It’s fascinating to learn early on in the book that those who in the 20th century worried about the application of a mathematical technique like statistics to the human populace had got things entirely back-to-front. Statistics originated as a collection of information on people, a crude form of census and developed into a mathematical discipline, rather than the other way round.
It’s a big book and it’s necessary to bear with Philip Ball through the rather (aptly?) ponderous chapter on Hobbes’ Leviathan up front, but once he gets into statistical physics he takes off.
There’s a lot on economics, on political power, globalization and even the Internet. Again and again the book comes back to the way that mass human action has some resemblances to the physics of large quantities of interacting objects. In physics this has produced a lot of theory based on statistics that does very well at predicting what will actually happen. When it comes to the human world, not entirely surprisingly, things are more complicated. Not only are most human masses not closed systems – so you have to take into account the impact of external forces – but a single individual can have a huge impact. When you are looking at gas molecules you aren’t going to have a Jesus or a Hitler – we, on the other hand, can expect that.
Because of this disparity, there are always problems with using the methods of statistical physics to make predictions. Ball spends ages describing how different models can be built, but often then has to come to the conclusion that while they can explain a lot that has happened, they aren’t much use at predicting the future – which is what we really want them to do. (One scientist Ball quotes did dare to make a prediction based on his model, that the UK housing price bubble would burst by the end of 2003 – while it will inevitably come, we’ve reached June 2004 without it happening.)
This gives us one of the two big problems with this book, and the reason it doesn’t score more than three stars. It is a great idea for a book, but everything’s really a work in progress. There are few conclusions it just goes on. And there’s the other problem. It goes on, and on, and on. There seems to be a bit of a Harry Potter phenomenon occurring with popular science books (it’s probably following a good statistical pattern) – the incidence of over-long books is on the rise. At around 640 pages, this book was twice the length the content deserved and in the end it was hard not to start skimming the material.
There is also one striking omission. Ball several times refers to fiction and speculative writing in considering the application of maths to mass human behaviour, but strangely never mentions Isaac Asimov’s remarkable 1950s Foundation trilogy, which features “psychohistory” a concept relying on a vast mathematical model of human space. While Asimov’s idea is not practically possible, neither are many of the others that Ball mentions, and nothing else has quite the magnificent sweep of Asimov’s confection.
Despite all this, it’s a fascinating subject and often Ball makes his points well, it’s just a shame he’s made it such a slog. If you’d like an easier time of it, try the overlapping and much better Sync.
Paperback: 
Using these links earns us commission at no cost to you
Review by Brian Clegg
Comments
Post a Comment