Complex Systems and Complex Failure

9 June 2011


There is a famous verse that exists in many different forms, but is most familiar, I think, in this version of For Want of a Nail:

For Want of a Nail

For want of a nail the shoe was lost.
For want of a shoe the horse was lost.
For want of a horse the rider was lost.
For want of a rider the battle was lost.
For want of a battle the kingdom was lost.
And all for the want of a horseshoe nail.

Today we know this as the “Butterfly Effect.” We are familiar with the Butterfly effect from popular expositions of chaos theory. Choas theory has rapidly become the central theoretical reference point for studying complexity, but we don’t even have to invoke chaos theory in order to discuss cascading failures in complex systems.

The “for want of a nail” scenario above is a classic evocation of a cascading failure, and indeed the story illustrates a small failure that cascades all the way to catastrophic failure. (NB: not all cascading failures culminate in catastrophic failure.) It is in part because such scenarios of cascading failure begin with small and simple events that it becomes easy to think of failure as a simple thing. But it is not. Failure is not always simple, though it may sometimes be simple.

Complex systems fail in complex ways. Moreover, the scope of a catastrophic failure of a complex system is commensurate with the scope of the complex system. This is easy to see intuitively since a catastrophic cascading failure in a complex system must penetrate through all levels of the system and encompass both core and periphery.

This sense of the complexity of failure was brought home to me when I received a copy of Emergency Management magazine. I don’t know how I got on their mailing list, but the magazine was of interest to me. The article Deepwater Horizon Oil Spill Is an Ominous Sign for Critical Infrastructure’s Future by Austen Givens explicitly formulated the equivalence of complexity of failure with the complexity of the system that fails:

“…as complex systems continue to proliferate — converging people, processes and technologies — equally sophisticated failures of those systems are likely to emerge.”

Complex failures are all about failures that are as sophisticated as the systems themselves that experience failure. Another article, Black Swan Events Require Expanded Thinking, Planning by Eric Holdeman, made a similar point:

“The world has been experiencing a series of extreme events — political, technological and natural… Japan was hit by a triple whammy: an earthquake, followed by a tsunami and then a partial meltdown of nuclear reactors. This natural disaster led to cascading events, including the deaths of tens of thousands, the evacuation of hundreds of thousands and electrical power shortages that will continue for months.”

While considering complex systems and their complex failures, I will also point out that this holds for the increasing complexity of human societies, and even the increasing moral complexity of human life. Life today is more complex than at any earlier time in our history. The number of concerns that must be juggled by the average individual is staggering, and all of the dizzying complexity of life in the contemporary world has a moral dimension.

I can see now that when I discussed the moral complexity of life in Spots Upon the Sun that I did not there go far enough. The revolution, terror, and genocide that marks our time, and the weaponization of eliminationism that I have discussed are instances of complex moral failures only possible in a morally complex world.

There is a sense in which the complex failure of complex systems is reliant upon ecological mechanisms as well as the fact that ecological mechanisms are all but inevitable to emerge within complex systems. If failures could be compartmentalized, failures would not cascade and would not prove catastrophic. It is only when each level of the functionality of a system is related to the functionality of every other level that a system as a whole is compromised when a part of the system is compromised.

. . . . .


. . . . .

Grand Strategy Annex

. . . . .


6 Responses to “Complex Systems and Complex Failure”

  1. Luke said

    Excellent stuff, I have been visiting your blog for a few months now and I really enjoy your thoughts and your writing. 
    I had to comment on this posting because it relates to a couple of things I have been thinking about for a few years now.
    1.       The more complex the system the more redundancies there are generally in place to avoid total failure. Take something as simple as a cigar box guitar. It is easy to understand what it would take to make it unusable and equally as simple to make it work again given the correct tools and parts. It’s not hard to break and it’s not hard to fix. Compare that guitar to a living creature and the understanding required to “fix” the creature (let’s say my dog Ivan) is light years beyond what can be observed (basic physics, acoustic properties of material under tension) in the guitar. It requires the “fixer” to devote time specifically to the “fixee”. It requires more than understanding general concepts that are shared by a wide variety of basic mechanical objects.
    As I have been writing this I realize that sometimes complex systems can fail in staggeringly simple ways. Take the nail for the horseshoe example. Another way the kingdom could be lost is that a meteorite hits it. That’s pretty simple. The crushing force of the impact destroys everything. It can’t get much more simple then that really. But I think that somehow a systems vulnerability to a “simple” catastrophic failure is proportional to the “power” of the simple fault/disaster/event and the redundancies/complexity of that system. I don’t know, what do you think?
    2.       The evolution of technology will ultimately lead to the manipulation of organic matter. What is the human body if not an intensely complex system?
    We pretty much have the technology to do it now but as you have stated before there are serious moral concerns that need to be addressed before This (inevitably?) will be achieved and accepted as the natural course of things.
    So I guess my question is: Does it require the moral system in a society to “break” or just change in order for progress to be made?
    This is the first blog I have ever commented on, in fact this is the only blog I have visited more than twice in my whole internet life I think… I just really enjoy your subject matter and find the way that you write/think very accessible and interesting. Thanks for your time and keep expanding the horizons.

    • geopolicraticus said

      Dear Luke,

      Thanks for your detailed comment. I really appreciate your careful reading, as well as the fact that you should honor me with your first blog comment.

      1. While complex systems may have staggering simple triggers, I think it is important to distinguish between the trigger of a complex failure and the complex failure itself. As I tried to point out in my post, the trigger can be dead simple, but for the complex system to be declared fully and completely failed, they must fail at all levels, and this is a complex process that takes us through all the levels and details of the system that fails.

      What is distinctive about such a failure of a complex system is that it can’t be readily resurrected. A complex system needs to be reconstituted in all its complexity. The obvious example of this is ecosystem failure. After a given case of ecological nudation (whether by natural genesis or anthropogenesis), a climax ecosystem doesn’t just spring up in place of that which was lost. The replacement of the climax ecosystem is described by the process of ecological succession, which may take hundreds of years. For example, if a forest has been destroyed, it takes several hundred years for mature trees to grow back again.

      Is systems vulnerability proportional to the power of the simple trigger and the redundancies of that system? This is an interesting question. We would need to quantify both kinds and degrees of triggers, and kinds and degrees of redundancies and complexities. It is important to point out that all of these things are different. Because triggers, redundancies, and complexity are different kinds of things, the answer to the question will itself be complex, and we would need to map out our answer in a parameter space defined by the variables you mention. So the short answer is that some of these failures will be proportional, while others are not likely to be proportional. And, obviously, the disproportional ones are going to be the most interesting — like the case of “for want of a nail.”

      2. The human body has already been modified by technology. As long as technological civilization persists, this modification is likely to continue and perhaps also to accelerate. The only development that would bring this technological acceleration to the modification of the human body to a halt would be a significant social or moral change that renders such modification unthinkable.

      I do not think that the moral system of a society needs to “break” for further, perhaps even unprecedented change to take place. There only needs to be time for that moral system to experience descent with modification. Of course, a moral revolution might speed things up, and these definitely do happen in history (though, like the abolition of slavery, such moral revolutions are rare). In most cases, however, it’s simply a matter of continuous developments – which shows us that while the “slippery slope” may be a logical fallacy, it is a psychological truth.

      Far from thinking that a moral “break” needs to occur, as I implied in the above, I think it would take a moral “break” – what I above called a “significant social or moral change” – to stop further incremental and gradual moral change from taking place. The sum of incremental changes over time will be sufficient to bring about dramatic change.

      Thanks especially for implying that I’ve been trying to expand horizons with what I write here, because that is exactly what serious thinking is about. I hope that you will continue to read that I post.

      Very Respectfully Yours,


  2. […] Complex Systems and Complex Failure ( […]

  3. […] Complex Systems and Complex Failure ( […]

  4. Fuckhumans said

    One things Iv’e learned is that humans in general have a specific gift of fucking up everything they touch. Which seperates them from other animals. I hope the plan backfires cause I know it will.

    • geopolicraticus said

      Dear Fuckhumans:

      Thanks for your comment.

      While is happens to be the case that our only detailed knowledge of a conscious and civilized species is of human beings, there is no a priori reason to suppose that other conscious and civilized species (if there are any) are any different.

      Just as it is anthropocentric to praise human beings at the expense of other species, it is equally anthropocentric to condemn human beings above and beyond other species. Both are equally instances of human exceptionalism.

      I attempted to discuss similar issues of anthropic bias in The Future Science of Civilizations, which I just posted today.

      Best wishes,


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

%d bloggers like this: