Primer Home /
Emergence / Emergence, the Offspring of Complex Systems
Emergence, the Offspring of Complex Systems
Topic: Emergence
by Pandu, 2019 Cohort
Imagine you’re a biologist trying to work out why an ant colony forms the way it does. Perhaps you isolate a few ants to study in isolation. Watching their behaviour reveals a set of basic instinctual actions that goes little way towards explaining complex behaviour of a large colony.
When looking at complex problems, it is often the case that its properties can’t be reduced to the sum of its constituents. This is the basis of emergence.
Emergence is a feature of many complex real-world systems, where system behaviour can’t be predicted by the behaviour of its parts. Examples include hurricanes, bird flocks and consciousness. These systems exhibit emergent behaviour resulting from their parts, such as air particles, birds and brain-cells respectively. In these cases agents’ local behaviour interact to produce system behaviour irreducible to the agent level.
While most of these examples are decentralised systems, it is important to note that top-down hierarchical systems, such as bureaucratic organisations, can also produce emergent behaviour.
When dealing with a complex system, recognising and looking for emergence can assist in understanding it. It is tempting to try resolving systems into a linear sum of parts because these are easy to understand, represent and perform calculations on.
Complex systems are often impossible to simulate with computers. Many parts lead to a staggeringly- huge number of interactions. We can look at complex systems instead as computational machines themselves, with emergent behaviour the solution to its parts’ interactions. This allows a degree of detachment between part and system behaviour. The problem can be looked at in separate scale levels: the less complex parts and the more complex whole. In an ant colony this would be the individual ants and colony-wide behaviour
Emergence contradicts reductionist philosophy. This is the philosophy that a phenomena is solely the sum of its parts. A reductionist explains temperature as solely an object’s atoms’ kinetic energy. How well does this apply to phenomena such as consciousness and personality? Can we explain someone’s personality by inspecting underlying neurones and neurotransmitters?
We look at the behaviour of South American army ants as an example. It is a complex system with many agents (a colony can contain half-a- million ants) and connections (ranging from chemical signals to physical interactions). They differ from common ants in that the colony is nomadic, and moves like a river cross the forest floor. Attempting a computer simulation may prove futile, as for every ant there is an exponential number of connections. Rather than trying to break down the colony’s behaviour to the ant level, we can reason about it easier if we understand that it only exists at the macro
Not all emergent behaviour is unpredictable: we could infer that the colony shares some goals similar to its constituent ants, such as searching for food (the army ant colony consumes any prey, unluckily, found in its path). However the living, coordinated protective ball used to shelter the Queen and larvae at night couldn’t be easily predicted from close examination of an ant.
Through the lens of emergence we can unravel complex societal problems, such as traffic jams, epidemics and asymmetric wars. Accepting that phenomena occurs on different levels of scale allows us to address each level differently, rather than focusing on breaking down high level behaviour to individuals.
Explore this topic further#
Return to
Emergence in the Primer
Disclaimer#
This content has been contributed by a student as part of a learning activity.
If there are inaccuracies, or opportunities for significant improvement on this topic, feedback is welcome on how to improve the resource.
You can improve articles on this topic as a student in "Unravelling Complexity", or by including the amendments in an email to: Chris.Browne@anu.edu.au
Imagine you’re a biologist trying to work out why an ant colony forms the way it does. Perhaps you isolate a few ants to study in isolation. Watching their behaviour reveals a set of basic instinctual actions that goes little way towards explaining complex behaviour of a large colony.
When looking at complex problems, it is often the case that its properties can’t be reduced to the sum of its constituents. This is the basis of emergence.
Emergence is a feature of many complex real-world systems, where system behaviour can’t be predicted by the behaviour of its parts. Examples include hurricanes, bird flocks and consciousness. These systems exhibit emergent behaviour resulting from their parts, such as air particles, birds and brain-cells respectively. In these cases agents’ local behaviour interact to produce system behaviour irreducible to the agent level.
While most of these examples are decentralised systems, it is important to note that top-down hierarchical systems, such as bureaucratic organisations, can also produce emergent behaviour.
When dealing with a complex system, recognising and looking for emergence can assist in understanding it. It is tempting to try resolving systems into a linear sum of parts because these are easy to understand, represent and perform calculations on.
Complex systems are often impossible to simulate with computers. Many parts lead to a staggeringly- huge number of interactions. We can look at complex systems instead as computational machines themselves, with emergent behaviour the solution to its parts’ interactions. This allows a degree of detachment between part and system behaviour. The problem can be looked at in separate scale levels: the less complex parts and the more complex whole. In an ant colony this would be the individual ants and colony-wide behaviour
Emergence contradicts reductionist philosophy. This is the philosophy that a phenomena is solely the sum of its parts. A reductionist explains temperature as solely an object’s atoms’ kinetic energy. How well does this apply to phenomena such as consciousness and personality? Can we explain someone’s personality by inspecting underlying neurones and neurotransmitters?
We look at the behaviour of South American army ants as an example. It is a complex system with many agents (a colony can contain half-a- million ants) and connections (ranging from chemical signals to physical interactions). They differ from common ants in that the colony is nomadic, and moves like a river cross the forest floor. Attempting a computer simulation may prove futile, as for every ant there is an exponential number of connections. Rather than trying to break down the colony’s behaviour to the ant level, we can reason about it easier if we understand that it only exists at the macro
Not all emergent behaviour is unpredictable: we could infer that the colony shares some goals similar to its constituent ants, such as searching for food (the army ant colony consumes any prey, unluckily, found in its path). However the living, coordinated protective ball used to shelter the Queen and larvae at night couldn’t be easily predicted from close examination of an ant.
Through the lens of emergence we can unravel complex societal problems, such as traffic jams, epidemics and asymmetric wars. Accepting that phenomena occurs on different levels of scale allows us to address each level differently, rather than focusing on breaking down high level behaviour to individuals.