Ubiquity - Mark Buchanan
Russell Degnan

I was really looking forward to this book. It had not been on my bookshelf long before I read it, but I thought it would be of particular interest. In a sense it was, but it is also an idea in search of a thesis; roughly summarised by the authors own statement near the end of the book:

"But from another more abstract point of view, great wars may take place simply because the collective attitudes, ideas and behaviours of the mass of humanity are subject to the same wild fluctuations of the magnet poised between its magnetic and its non-magnetic phases. It goes without saying that nothing I have mentioned in the past few chapters proves this. The 'take-home' message is simply that this is a real possibility."

The research has not been done, if anything it is a call to do it. Hence the book contains two parts. The first, an occasionally frustratingly shallow look at 'historical science' that look at systems that grow and evolve, where each previous state depends on the previous arrangement of an enormous number of elements. States that are endlessly unpredictable in their consequences, because of their size and chaotic nature, while at the same time having a basic mathematical pattern in their fundamental properties. In particular, the tendency to periodically 'fail' - sometimes spectacularly.

Buchanan's book then goes off the path by focusing on the 'games' that physicists use to describe these systems. He constantly tries to draw the line between these and 'real' systems ignoring the most important fact that in order for a model to be of use they must be useful for prediction. They aren't, not least of which because it is probable that these systems are not, but also because it is very easy to create a game that will act like a system in a critical state if your goal is to create a game that will act like a system in a critical state. And many of the games Buchanan describes seem to do precisely that.

The second part of the book calls on social scientists to once again plumb the depths of scientific analogies by using similar models to understand their own fields. It is an area that has often proven fruitless if not disastrous; that is not a reason not to do so, merely an observation. Again however, Buchanan is short on detail and long on speculation. Far too often he draws a conclusion that can roughly be described as "because this system has a failure rate similar to a power law, it must be in a critical state, and moreover, massive failure may be an inevitable part of the system rather than because of a massive cause". Be that effect warfare, the size of cities, or the collapse of financial markets.

Buchanan makes the - correct I think - claim that major wars are no different from minor ones. The difference is that in a major war, such as the First World War, a large number of actors are poised in the critical state and can set off a chain reaction of events. However, he comes dangerously close to supposing that because noone can predict when a small conflict will become a major war that a major war could occur at any time and at any place. This is incorrect. Under his theory, a large war supposes a long chain of interconnected components. Western Europe is not likely to explode into any sort of war of any size at the moment, even if the Middle East could. If the historian's task is to enumerate a large number of small componenets in a large conflict instead of a small number of large components then that is what they must do. There is no effect without some cause, no matter how numerous or small those causes might be. A similar argument must hold for each of the other areas Buchanan discussed.

Having said that, I still think this book has something to add, if only as a signpost for where something might lie. Unfortunately it was not until the very end that Buchanan gave the hint of where that value may lie:

"Imagine that you lived in that magnetic world, and that its temperature was held below that critical point. [...] The history of your world would have fixed laws and undending peace and lack of change stretching back through time. [...] [T]here would be no history at all, since a record of unbroken sameness is not history, but the lack of it.

"Now suppose that the temperature were raised well above the critical temperature. [...] [T]he past would be a wholly senseless record of absolutely random changes. Again, a boring world, since the only thing you could say of it is - it is random.

"How much more interesting things would be if the temperature were brought close to the critical point. [...] [Y]ou would find a fascinating pattern that had structure and randomness all mingled in some perplexing but intoxicating way."

It has been sometimes supposed, that the second law of thermodynamics - that all systems will tend towards entropy - would suppose that life on earth is impossible, or at least so unlikely as to be nearly so. This is the first "no history" side of the coin. A bouncing ball that bounces a little bit less each time until it bounces no more. An organism that dies precisely after it is comes into being. No growth. The other side of the coin is a world where so much energy is available no patterns occur at all. The world where atoms form into an infinite variety of combinations but are blown apart before they can do anything with it.

There are two elements to the critical state that are vitally important here. The first is that the critical state is necessary for any form of growth to occur at all in any organic system - and earthquakes aside, it is organic systems that are most in need of a better understanding. Without it there will be either nothing or no organisation. History would not be boring, it would be non-existent. To look at the social sciences, in the Ascent of Man Bronowski looks at the remaining nomadic tribes in Iran and describes them as being the same, from generation to generation, wach following the last. This is that sameness in action, a point of equilibrium that will never move without an external push.

The second is that failure - not success - is the predominant theme in these systems. Rocks could potentially move constantly if not for friction. The stresses that build up before an earthquake are the results of those failures, that explode in sudden pockets of success before returning to the way they were.

Similarly, an economic system is one where we constantly fail to trade. In any day, I reject an almost infinite number of potential transactions, because of distance, because of a failure of my demand, of their supply, or because we couldn't agree on an acceptable bargain. When people talk about 'market failure' it is a misnomer, the market fails to transact almost constantly. What appears to be market failure or a collapsing economy is in fact the destruction of previously viable pockets that - like rocks in an earthquake zone, or particles in a particle accelerator - had actually been moving.

These small pockets of success that occur seemingly randomly, but grow and expand on each other are the key to understand the vast range of natural phenomenen - including social phenomenen - that are described in the book. It may, in the future, be an interesting and fruitful area of research, but there is a lot of work to be done yet.

Finer Things 13th December, 2004 22:01:07   [#]