Social Synthesis

Finding Dynamic Patterns in Complex Social Systems

Philip Haynes

• Publisher: Taylor & Francis
• Year: 2017
• eBook ISBN: 9781315458519
• Edition: 1
• Topic: Social Sciences
• Subtopic: Sociology
• Index: Social Sciences

How is it possible to understand society and the problems it faces? What sense can be made of the behaviour of markets and government interventions? How can citizens understand the course that their lives take and the opportunities available to them?

There has been much debate surrounding what methodology and methods are appropriate for social science research. In a larger sense, there have been differences in quantitative and qualitative approaches and some attempts to combine them. In addition, there have also been questions of the influence of competing values on all social activities versus the need to find an objective understanding. Thus, this aptly named volume strives to develop new methods through the practice of 'social synthesis', describing a methodology that perceives societies and economies as manifestations of highly dynamic, interactive and emergent complex systems. Furthermore, helping us to understand that an analysis of parts alone does not always lead to an informed understanding, Haynes presents to the contemporary researcher an original tool called Dynamic Pattern Synthesis (DPS) – a rigorous method that informs us about how specific complex social and economic systems adapt over time.

A timely and significant monograph, Social Synthesis will appeal to advanced undergraduate and postgraduate students, research professionals and academic researchers informed by sociology, economics, politics, public policy, social policy and social psychology.

1
Methodology

Towards a representation of complex system dynamics

Introduction

This chapter considers the use of complexity theory in the social science in the last 25 years. The growing application of complexity theory in the social sciences followed the adaption of ideas from complexity science into the natural sciences. The method of Dynamic Pattern Synthesis (DPS) as proposed, developed and demonstrated in this book has its roots in two theoretical and methodological approaches to social science: complexity theory and critical realism. The growing recognition that complexity theory has implications for social science is changing the approach to the research methods used by social scientists (Byrne & Callaghan, 2014; Castellani & Hafferty, 2009). Social researchers informed by complexity theory see the need for methods that can take account of the strong interactions between people in social systems and that the way these interactions occur leads to indeterminate and difficult to forecast outcomes. Rather than looking for fixed causal laws, researchers turn their attention to less stable patterns of association that might suggest temporal causal patterns, these being patterns that change over time and can even reverse in their interactive effect in future relationships. Kauffman (1995, p. 15), a seminal figure in complexity science, notes that all of life ‘unrolls in an unending procession of change’. It is this constant change that feeds the need for research to understand dynamics. This chapter will explore the theory of complexity theory and consider how it is influencing social science methodology and its connections with the domain of critical realism and the mixed methods that result. As Byrne and Callaghan (2014, p. 8) remarked in their seminal review of the influence of complexity science on the social sciences

When we say complexity theory we mean by theory a framework for understanding which asserts the ontological position that much of the world and most of the social world consists of complex systems and if we want to understand it we have to understand those terms.

Complexity science

The classical reductionist method

Complexity Theory has its origins in contemporary scientific methodology and it questions the universality of the assumptions of Newtonian reductionist methods. Sir Isaac Newton, argued to be one of the greatest scientists of all time, and Luca-sian Professor of Mathematics at Cambridge University from 1669–1701, was much involved in the discovery of some of the most important predictable laws of science, for example, gravitation force and its effect on objects on the earth, including the motion of tides, but also its effect on planets in the solar system.

Similar to Newtonian laws, reductionist methods premise that if research analyses the micro detail of physical matter it will be possible to deduce how higher order entities and their forms function. For example, the discovery of microbiology in medicine explained the causes of bacterial related diseases and therefore if drugs could be discovered to kill the destructive microbes, a cure was guaranteed. This classical scientific approach has certainly had historical influence on the development of the social sciences. As Byrne (2002, p. 6) writes:

Traditionally, quantitative social scientists have tried to construct a social mechanics that can generate predictions of future states on the basis of the measurement of variables in the same way in which Newtonian mechanics predicts through the measurement of forces.

Some early social scientists questioned such a mechanistic view. For example, Durkheim, himself, questioned the Newtonian principles of science, even though he is often regarded as a social scientist with much investment in a classical scientific method.

It is in the very nature of the positive sciences that they are never complete. The realities with which they deal are far too complex ever to be exhausted. If sociology is a positive science, we can be assured that it does not consist in a single problem but includes, on the contrary, different parts, many distinct sciences which correspond to the various aspects of social life.

(Durkheim, in Thompson, 2004, pp. 13–14)

But it is not only social science that has challenged the classical and mechanistic view of science. Much of the new science of complexity has emerged from within the physical sciences also.

Beyond reductionist science

Twentieth century postreductionist scientific theories like thermodynamics, quantum mechanics and complexity theory supplemented the Newtonian explanation by exploring how dynamic and creative change can occur through the emergence of order at the micro level. One such example is the law of thermodynamics.

In classical mechanics, time was reversible, and therefore not part of the equation. In thermodynamics time plays a vital role. This is perhaps best expressed in the second law of thermodynamics, which states that the entropy of a system can only increase. Entropy can be seen as a measure of disorder in a system.

(Cilliers, 1998, p. 8)

In addition, quantum theory or quantum mechanics is the science of the smallest building blocks of physical materials, atomic and sub atomic particles. Instead of reducing the particles to predictable laws, quantum mechanics demonstrates that such particles behave probabilistically rather than deterministically. In ‘Quantum leaps’, particles will pass through barriers and boundaries, or behave differently in sub atomic structures, in unpredictable probabilistic ways, rather than always responding to precise deterministic rules (Frisch, 2014). Sub atomic matter may behave simultaneously as a particle and a wave and may demonstrate surprising future behaviours in terms of their movement, location and consequences. Quantum physics and the study of sub atomic particles resulted in a science of wave motions and probability events, rather than a fixed and highly predictable movement of its component parts.

Such phenomenon challenge single dimensional and highly predictable versions of science. Quantum theory implies that increased analysis of scientific detail will not solve all understanding, but instead micro detail may have dynamic, random and chaotic characteristics that are not easy to predict. This has similarities with complexity theory, where the interaction of physical materials can be dynamic and unstable, resulting in analogous historical patterns of interaction, but never a completely identical outcome. For example, different outcomes might result when the same apparent processes are observed at differing levels and scale. Weather systems are the most obvious physical examples, but neurological and cognitive systems can also be understood by applying complex theoretical and methodological frameworks.

One of the earliest technical acknowledgements of scientific complexity that could not be easily explained by fixed rules of motion was Poincare’s three body problem. This demonstrated that the movement of three celestial bodies subject to gravitational forces was theoretically unstable and chaotic and not easy to determine by the laws of motion. In general, the movement of three celestial bodies was found to be nonrepeating and thus not reducible to an outcome based on the known laws and rule of gravitational forces. Subsequent reflections on Poincare’s problem suggest that the patterns of stability usually observed in the relationship of three celestial bodies are created by a combination of resonance and gravitation. The influence of complexity science has therefore resulted in a different view of the universe. It is no longer seen as governed by entirely mechanistic laws, but rather viewed as an evolving complex system. It is dynamic not static. This book will argue that this complex reality means that social and economic researchers need to be careful about the methods they use and should consider selecting methods which can see the changing dynamics of complex systems and are not blind to the likelihood of unstable change. The relationship of time with social interactions and change is episodic and not stable or linear in its combined impact. An example of this is ‘sensitivity to initial conditions’.

Sensitivity to initial conditions

In the 1960s, an American meteorologist, Lorenz, was programming a mainframe computer to model the weather when he noticed that a small mistake in an input led to very large results in the model output. This was heralded as the discovery that small changes in initial conditions could have exponential results. It has since been referred to as the ‘butterfly effect’, given the conclusion that a butterfly flap-ping its wings in Brazil might cause a hurricane in Texas. More fundamentally it raised significant challenges for Lorenz in terms of the theoretical possibility of building a computer program that could really model a large-scale weather system based on local data inputs. Many scientists came to know this phenomenon as ‘chaos’ (Gelick, 1988).

Such sensitivity to initial conditions has been argued to apply to the historical evolution of economies in the study of social systems. Getting products to market first, and marketing them first, may make exponential success much more likely. Once that a given number of consumers have made the purchase and others begin to copy their behaviour (because they also want the product), a rapid positive feedback loop develops. It is much harder for later producers of the same invention to be as successful. But within this change may be a random effect. For example, if two companies get similar inventions to the market at the same time, it may be luck, or randomness, as much as their advertising strategy that makes one more successful than the other. Examples of this might be the PC and Apple computer, the Betamax and VHS video tape. While companies claim their success is down to marketing and product quality, we cannot always be sure that this was the reason and cannot rule out an element of randomness in rapidly generating consumer patterns of feedback.

If one of two similar products get market supremacy early on in a product life cycle, due to customer behaviour and feedback, this can also be linked with the concept known as ‘path dependence’ (Room, 2011, pp. 16–18). In other words, if the historical path of events is tracked, the observer finds critical points of time and transforming events that explain the longer-term trajectory and a resulting period of stability and dominance which follows the initial conditions (Boulton, et al., 2015).

These early theoretical observations about the importance of initial conditions and subsequent different random paths were later replaced by an appreciation that disorder and instability is often mixed in with periods of order and stability over time. As the late Paul Cilliers said:

Complex systems are neither homogeneous nor chaotic. They have structure, embodied in the patterns of interactions between components.

(2001, p. 140)

Change becomes episodic (Boulton, et al., 2015) rather than chaotic. Systems might periodically get into situations that put them on the ‘edge of chaos’ and more likely to experience substantial change, but such a system state is not permanent. This mix of order and disorder is what we have come to know as complexity, rather than chaos.

The dynamism of micro order in complex systems and the ability of ‘small things’ to create novel outcomes can also be considered with regard to the emergence of new forms of order.

Emergence

Perhaps the most important scientific finding with respect to the understanding of complexity science rather than chaos was the discovery of emergent order in biochemistry by the Nobel Prize winner Ilya Prigogine. He demonstrated that open physical systems could result in dissipative structures that were dynamic and far from equilibrium. Given that simple scientific rules at a micro and local level could result in innovative and creative structures led Prigogine and others to conclude that a given state of order is not deterministic over time. Knowing things by reducing them to detail and examining their past cannot necessarily guarantee a predictable outcome. The chemical reactions studied by Prigogine and his research team showed that simple rules and behaviours led to the emergence of complex patterns rather than simple predictable outcomes.

Smith and Jenks (2006) summarise the importance of Prigogine’s work. Systems reorganise and change far from equilibrium when driven by entropy. This is not a mechanical or deterministic process. Such change is irreversible and unlikely to return to an exact previous system state. The nature of such change is, in part, statistical and random, following patterns of probability. At the micro level, Smith and Jenks (2006, p. 95) conclude:

Prigogine’s studies lead us to a new conception of material probablity. The microspcopic properties of matter, such as substance, particles, molecules, largely independent of each other at equilibrium levels, begin to act together at macroscopic levels at far-from-equilibrium conditions, charged by thermo-dynamics.

This dynamic change also takes place at a relatively high level in the system, rather than in just micro parts, so that a large number of elements and their interactions are involved. This is the discovery of a creative and dynamic world that is best understood by searching for similar patterns rather than singular causes. At the macro level, Pawson and Tilley (1997, p. 72) state: ‘The balance of mechanisms, contexts and regularities which sustains social order is prone to a perpetual and self-generating reshaping’.

Relatively new research methods, like Agent Based Modelling (ABM), put more emphasis on the bottom up emergence of actor behaviour to create novel social behaviour and seek to demonstrate the collective results of models of agent behaviour (Gilbert, 2008). Perhaps the best-known example is Schelling’s (1971) model of ethnic segregation in urban populations, where it was demonstrated that, even if agents were tolerant of ethnic difference, if they nevertheless mildly preferred ethnic similarity in their choice of neighbourhood, over time ethnic segregation would follow. In more recent times, Bar-Yam (2005) has presented a similar type of argument about the causes of ethnic violence in civil stress such as was experienced in former Yugoslavia in the 1990s. In sum, relatively small scale incidents and decisions can quickly get scaled up to represent a social phenomenon with devastating larger scale effects in the macro system.

Byrne (2004, p. 55) neatly summarises the importance of emergence for the social sciences.

The important thing about complex systems is that they have emergent properties. They are inherently non-analysable because their properties, potentials and trajectories cannot be explained in terms of the properties of their components.

If complexity theory demonstrates the importance of system dynamics and reveals the ever-important emergence of change and innovation, it is also important to rethink the notion of the relationship between the individual and society and how separate and joined social sub systems and systems interact in an unstable world. The scientific concept of ‘autopoiesis’ has been used to explore these challenges in the social sciences.

Autopoiesis

Another seminal moment in the development of complex systems theory was Maturana and Varela’s (1992) proposal of the concept of ‘autopoiesis’. This is a cell or system that can maintain or reproduce itself. It can refer to itself in a social process, it is ‘self-referential’. Luhmann (1995) expanded these concepts in sociology and applied the principles of autopoiesis to social systems. Here was a branch of systems theory that turned its attention to the unity of a case within a system, or a part of a system (or sub system) and its ability to keep itself discrete from external systems and the environment. This allowed one component to be identified as ‘different’ while still operating as part of something else. This leads to an important area of research and scholarship that debates the relative openness or closure of complex systems.

Autopoiesis is something of a paradox for social systems in terms of social differences to physical systems. Cilliers (2001, p. 141) expresses the following concern.

An overemphasis on closure will also lead to an understanding of the system that may underplay the role of environment…if the boundary is seen as an interface participating in constituting the syste...

Table of contents

• Cover
• Title
• Copyright
• Contents
• List of Illustrations
• Acknowledgements
• Abbreviations
• Introduction
• 1 Methodology: towards a representation of complex system dynamics
• 2 The method: introducing Dynamic Pattern Synthesis
• 3 Macro examples of Dynamic Pattern Synthesis
• 4 A meso case study example: London boroughs
• 5 Micro case study example: older people in Sweden
• 6 Conclusions
• References
• Index