A quiet revolution – the science of complex systemsIf you haven't heard of complex systems don't worry, you are not the only one. Scientists have been quietly puzzling over the complex interactions that define so many things in our world.
Key text
Key textMost of us at some stage have thought that the world is a complex place. Things are never straightforward or simple. We understand bits and pieces of the world around us, but when it comes to real life, things often happen in ways that we couldn't have predicted.Most of us believe that we simply don't know enough about the thing in question – be that our home, our social network, the stock market or the weather. Many scientists have noticed the same thing in their research and wondered why this might be the case. They realised that knowing about the parts of a system does not often tell us how things work as a whole. The sorts of systems that behave this way are called complex systems, and efforts to better understand them are creating a revolution in science. It's a revolution because working with complex systems goes against traditional science practice. Until now, scientists have spent a lot of time breaking things down into ever smaller component parts – known as reductionism – to understand how each part works in isolation of other parts, only to find that this does not help to understand how the whole system works together. Now, scientists are bringing the pieces of the puzzle together to look at the interactions between components of a system, to understand how the whole system works. Complex systems science is about new ways of learning and understanding: indeed, many believe the rise of complex systems science represents a paradigm shift in thinking. What is a complex system? So what is a complex system? A system is a group of two or more parts which interact to function as a whole. The root word systema means 'organised whole'. The parts of a system are interconnected, and every system is composed of subsystems nested within larger systems. For example, a person is part of a family, which is part of a community, state, nation and world.
The origin of the word complex means 'twisted together' as in the individual threads that make up a tapestry. This suggests the parts of the system are linked in ways that creates a whole, with properties above and beyond those of the parts in isolation. Although there are a wide variety of systems that are complex, they all have two elements in common. They all exhibit emergence and self organisation. Emergence Emergence is the formation of complex but regular patterns from the interaction of the many simple parts of a system. The emergent collective behaviour of a system cannot be predicted merely by understanding its individual elements, or from understanding the interactions between these elements, but it can in principle be predicted by seeing how all the elements work together. It is this element of regularity in the emergent behaviour that distinguishes complex systems from complicated and chaotic systems. Self organisation Self organisation is closely related to emergence and refers to the ability of the system to organise itself. The emergent features of the system appear spontaneously. There is no one in control of the system. We humans like to believe we are in control of the things we construct, and that such mastery should extend over the natural world as well (Box 1: Complex systems, resilience and ecological sustainability). But complex systems are forming around us all the time, often without us being aware that it's happening. Connect up a few computers, as happened when the internet was born, and you have a simple system that's predictable and controllable. Connect up several million computers around the planet and you have a complex system behaving in ways no one could have imagined – and nobody is in control. The same applies to power systems, telephone systems and stock markets (Box 2: Building reliable networks). In addition to emergence and self organisation, complex systems often share other features in common. Other features of complex systems
The importance of local interaction in complex systems How can we make sense of complex systems? In the past, working with systems involved making a number of assumptions, and using relatively simple mathematical models to predict the behaviour of the system. For example, epidemiologists often use models of disease transmission that assume that populations of people are well mixed and everyone can contact everyone else. But such models are often unable to predict how an outbreak will move through a community. Simplified models of complex systems frequently overlook one key feature of those systems – local interaction. The parts of a system tend to only interact with a small subset of the whole system, known as a local neighbourhood. For example, during our everyday lives we interact with friends, neighbours and work colleagues, as well as people on the street, but never everyone in the world. If a flu outbreak reaches a country via an infected tourist arriving at the airport, it's unlikely that you will be infected by direct contact with that person. If the disease spreads, and you become infected, it will probably be via people you are regularly in contact with. The main reason local interactions haven't been included in models in the past is because the maths involved is extremely difficult. With the rise of computing power, models based on local interactions have been compared with simple models based on average properties. To the surprise of many, the models that incorporate local interactions display behaviour that is a lot closer to what is observed in the real world. Mapping complex systems Complex systems such as metabolic pathways, ecosystems and the internet can all be represented as a network of nodes. The nodes of the internet are computers and they are said to be linked if there is any interaction between them. For computers this might be an exchange of information. For other networks the interaction might involve the exchange of materials, energy or diseases. Scale-free networks You can deduce a lot about how a complex system behaves by understanding its network structure. Metabolic pathways in cells, ecosystems and the internet are scale-free networks. They are resilient structures because the random removal of any particular node is unlikely to stop the network from functioning as a whole. On the other hand, if a node with many links was targeted and removed, it could create a large system-wide disturbance – which is just what you want to do for terrorist or drug-trafficking networks.
Complex systems science in Australia
Life is complex, it always was and always will be. However, by acknowledging the true nature of complex systems and focussing on ways of better describing them, we're all better placed to manage in a complex world. Related Nova topic:
Box 1: Complex systems, resilience and ecological sustainabilityEcosystems are complex systems. They are made from many parts that are strongly connected, they behave in non-linear ways, are largely unpredictable, and can exist in alternate stable states in which their function, structure and feedbacks are different.Complex systems can exist in alternative states An example of an ecosystem of alternate stable states is a lake with clear water that receives too much phosphorus over time. Up to a certain point the lake can store phosphorus in the sediment and the water will remain clear. Beyond a certain level or threshold, phosphorus is released into the water, algae multiply and the water becomes murky – and stays that way. Redefining sustainability If you accept that ecosystems are complex systems, then definitions of sustainability need rethinking. In its broadest sense, sustainability is all about maximising the use of our resources today so that we don’t limit the options of future generations. It usually translates to being more efficient with our resources. But many scientists are now worried that by focussing only on efficiency we are more likely to allow ecosystems to cross thresholds into undesirable stable states, such as murky lakes, salinised catchments and depleted fisheries. These scientists believe that resilience is the cornerstone of sustainability, not efficiency. Resilience of ecosystems Resilience is the capacity of a thing to absorb disturbance without shifting to a new stable state. A resilient ecosystem has a greater capacity to avoid unwelcome surprises that occur when you shift to a new stable state, and so has a greater capacity to continue to provide us with the goods and services that support our quality of life. Efficiency, if applied to a complex system without consideration of how it functions over several scales, can result in the removal of redundancy, resulting in a reduction in resilience and the capacity to remain sustainable. As is often the case when dealing with complex systems, sometimes the obvious approach to a problem ends up creating unwanted outcomes. Related sites
Box 2: Building reliable networksScientists who study complex systems are helping us to understand what’s needed to design networks that can cope with a range of disturbances.Modern society depends on the reliable delivery of power, water, telecommunications, finance and government. It’s easy to visualise these services as being delivered to us via a network. To guarantee that these systems don’t shut down from small disturbances – such as the failure of a few components – care must be taken in the way components are connected to each other. An example of the fragility of our large scale infrastructure networks occurred in August 1996, when the western United States power grid collapsed. A sagging power line touched a tree limb, causing a short circuit that triggered a cascading failure, where the backbone of the power grid fell like dominoes. Within minutes, the failure swept across Idaho, Utah, Colorado, Arizona, New Mexico, and Nevada, fracturing the entire system into four disconnected islands, leaving 7.5 million people without power and incurring a damage bill of US $2 billion. The power grid was designed for the efficient delivery of electricity, with little thought given to redundancy. The drive for efficiency produced unintended consequences by increasing the vulnerability of the system and the risk of failure.
Box 3: It's a small world after allMany complex systems are scale-free networks in which some nodes have many links, many nodes have just a few links, and the remaining nodes lie somewhere in between. In this network there is no clear average number of links per node. The internet can be represented as a scale-free network and so can the human population.One of the properties of a scale-free network is that all nodes are only separated by a handful of links – everything is connected. That’s easy to appreciate for computers, but does it apply to humans? Many researchers believe it does and have postulated that any person on the planet is only separated from any other person by only six degrees of separation. In other words, anyone on earth can be connected to any other person on the planet through a chain of acquaintances that has no more than four intermediaries. This is sometimes referred to as the small-world phenomenon and underscores the connected nature of the human world. While the six degrees of separation concept has never been proved absolutely, the concept of the small-world phenomenon has played a valuable role in developing reliable communication protocols for the internet and ad-hoc wireless networks. Related sites
Activities
Further readingAustralasian Science April 2006, pages 26-29 Can science foretell the future (by Julian Cribb) Reports on the analysis of thresholds to predict Earthquakes, floods, stockmarket crashes and other major events.
June 2005, pages 32-34 The science of complex systems (by John Finnigan) Explains how scientists are probing the complex interactions that influence the behaviour of bushfires, cyclones and the stock market.
Business Week 17 May, 2002 Stephen Wolfram's simple science (by Michael Arndt) An article about Stephen Wolfram and his idea that all phenomena are programmed by basic rules.
Cosmos Nature 29 January 2004, page 399 Engineering complex systems (by J M Ottino) Looks at the role of engineers in the development of complex systems.
New Scientist 9 August 2008, pages 28-31 Why complex systems do better without us (by Mark Buchanan) Proposes that complex systems are more efficient if allowed to self-organise.
5 April 2008, pages 28-31 Will a pandemic bring down civilisation? (by Deborah MacKenzie) Raises the question of whether a pandemic could cause social collapse
5 April 2008, pages 33-35 Why the demise of civilisation may be inevitable (by Deborah MacKenzie) Discusses the vulnerability of civilisation due to its complexity.
4 July 2006 The net reloaded (Kim Krieger) Looks at the 'scale-free' network theory.
26 February 2005, pages 32-35 Too much information (by Mark Buchanan) Discusses cellular automata and the ability to predict events.
7 August 2003 Email experiment confirms six degrees of separation (by Will Knight) Describes the email version of the experiment that demonstrated the six degrees of separation.
13 April 2002, page 24 All the world's a net (by David Cohen) Describes a variety of networks that are complex systems.
Science 2 April 1999 This special issue is dedicated to complex system science.
2 April 1999, page 79 Beyond reductionism (by Richard Gallagher and Tim Appenzeller) Provides an introduction to the technical articles in the special issue.
Scientific American March 2006, pages 54-61 The limits of reason (by Gregory Chaitin) Provides an historical background to complexity and a discussion of mathematical axioms.
May 2003, pages 50-59 Scale-free networks (by Albert-László Barabási and Eric Bonabeau) Provides an overview and potential implications of scale-free networks.
The Chronicle 10 May 2001 Complexity research and its challenge to other disciplines A discussion about how complexity theory can reshape the way scholars examine evolution, economics and many other fields.
The Economist 9 October, 2003 Modelling complexity Looks at agent-based networks, including financial markets, crime networks and commercial supply chains.
Useful sitesAustralian Research Council Complex Open Systems Research Network
New England Complex Systems Institute, USA
Insights from complex systems (Serendip, USA)
Provides an eleven point summary of the features of complex systems.
Mathematics in environmental science (Macquarie University, Australia)
Discusses the use of mathematical models in environmental science.
Resilience (Resilience Alliance, USA)
Defines resilience and how it may be altered in ecosystems.
Cellular automaton (Mathworld, Wolfram Research, USA)
Provides examples of simple mathematical rules that generate complex outcomes.
Life’s games (University of Queensland, Australia)
Describes the rules of the ‘Game of Life’ cellular automaton.
Australian Broadcasting Corporation
Glossaryecosystem. A term used to encompass all the organisms in a community together with the associated physical environmental factors with which they interact (eg, a rockpool ecosystem, a forest ecosystem).epidemiologists. Researchers who study diseases or conditions in human populations and the factors that influence their incidence and prevalence. equilibrium. When a reaction and its reverse occur at equal rates, they effectively cancel one another, so there is no net change. feedback. The process whereby the output of a system affects the input. Positive feedback reinforces or increases something; negative feedback acts to keep a process within certain limits. Positive feedback can work in systems by amplifying a very small effect, changing the previous equilibrium. metabolic pathways. A group or series of chemical reactions occurring within a cell, catalysed by enzymes. Pathways can breakdown compounds to yield energy, or involve the step by step modification of an initial compound to create a new product. model. Solving complex problems associated with real situations is often made easier by setting up a model of the situation – a mathematical description of the problem. To set up a model, a problem is simplified and only those aspects that can be represented mathematically are included. After the problem is solved mathematically, tentative solutions are translated back to the real situation, as possible real solutions. At this stage the inadequacy of the simple model may be revealed, and some parts of the process may need to be changed. More information on models and modelling can be found at What is modelling? (Nova: Science in the news, Australian Academy of Science). non-linear. For non-linear systems, a small perturbation may cause a large effect, a proportional effect, or even no effect at all – the behaviour of the system is not simply the sum of its parts. In linear systems, effect is directly proportional to the cause. Many systems are best represented by non-linear equations that are difficult to solve, but can give rise to interesting phenomena. scale-free networks. A network pattern commonly seen in living systems that has some nodes with many links, many nodes with a few links, and the remaining nodes lying somewhere in between. In this system, known as a scale-free network, there is no clear average number of links per node. Scale-free networks are resilient structures because the random removal of any particular node is unlikely to stop the network from functioning. On the other hand, if a node with many links was targeted and removed it could create a large system-wide disturbance. For more information, see Scale-Free Networks (Computerworld, USA). In some networks the nodes are connected randomly, in others each node has a fixed number of links to adjacent nodes, giving rise to a regular pattern. But most systems observed in nature fall somewhere between these two extremes.
External sites are not endorsed by the Australian Academy of Science. Posted October 2006. The Australian Foundation for Science is also a supporter of Nova. This topic is sponsored by the ARC Complex Open Systems Research Network.
© Australian Academy of Science |