Books We Loved

'Critical Mass', by Philip Ball

Ever wondered whether it's possible for the good guys really to end up on top? - or will double crossers always do better? Ever thought it wierd that sometimes on a motorway the traffic grinds to a halt, even though when you finally get moving there seems to have been nothing to cause it? Ever wondered whether house price crashes are bound to happen eventually? All of these questions and more are covered by the newly revitalised idea of developing a 'Physics of Society'. That is what 'Critical Mass' by Philip Ball is about. How far can you get in understanding Society by using Physics?

Bouncing Balls

It boils down to the question of what happens if you take the tools that Physicists have developed over the years and apply them to people instead of to physical particles? It turns out that lots of the behaviour social scientists see as they study 'society' has remarkable similarities to the way, for example, that solids turn into liquids and gases. Why? The former is group behaviour of individual people interacting with each other, the latter is the group behaviour of molecules interacting with each other. Simplistic as it seems, some of the early fundamental breakthroughs in understanding the physics of matter came from treating molecules as though they were just millions of Billiard Balls bouncing around. They are far more complex than that of course, but from simplifications like that and asking what the average behaviour of millions of them was like, a deep understanding of solids, liquids and gases was obtained. Perhaps a similar approach can help us understand the complexities of society: from the economics of global markets to the way panicing crowds behave?

Worlds within Worlds

The book is not just about Physics and the Social Sciences though. It doesn't get an explicit mention but computer science runs through the whole story too. The book is all about how Physics is revolutionising the Social Sciences, but it has only come about because ideas from Computer Science already changed the way physicists do Physics. The Physics of Society is about 'computational modelling' and that is Computer Science through and through.

What is computational modelling? It is where you create a virtual 'micro-world' of individuals each following some set of rules. Social scientists might think of them as a whole range of things from cars travelling along roads, market traders deciding when to sell, or nation states deciding whether to unleash a holocaust on their neighbours. It depends on the rules. The physicists call them particles. Computer Scientists? They call them 'agents'.

Start the clock

How do micro-worlds work? Time runs in a series of ticks, and on each tick the world moves on - each agent following its rules, whether it be to move forward a little (if the world was of cars as agents), sell (if the world was one of market traders) or attack a neighbour (if the world is of nation states as agents). By settiing the worlds running for thousands of ticks and then doing that over and over again you can look for the similarities across all the different runs. You can also look at how sensitive it is to changes in the rules or the starting conditions...You can start to tease out the laws behind how the patterns in the behaviour of the many results from the interacting behviour of the individuals.

Complexity from Simple things

The rules each agent follows are simple, even simplistic, but that is the point. When they all start to interact, often new behaviour emerges: behaviour that is nothing like that of the individuals themselves. That is where it gets interesting, because often complex, unpredictable behaviour emerges; behaviour just like the behaviour that social scientists see in the real world. When the same patterns are seen over and over across different scenarios it suggests something fundamental is going on. It certainly shows that you do not need to look for complicated individual behaviour to find the causes of complex group behaviour.

The book shows that laws do emerge from the group behaviour we call Society, and that Physics and Computational Modelling can help us to understand them. What the book also makes clear is that working out what to do with the knowledge needs not just a Physics of Society but an understanding of morality too.

A Physics of Society certainly helps, but the laws of morality are a whole new Ball game.