Primer Home / Emergence / Summary on emergence

Summary on emergence

Topic: Emergence
by Dmitry, 2018 Cohort

Note: This entry was created in 2018, when the task was to “summarise a key reading”, and so may not represent a good example to model current primer entries on.

Emergence is when the behaviour of a bunch of parts together is hard to understand by examining the parts alone, but the picture becomes clearer when you look at them as a whole. Melanie Mitchell considers the existence of emergent properties to be a defining feature of complex systems. That is, without emergent properties you can always understand a system by decomposing it into parts, and so the system might be complicated, but it does not deserve the label complex.

As an example, you’d be hard pressed to predict the intricate shapes of snowflakes by looking only at the simple hexagonal patterns that water molecules form when they crystallise into snow. Thus, we can call the beauty of snowflakes emergent, because it is the result of interactions among many parts, and it’s not obvious how the result comes from the behaviour of the parts. In contrast, if you look at a salt crystal, you can easily predict the shape (a cube) from the behaviour of the atoms (arranged in a rectangular grid). The shape of a salt crystal is not emergent (or complex).

Another good example comes from nature. Ant colonies can explore and find food in complex environments, even though individual ants have very simple behaviour. If there was food right outside the colony, and each ant had the behaviour go out, pick up food, bring to colony, it would be obvious (and not emergence) that the colony’s food stockpile increased. But the ants can find food no matter where it is hidden away. I’m sure if my keys were made of honey, ants would be better at finding them than I am. And that is an emergent property.

Notice that nowhere in these examples is there some magical force causing the emergent properties. They do result purely from the behaviour of the individual parts if you simulated all the parts in detail, you would see the emergent property also. The word emergence just captures the idea that it’s hard to predict the result of that simulation before doing it, because it depends in a complex way on the interactions between these parts.

With the concept of emergence in hand, you can look at a system and say This doesn’t behave at all like what I’d expect, given the parts it’s made of! Maybe that means if I want to understand it, I have to zoom out a bit and look at the interactions between the parts. Often that means working across disciplines: getting all the people who understand the different parts in a room together so they can talk about the interactions.

It’s nice to study parts in isolation because it’s simpler and you can learn more and faster. But to understand emergent properties, that doesn’t cut it. And that is one of the key ideas of complex systems science in general.

References / Extra Readings#

The core reading for this topic is Chapter 1 of Complexity: A Guided Tour by Melanie Mitchell. It mentions emergence only in passing, but gives a very clear and simple definition, and the rest of the chapter provides nice examples.

Chapter 1 of Making Things Work by Bar-Yam and Chapter 2 of Seeking Sustainability in an Age of Complexity by Harris both have whole subsections on emergence. In my opinion, the former discusses the concept more clearly.

P.S. If you read more about emergence, you might notice some debate over whether an emergent property can always be understood from the behaviour of parts in principle (though maybe not in practice). My opinion comes down very strongly on the yes side behaviour that is not fully reducible to the parts violates all known laws of nature, and there is no evidence of any phenomena where we have to posit such things. Saying we can’t understand it because it’s too hard seems to always be enough, and it is not necessary to add and nobody could ever understand it even in principle. Of our readings, Mitchell and Bar-Yam define emergent properties as hard to understand from the parts, while Harris defines them as impossible to understand from the parts.

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

bars search times arrow-up