, a soldier, philosopher and the 2nd Prime Minister of South Africa from 1939-1948). Arthur Koestler (1905-1983) wanted to integrate reductionism with holism and coined the term holon as a whole (Greek holos) and part (suffix on). They all recognised the venerable history of hierarchy.

Systems dynamics came into public view when Jay Forrester (1918- ) with his MIT team built a global computer model ‘World3’ for the Club of Rome. That work led to the book Limits to Growth [Meadows et al. 1972]. The message was that humanity was living beyond its means – and it sold millions of copies. It wasn’t the first book with this message but was one of the first based on science in the form of systems dynamics models. Systems dynamics is computer simulation of coupled non-linear first order differential equations through time steps. Forrester is an engineer whose early work in flight simulators and control systems was influential in his thinking about management and wider issues. Systems dynamics has close parallels with cybernetics. The structure of a system as a set of interacting feedback loops drives the results. Peter Senge (1947- ) worked with Forrester and is perhaps most well-known for his excellent book The Fifth Discipline [1990] which draws upon previous work in related fields to describe the characteristics of a learning organisation largely in systems dynamics terms. Engineering systems dynamics [Shearer et al 1967] focusses almost entirely on energy and power in mechanical, fluid, thermodynamic and electromagnetic hard systems.

Soft and critical systems developed from the need to locate human being as different from the other components of a system. C West Churchman (1913-2004) brought ethics into systems-thinking – he thought that there was too great an emphasis on mathematical modelling. He moved operational research towards general systems theory. Peter Checkland (1930- ) is most notable for his development of a soft systems-thinking methodology – a set of principles built around problem solving to achieve accommodation or consensus between various players. One of his principal critics has been Mike Jackson (1951- ) who wanted to see more emphasis on power and conflict. He argues that building consensus may not deal with conflict or coercion. The result was the Critical Systems-thinking (CST) approach through which Jackson wants us to go beyond the status quo.

Complexity theory has several interpretations varying from non-linear chaos theory to adaptive systems and is being applied to a wide range of issues from the stock market to insect colonies, ecosystems, the brain, immune systems, manufacturing, political parties, social communities and many more. Indeed it can be applied to any systems of multiply interconnected elements. IIya Prigogine (1917-2003), Nobel laureate, was one of its founders. He developed a theory of dissipative systems from thermodynamics which describes the behaviours of self-organising open systems far from equilibrium.

The press release from the Royal Swedish Academy of Science [1977] reads ‘The great contribution of Prigogine to thermodynamic theory in his successful extension of it to systems which are far from thermodynamic equilibrium………..Quite generally it is possible in principle to distinguish between two types of structures: equilibrium structures, which can exist as isolated systems (for example crystals), and dissipative structures, which can only exist in symbiosis with their surroundings. Dissipative structures display two types of behaviour: close to equilibrium their order tends to be destroyed but far from equilibrium order can be maintained and new structures be formed.’ Prigogine’s great contribution was to address the clash between the view of classical science that time is deterministic and reversible and the common sense view that it isn’t. Einstein is widely quoted as saying time, the separation of past present and future, is an illusion. In other words in classical science (Newton to Einstein) being was the same as becoming. Prigogine showed it is possible to have scientific models in which time is an irreversible arrow. Becoming is no longer an illusion, an expression of ignorance, but rather an expression of instability. A dissipative structure goes through a series of unstable bifurcation points. It is not possible to predict at any one bifurcation what will happen next. Self-organisation emerges at certain critical values of the parameters that control behaviour – without external control. Self-organisation is a process where some form of order (structure) emerges out of local interactions between parts. A system in equilibrium requires little effort to retain its structure and much effort to change it. A dissipative system far from equilibrium needs great effort to retain its structure and little to change it. Prigogine wrote ‘where classical science used to emphasise permanence, we now find change and evolution’. ‘All of chemistry…..involves …irreversible processes.’  He wanted scientific laws that contain the possibility of novelty.

Kurt Lewin (1890-1947) is quoted as saying ‘There is nothing so practical as a good theory’. He is often quoted as the founder of social psychology and coined the term ‘action research’ which embodies the idea that theory led to action and action led to theory. Chris Argyris (1923- ) and Donald Schön (1930-1997) developed the ideas of single and double loop learning (Section 7.3) and the latter was deeply influenced by the work of John Dewey as he elaborated his idea of reflective practice. These ideas led to the work of Claxton, Broadfoot and Deakin Crick.

The major premise of those who promote process philosophy is that being (existing) is dynamic and should be our primary focus. Process thinkers see the world as an assembly of interacting physical, organic, social, and cognitive processes. They debate about how such a world of processes is to be construed, how it relates to the human mind (which is another process) and how the dynamic nature of reality relates to our scientific theories. Two points are worth noting – process philosophy: a) refers to the directionality of time but does not live easily with our Aristotelian and common-sense notion of a process as generating a product – processes beget further processes; b) has a history that extends far into antiquity, both in eastern and western thought – so it is intercultural and no longer parochially western.

The Greek Heraclitus of Ephesus (born ca. 560 B.C.E.) is commonly recognized as the founder of the process approach in the west. He said ‘everything flows’ and that ‘no man ever steps into the same river twice’. In the late 18th and early 19th century Johann G. Fichte (1762-1814), Friedrich W. J. Schelling (1775-1854), and Georg W. F. Hegel (1770-1831) discussed reality as processes of an unfolding dynamic web of dependencies. In the early part of the 20th century Alfred North Whitehead (1861-1947) argued that reality consists of processes rather than material objects and that processes are best defined as a web of interrelations. John Dewey (1859-1952) said that meaning is not abstract but results from human cooperative behaviour. George Herbert Mead (1863-1931) added that mind emerges from social communicative actions. More recently, and as noted earlier, Prigogine has emphasised the arrow of time in process. Nicholas Rescher (1928- ) has argued that rational inquiry is interminable and progressive since new questions are raised so reality is ‘cognitively inexhaustible’ and scientific knowledge increasingly complex.

Many argue  that process ideas promote the best explanation of emergence, self-organization, chaos and complexity. They enable better descriptions varying, for example, from quantum entanglement to consciousness, from computation to feelings and from traffic jams to climate change. Process thinkers seem to be of two main types – the teleological (and often theological) that sees nature moving towards a positive destination and the naturalistic (and generally secularist) that sees nature as dynamic but without any directedness towards a specifiable destination. By this view the arrangements that succeed in establishing and perpetuating themselves will have done so because they are actual improvements.

Over many centuries human beings have searched for certainty and found it in various ways – largely through belief with or without evidence. Aristotle believed that science is universal and invariable and superior to practice which is contingent and variable. Since then knowing has been more valued than doing – like substance it is an idea deeply embedded in our western culture. After Newton we thought we had found certainty in physics. Unfortunately in the 20th century quantum mechanics destroyed this ideal. As we discussed earlier, Prigogine emphasised the role of instability in complex systems. Some philosophers such as Nancy Cartwright have begun to argue that science is about incomplete modelling for a purpose.

Attempts to quantify uncertainty have evolved around the idea of chance. Theories of frequency, betting, randomness and probability did not surface until the 15th century and were not really developed until the 17th with the work of Blaise Pascal (1623-1662). Probability emerging from that time is essentially dual—aleatory (gamble or stable frequency) and epistemic (belief). Aleatory is effectively another name for randomness and epistemic is ‘what we know’.

In the 17th century the word probability meant the extent to which it is right to do something i.e. worthy of approbation, approval, commendation and sanction by an authority such as the church or a wise person. Knowledge was not simply justified true belief but rather was necessary (true in all contexts) demonstrable and universal. Opinion was not demonstrable but referred to belief resulting from reflection or argument. Increasing the probability of opinion might bring certain belief but it would not bring knowledge.

At that time testimony as the evidence of witnesses and authority was ample but evidence provided by things as they demonstrably appeared in reality was lacking. The latter idea was required in order to state the problem of induction which David Hume (1711-1776) did in 1739. Hume doubted that evidence from the past could be used as evidence for the future. Such scepticism however does not prevent people from distinguishing good inductive reasons from bad ones and a new logic of induction was born. Frequency and credibility were linked and a dual concept of probability became possible – men could start to order the different degrees to which hypotheses are supported. Knowledge was not now considered to be universal truth but rather the use of first principles, demonstrations and comparisons of ideas. Cause was not necessary (i.e. in all possible circumstances) for effect but rather constant ‘habitual’ conjunctions.

Risk is a generic term which involves chance, uncertainty and consequences – all modelled in a context. Unfortunately developments of the concept in the 20th century have largely been in subject silos with very little interdisciplinary understanding or tolerance of differing perspectives. This has been particularly so in the mathematical modelling of uncertainty and chance. The various disciplines include mathematical statistics and probability theory, finance and insurance, safety and reliability, disaster studies, engineering (including information technology), health, medicine, philosophy and social science (e.g. psychology, sociology and politics). So financial risks may be, for example, those that are insurable or not, market, default, operational or liquidity. Engineering risks may be safety, reliability and security. Medical risks include disease, health and diet, old age. Social scientists concern themselves with personal everyday activities, how people make risk decisions, risks due to crime, social inequalities and public policies. The major rift in risk studies has been (and still is) huge – between the quantitative (hard systems) and the qualitative (soft systems). Bridging that divide requires systems-thinking.

This brief history demonstrates quite clearly that the ideas of systems-thinking have had a long gestation. Modern thinking in complexity theory and in quantum mechanics links back to the way the ancients thought about change. The unity of the whole and the behaviour of its parts, layers of hierarchy of sense making, the importance of interconnectedness and interaction are coming together into a powerful new philosophy. The methodologies have yet to be fully integrated and appreciated.

This web site is a contribution to the wider development of systems-thinking for practical problem solving. We believe that one of the major barriers to joined-up thinking is a silo mentality. Systems thinking provides a way of thinking and a set of tools with which to overcome disciplinary silos as we face the complex issues of the 21st century.