Customer Reviews

Ubiquity: Why Catastrophes Happen

5 star:		(11)
4 star:		(5)
3 star:		(1)
2 star:		(3)
1 star:		(1)

 

 

Who hasn't wondered why catastrophes happen, and if they can be predicted or avoided? Economists and investors try to understand why markets crash, seismologists struggle to understand and predict great earthquakes, and historians speculate why empires crumble and global cataclysms such as the First and Second World Wars occur.

Physicist and science journalist Mark Buchanan brings the science of what he calls "historical physics"--the study of systems that are far from equilibrium and, as he puts it poised "on the knife edge of instability" to bear on these questions.

He describes a much-studied model of such catastrophe-prone systems, a simple sandpile. Build a sandpile by dropping one grain at a time on the top of the heap. It will eventually reach a critical state at which a grain can either make the pile a bit taller or start an avalanche, small or large. Scientists experimenting with real and virtual sandpiles have observed several important regularities:

1. The time between avalanches is extremely variable, making it essentially impossible to predict when the next avalanche will occur.

2. The size of avalanches is also extremely variable, making it essentially impossible to predict whether the next avalanche will be tiny or huge.

3. A big avalanche doesn't need a big cause; one grain can trigger a sandpile-flattening event.

4. Avalanche sizes follow what mathematicians call a power law. What that means is that large events happen less frequently than small ones according to a fixed ratio. For sandpiles the frequency goes down by a factor of 2.14 for each doubling of avalanche size. For earthquakes the frequency goes down by a factor of four for each doubling of released energy.

5. Any process that follows a power law shows two key features. The events are "scale invariant," meaning that no particular size of event is favored. And large events--big avalanches, 8.0 earthquakes, "1000-year floods" and many other kinds of catastrophic events occur far more frequently than common sense would suggest.

We tend to assume that events distribute themselves along the familiar normal curve--like height, weight, IQ scores, etc. These distributions do have a favored scale--most people cluster around the average height, weight, or IQ, while the number of people with extremely low or extremely high scores is very small.

Buchanan shows that many events that greatly impact our lives represent changes in sandpile-like systems, and so are not just hard to predict, but inherently unpredictable. The one thing that can be predicted is that huge events will occur far more often than our intuition prepares us for.

Many natural events follow power laws, including earthquakes, forest fires, floods and the mass extinctions that have punctuated the history of life on earth. And many human events also show these regularities, including traffic jams, market crashes, the collapse of nations and empires, and wars.

Buchanan's presentation of these regularities and their implications is well reasoned, well documented and well written. Read it for yourself, and see if the ideas he presents don't help you to understand what seems to be a profound pattern that underlies many of the events that shape and shake our lives.

Robert Adler, author of Science Firsts: From the Creation of Science to the Science of Creation (Wiley & Sons, September 2002).

This is not a hard book to read, but it is difficult to integrate into the way you look at the world. Mark Buchanan is a science writer who has worked on the editorial staff of Nature and as a features editor New Scientist. In this book he is writing about the development of a growing field of physics - complexity. Complexity is chaos in critical states. A critical state exists in a system that is not in equilibrium. You may have heard of the "butterfly effect". That is, there is a possibility that a butterfly flapping its wings in South America can cause a storm in Europe weeks later. However, that same butterfly can flap all in wants inside a closed balloon with no effects, other than maybe slightly increasing the temperature of the air in the balloon. The air inside the balloon is in equilibrium, even though the molecules exhibit chaotic behavior. The atmosphere is in a critical, i.e. non-equilibrium, state. A small perturbation somewhere can lead to very big changes.

If the air inside the balloon is in equilibrium, its past, present and future are all the same. It has no "history". When things are in non-equilibrium, history matters since what happens now can never be washed away but affects the entire course of the future.

The applications of this model extend from the piling of grains of sand in an hourglass to economics.

"Despite what scientists had previously believed, might the critical state in fact be quite common? Could riddling lines of instability of a logically equivalent sort run through the Earth's crust, for example, through forests and ecosystems, and perhaps even through the somewhat more abstract "fabric" of our economics? Think of those first few crumbling rocks near Kobe, or that first insignificant dip in prices that triggered the stock market crash of 1987. Might these have been "sand grains" acting at another level? Could the special organization of the critical state explain why the world at large seems so susceptible to unpredictable upheavals?

A decade of research by hundreds of other physicists has explored this question and taken the initial idea much further. There are many subtleties and twists in the story to which we shall come later in this book, but the basic message, roughly speaking, is simple: The peculiar and exceptionally unstable organization of the critical state does indeed seem to be ubiquitous in our world. Researchers in the past few years have found its mathematical fingerprints in the workings of all the upheavals I've mentioned so far, as well as in the spreading of epidemics, the flaring of traffic jams, the patterns by which instructions trickle down from managers to workers in an office, and in many other things. At the heart of our story, then, lies the discovery that networks of things of all atoms, molecules, species, people, and even ideas have a marked tendency to organize themselves along similar lines. On the basis of this insight, scientists are finally beginning to fathom what lies behind tumultuous events of all sorts, and to see patterns at work here where they have never seen them before."

The mathematical models of this science don't really exist yet, and may never exist. We have empirical observations and we have games. The empirical data suggests that all these phenomena follow a power curve, and all with roughly the same shape. For example, looking at earthquakes, as the strength of the earthquake doubles, the frequency of occurrence drops by one fourth. This simple rule seems to apply to many examples.

So what does this have to do with creativity, strategy, leadership and innovation in organization? Well, I'm not sure yet. My intuition tells me that this is very important to those concepts. It may help us understand the frequency of occurrence of breakthrough ideas and innovation. It may help explain why some innovations cause such change and others do not. It may help produce better strategies to deal with chaotic and unstable markets. And, it may provide lessons for leaders in chaotic times.

This is the book that I would like to have written. Although being a popular account, it is scientifically accurate and carefull in its suggestions, always informing the reader what is consolidated science and what is scientific speculation.
In contrast to a previous review, I have read all the pages of this book. Since I am a physicist working in this very subject (self-organized criticality), I probably can say that if someone use the example of a Gaussian (bell shaped curve) to illustrate that the power laws discussed in the book are trivial, well, this person have not understood anything.
Gaussians have exponential decays, so they predict that very larg events (catastrophes) will occur with vanishing probability. For example, the heigh of people is distributed as a Gaussian. What is the probability of finding a 3 meter person?
Zero.
Distributions wich have power law tails, depending on the power exponent, may have no well defined variance or even average value. This means that there is no "average" earthquake, and that very big earthquakes (or other cathastrophes) are not "acts of God" but have a no desprezible chance of occur due to simple chain reactions of events.
I have introduced my students to ideas like critical states and modern physical thinking by using this book. So, I can recommend it to any reader without reserve. The emphasis by the author that critical chain reactions of events must be accounted by any view of History and Society is an important mind tool in our increasing interconnected (and, because it, prone to global chain reactions) world.

One of the ways I judge a book's worth to me is how often I think about it during the day, and in the days after I've finished it, and if it has added a valuable viewpoint reference. This book has scored high on that scale. This is a book that will present a new way of viewing, something to think about, and I think most would enjoy reading it. The first reviewer had an excellent synopsis of the substance of the points of the book.

I am intrigued by systems thinking and explanations, a way to distill the random patterns of whatever a person is dealing with in their daily life, from the rare to the mundane. I was fascinated with the modeled games presented and how they illuminate the heart of the underlying mathematical "engine" that permeates our world. To see these findings correlate with the recent face of physics theories was compelling.

I had recently finished "The hole in the Universe" by K.C. Cole, and thought about the condition of our universe being in a particular "frozen" state of conditions and how it's possible that that state could be "kicked down" to another "rung" on the ladder-not predictable as to when or under what conditions. It seemed to be a reiteration of these principles, seen on an immense scale.

This picture could be disconcerting as to the randomness of chaos potential, but at the same time it presents a view of the dynamism of life and possibilities, how it can't be any other way if we are able to move and have effect and not live in a static world.

How does this affect one's world view? It's the same as if you haven't read the book; navigate through as best you can, and appreciate life.

Its an interesting read. The reason I didnt give it 5 stars is that I have already read one of Marks previous books (Nexus) which has some overlap (not a lot) with this book. In fact it would be beneficial to readers to read the Nexus book before reading this one as what he writes about in that book really helps to understand this book.

I was really hoping for some more answers on how to predict things based on what Mark talks about but that is the essential outcome of the book, you cant predict things!

I found this book incredibly boring. OK, I know this goes against the grain of other reviewers here. And I'll admit I'm only through the first 85 pages. But I already have that Ayn Rand feeling that the entire book is just going to rehash what's already been said.

I agree there are some interesting ideas, basically that we can't predict stuff very well. But here are a couple of examples of where Buchanan makes me suspicious that he really has the "Ph.D. in theoretical physics" stated on the back. The most egregious example so far, I think, is his statement on p. 85 that "take some really small number, such as .0001, change it by 10 percent, or even multiply it by 2 or 10 or 100 and you still have a very small number."

This, coming from a guy who has written page after page about scale invariance, seems just ludicrous. What on earth can he possibly mean by this nonsense??

Another example is his discussion of getting the friction issues wrong with the sliding blocks. Then he says, hey, but what about heat? (Bottom of p. 59). As if this great insight takes care of the problem of friction. How can a Ph.D. physicist make the mistake of thinking that heat is a new way of dealing with friction (duh?)

Another is his comment (bottom of p. 80) that the use of constants in the great differential equations of physics is some way mitigates the problem of tuning the blocks. He gives "c" in Maxwell's Equations and "h" in Quantum Mechanics, as examples of these. He writes "almost every good theory in the world has some numbers in it that have to be tuned to make the theory fit reality." But the "tuning" he's talking about with the sandpile and other games, has to do with the basic structure of the differential equations. "c" is a RESULT of Maxwell's Equations, not some "tuning" factor. It is true that the existence of Planck's constant is a fundamental feature of the equations, but its VALUE is simply a number that makes experimental observations work in SI units. Now Einstein's "cosmological constant" is much more like what Buchanan is talking about. But by this time one wonders if he really has a point here, or is just rambling on to cover up his hand-waving, and hoping he can get his book sold.

In my opinion, here is what this book is saying:

Let's take the example of the average temperature for my city on a given calendar date. The facts are these: the temperature over history for that date follows a bell-shaped curve. There IS a typical temperature. But the VARIATION of the actual temperature TODAY (i.e. a particular day) from the typical temperature is NOT very easy to predict. In fact, there is no TYPICAL VARIATION.

All we can say is that most days will have a small variation from the norm, and fewer days will have a larger variation from the norm. The "power law" concerns this variability. The larger the variability, the less likely it is to occur. If the average over time is 65 degrees, a lot of days in history will have had an average of 66 degrees on that date, and only a very few will have an average of 76 degrees on that date. Why is this fact worth writing an entire hold-your-breath book about?

Most days have no earthquakes. But when an earthquake DOES occur, we can't predict how big it will be. All we can say is that there will be more small earthquakes than large ones.

Well, duh!!

Yawn.

I read this book cover to cover. This book has too many pages for little content. The thesis of the book could have been explained much more economically. One other problem I have with this book is that it is superfluous if you have read or own a copy of Per Bak's How Nature Works: The Science of Self-Organized Criticality, just that Bak's book is out of print and expensive.
The thesis is that everything in the world (almost everything) is in a self-organized critical state, therefore prone to upheavals caused by minor disturbances, and the latter is logical equivalent of a power-law. Well, let us call "everything is in a self-organized critical state" statement A, and "something follows a power-law" statement B. A straightforward mathematical reasoning can show us that A results in B, or A->B. The book is not able to show us any A->B save for the sand pile games (there are maybe hundreds of its variants in the physics literature), and wants us to believe that since we see empirical evidence of B, then everything is in a self-organized critical state. Of course this could be the case, but it could not be as well. B is not necessary and sufficient condition for A, just because A->B is true. In other words, all the thesis of this book is a speculation about state of the affairs of the world. It is most probably a very plausible thesis, but plausibility is not enough to verify a thesis as this one, because A (self-organized critically) is a mechanistic theory of the world in that state, as we know now, but what excludes any other possible mechanistic theory that would/could/might result in the same power law? The book never even mentions this question, and never even tries to answer it. This is either intellectual indolence or maybe even worse, intellectual dishonesty.

Wow! I am constantly amazed when someones careful investigation, and word crafting puts together a book like this.
The Author charts out the research path of scientists working in diverse fields as seismology and fire science,and takes the work on these seemingly random events and shows that there is a demonstrable, repeatable science to them.
More amazingly, he takes recent research work and shows that a large number of seemingly controlled things of mankind, such as city size, are driven by these same mathematical laws.
Even though the author alludes to the fact that these are events that just happen, and implies that they are beyond our control, I am hopeful that people are extending on this wonderful work to see the horizons of mankinds future in many different fields.
Brilliant work. If you are interested in the line of human history, and in the future, this is a great read.

Adding my positive response to the others here is easy. This is a thought provoking work. Which thoughts it stimulates in you will be different from mine, but just as interesting. The fact that this work is so easily readable may be not just a pleasure, but a drawback. I say this in reference to the subtitle. I fear (perhaps unjustly) that readers may think all critical changes lead to catastrophe. It's better to take a neutral position on the value of change at least in a theoretical frame.

That small comment aside, I recommend this book to all readers with a general interest in science or at least what's going on around us.

I, like some of the other reviewers, found that I could not read this cover to cover. And, I thirst for topics that look at the not-so-obvious faults and currents which underlie everything. Thus, I was eager to read about work being done with sandpile theories. Yes, there was mention of it. And, there were callbacks to it. But, whatever whole and complete picture Buchanan was trying to paint never came together. There were so many side trips and diversions that I just had to skim whole chapters in search of where I might pick up the trail again. I felt I was left in the woods, deep with discussion of the science itself rather than the topic this science was supposed to be exploring. Some of the stories were interesting. And, for me, there was a take away: the notion of maladjustment -- the build up of stress preceding any ...upheaval. In this case, the book might have a case of mal-editing.