Saturday, May 27, 2006

Critical Mass

One of the few authors not mentioned in Philip Ball's Critical Mass is Isaac Asimov. However, it was in his Foundation Trilogy that I first came across the thesis that animates Ball's book and the occupies many of the finest minds of today: that given a statistical sample large enough, you can not only understand but even predict the behaviour of large groups of people. Now Asimov went just a touch further; in his version, the fate of the galaxy and its thousands of planets was forecast over a stretch of centuries. Ball's ambitions are more modest.

It is a fascinating intellectual journey, this birth of the new science of Social Physics. As the name implies, it is the attempt to understand social phenomena by means of techniques and concepts more usually employed in Physics. One of the most important of these concepts is that group, or mass behaviour cannot always be understood or predicted by 'scaling up' the individual. As readers of Elias Canetti's Crowds and Power will recall, the pack obeys the laws of the pack, an organism in which the multiplicity of individual desires is subsumed into a simpler and more predictable paradigm. So too here. Physicists long ago abandoned the attempt to understand the behaviour of a gas by calculating the movements of individual molecules and concentrated rather on averages in the movement of millions. Now that the required huge numbers can be studied in human affairs, here as well the picture is simpler and more comprehensible than had once seemed possible.

There are new concepts aplenty for the neophyte: phase transitions, self-organising patterns, scale-free networks and power laws. That last one is less and more intimidating than it sounds. It is a mathematical term that concerns the relationship between two numbers. Each time one of the numbers is doubled, for example, the other number will increase by a fixed rate, which may be 4 times its value, or 8, or 16. This is of importance because these power law values are found everywhere; in nature, at the critical point of transition between a gas and a liquid, and in human affairs, such as market fluctuations and the size of companies. It means that you don't have a nice even rate of increase. No, the largest companies are vastly bigger than the next largest. Plotted on a graph, the closer you get to the top, the more vertical is the line. It seems to be the same with wealth. Here, too, the richest are almost infinitely richer than those just behind them in the rankings. Furthermore, it seems to be have been as true 3,500 years ago in Egypt as it is now. You will recognise Pareto's Law here, that 80% of the wealth will be held by 20% of the population, proportions that seem to remain constant in different types of economies as well as in different epochs. Here you face the great question: is this how it was 'meant to be'? Ball, when he talks about politics, is rather wishy-washy and doesn't try to give a definitive answer, apart from pointing out that we must differentiate clearly between what is, and what possibilities there are to change that. Very sensible, but not very helpful.

That last example will give you some idea of how explosive some of this number-crunching may become. There's a lot of it here, as well as innumerable models and graphs. It may get a bit much (it did for me). However, I know that this will be a book to return to as one of the many topics it covers pops into the news. Its range is vast, and the potential of the studies it dips into has been barely tapped.

No comments: