Kepler and the other planet-finding missions have begun to bear fruit. We now know that most stars have planets, and that a surprising percentage will have Earth-sized worlds in their habitable zone--the region where things are not too hot and not too cold, where life can develop. Astronomers are justly fascinated by this region and what they can find there. We have the opportunity, in our lifetimes, to learn whether life exists outside our own solar system, and maybe even find out how common it is.

We have another opportunity, too--one less talked-about by astronomers but a common conversation among science fiction writers. For the first time in  history, we may be able to identify worlds we could move to and live on.

As we think about this second possibility, it's important to bear in mind that habitability and colonizability are not the same thing. Nobody seems to be doing this; I can't find any term but habitability used to describe the exoplanets we're finding. Whether a planet is habitable according to the current definition of the term has nothing to do with whether humans could settle there. So, the term applies to places that are vitally important for study; but it doesn't necessarily apply to places we might want to go.

To see the difference between habitability and colonizability, we can look at two very different planets: Gliese 581g and Alpha Centauri Bb. Neither of these is confirmed to exist, but we have enough data to be able to say a little about what they're like if they do. Gliese 581g is a super-earth orbiting in the middle of its star's habitable zone. This means liquid water could well form on its surface, which makes it a habitable world according to the current definition.

Centauri Bb, on the other hand, orbits very close to its star, and its surface temperature is likely high enough to render one half of it (it's tidally locked to its sun, like our moon is to Earth) a magma sea. Alpha Centauri Bb is most definitely not habitable.

So Gliese 581g is habitable and Centauri Bb is not; but does this mean that 581g is more colonizable than Bb? Actually, no.

Because 581g is a super-earth, the gravity on its surface is going to be greater than Earth's. Estimates vary, but the upper end of the range puts it at 1.7g. If you weigh 150 lbs on Earth, you'd weigh 255 lbs on 581g. This is with your current musculature; convert all your body fat to muscle and you might just be able to get around without having to use leg braces or a wheelchair. However, your cardiovascular system is going to be under a permanent strain on this world--and there's no way to engineer your habitat to comfortably compensate.

On the other hand, Centauri Bb is about the same size as Earth. Its surface gravity is likely to be around the same. Since it's tidally locked, half of its surface is indeed a lava hell--but the other hemisphere will be cooler, and potentially much cooler. I wouldn't bet there's any breathable atmosphere or open water there, but as a place to build sealed domes to live in, it's not off the table.

Also consider that it's easier to get stuff onto and off of the surface of Bb than the surface of a high-gravity super-earth. Add to that the very thick atmosphere that 581g is likely to have, and human subsistence on 581g--even if it's a paradise for local life--is looking more and more awkward.

Doubtless 581g is a better candidate for life; but to me, Centauri Bb looks more colonizable.

A definition of colonizability

We've got a fairly good definition of what makes a planet habitable: stable temperatures suitable for the formation of liquid water. Is it possible to develop an equally satisfying (or more satisfying) definition of colonizability for a planet?

Yes--and here it is. Firstly, a colonizable world has to have an accessible surface. A super-earth with an incredibly thick atmosphere and a surface gravity of 3 or 4 gees just isn't colonizable, however much life there may be on it.

Secondly, and more subtly, the right elements have to be accessible on the planet for it to be colonizable. This seems a bit puzzling at first, but what if Centauri Bb is the only planet in the Centauri system, and it has only trace elements of Nitrogen in its composition? It's not going to matter how abundant everything else is. A planet like this--a star system like this--cannot support a colony of earthly life forms. Nitrogen is a critical component of biological life, at least our flavour of it.

In an article entitled "The Age of Substitutibility", published in Science in 1978, H.E. Goeller and A.M. Weinberg proposed an artificial mineral they called Demandite. It comes in two forms. A molecule of industrial demandite would contain all the elements necessary for industrial manufacturing and construction, in the proportions that you'd get if you took, say, an average city and ground it up into a fine pulp. There're about 20 elements in industrial demandite including carbon, iron, sodium, chlorine etc. Biological demandite, on the other hand, is made up almost entirely of just six elements: hydrogen, oxygen, carbon, nitrogen, phosphorus and sulfur. (If you ground up an entire ecosystem and looked at the proportions of these elements making it up, you could in fact find an existing molecule that has exactly the same proportions. It's called cellulose.)

Thirdly, there must be a manageable flow of energy at the surface. The place can be hot or cold, but it has to be possible for us to move heat around. You can't really do that at the surface of Venus, for instance; it's 800 degrees everywhere on the ground so your air conditioning spends an insane amount of energy just overcoming this thermal inertia. Access to a gradient of temperature or energy is what makes physical work possible. 

Obviously things like surface pressure, stellar intensity, distance from Earth etc. play big parts, but these are the main three factors that I can see. It should be instantly obvious that they have almost nothing to do with how far the planet is from its primary. There is no 'colonizable zone' similar to a 'habitable zone' around any given star. The judgment has to be made on a world by world basis.

Note that by this definition, Mars is marginally colonizable. Why? Not because of  its temperature or low air pressure, but because it's very low in Nitrogen, at least at the surface. The combination of Mars and Ceres may make a colonizable unit, if Ceres has a good supply of Nitrogen in its makeup--and this idea of combo environments being colonizable complicates the picture. We're unlikely to be able to detect an object the size of Ceres around Alpha Centauri, so long-distance elimination of a system as a candidate for colonizability is going to be difficult. Conversely, if we can detect the presence of all the elements necessary for life and industry on a roughly Earth-sized planet, regardless of whether it's in its star's habitable zone, we may have a candidate for colonizability.

The colonizability of an accessible planet with a good temperature gradient can be rated according to how well its composition matches the compositions of industrial and biological demandite. We can get very precise with this scale, and we probably should. It, and not habitability, is the true measure of which worlds we might wish to visit.

To sum up, I'm proposing that we add a second measure to the existing scale of habitability when studying exoplanets. The habitability of a planet actually says nothing about how attractive it might be for us to visit. Colonizability is the missing metric for judging the value of planets around other stars.

Yesterday it was announced that an earth-sized planet has been discovered circling the nearest star, Alpha-Centauri (around the smaller of its two main stars, actually, Alpha-Centauri B). The planet, Bb as it's currently called, has a six-day year and a surface temperature of 1500C. Not very hospitable, perhaps--but I'm about to argue that it's just fine. If we can get to this star system, we can settle it.

Let's look at two scenarios, a worst-case and a best-case, and see what's possible with each.

The Worst Case

This is a boiling hot planet. Actually, far hotter than boiling. At 1500 degrees, it's hot enough to melt rock. In a worst-case scenario, Bb has the kind of rotational resonance that Mercury does: it is not fixed with one face pointed forever at its star, like our Moon is to the Earth, but rotates so that the whole planet is regularly bathed in the blowtorch heat of the star. If there is an atmosphere, it's mostly composed of evaporated rock.

In this case, much or all of Bb's surface is a lava sea. Oh, and since this is a worst-case scenario, let's say that there are no other planets in the system, not even any asteroids. Bb is it.

If your idea of habitability is finding a more or less exact copy of the Earth and settling down on it to farm, then things are looking kinda bleak. But, if we have the technology to get to Bb, then we have the technology to live and thrive there.

Not on the surface, of course. Not even in a nearby orbit. But even if Bb is uninhabitable, it is still a great source of building material. If we have the technology to get to it, we'll have the technology to mine it, if only by dangling a skyhook down from the L2 point (or from a heliostat) to dredge the magma ocean. Haul the magma up, render it in the terrible light of the star, and ship the refined goods to a higher orbit where the temperature's a bit better. There, we can build habitats--either O'Neill colonies or, if we can harvest enough material, the coronals I describe in my novel Lady of Mazes.

With unlimited energy and (nearly) unlimited building materials, we can construct a thriving civilization around Alpha Centauri B, even if all we have to work with is this one piece of melted rock. (In terms of details, it would be a bootstrapping operation, with an initial small seed of robot miners constructing more or bigger skyhooks, more miners, etc. until exponential growth sets in, by which time it's safe for the human colonists to show up.)

The Best Case

 Even for the best case scenario, I'm going to assume that Bb is the only planet in the system. It's more likely than not that Bb actually will be tidally locked to its star--i.e., it has one face permanently aimed at its sun, and the other permanently in darkness. The point that's under a permanent noon (the 'solar pole') will indeed be a lava hell. What's interesting, though, is that some simulations show that the temperature in the twilight zone around the 'equator' and further into the night side could be quite cool. Cold, even, if you go far enough. If there's an atmosphere, there might be water and a zone of permanent rain around the mid-latitudes of the dark side, in a kind of hemisphere-wide hurricane with its eye at the anti-solar pole. And there, we might settle.

I doubt there'd be any oxygen to speak of, but we can generate that ourselves. What I find interesting, though, is that this 'dark side' is not really dark at all. Because Alpha Centauri is a binary star system, Centauri A will be visible in the 'night' sky of Bb during half its year. ...Which is only three days long. So A will cross the sky in about 75 hours, and then there'll be true night for 75 hours. This has been the pattern on Bb now for more than four billion years; it's pretty stable.

Centauri A appears very dim from Bb compared to our sun, but it's still too bright to look at and has a visible disk. It's dimmer than daylight, but much, much stronger than Earthly moonlight. Granted the luminosity range at which photosynthesis happens on Earth, I'd think plant life would do quite well on Bb's 'dark' side.

If the rain's not too bad, much of the 'dark' hemisphere might be settled. Remember that Earth is mostly covered with water; if there's no significant oceans on Bb, but enough water for rivers and lakes, then the habitable land area of Bb might be greater than Earth's. Gravity is the same as Earth's, and in fact the only major difference will be atmospheric composition/density, and the length of the day. And who knows? Maybe we can game those too, by geoengineering the atmosphere, and using a combination of distant orbital sunshades and orbiting mirrors to generate a 24-hour diurnal cycle. Ultimately, Bb could be very earth-like indeed.

The Happy Medium

I expect the reality of Bb's habitability lies somewhere in between the two extremes I've just described. In all likelihood, Bb is not alone; at the very least, there should be asteroids or planetoids of Ceres-size or larger. Bb itself might have a safe spot where industrial operations can be set up, even if it's not a place where you could live. It can export vast quantities of raw materials to colonists elsewhere in the Centauris.

All of which means one thing: Alpha Centauri is now a viable destination. If we can get there, we can live there. And knowing this makes real possibilities that, until yesterday, we could only dream about.

I started reading Ian Bogost's latest book last night. Alien Phenomenology, or What it's Like to be a Thing seems an unlikely excursion for a theorist whose major work so far was a literary theory for video game criticism. (I used the ideas in Bogost's book Unit Operations as a major theoretical framework for the scenario-fiction writing technique I outlined in my Master's thesis.)

 It's not often that I have the experience of hopping up and down, gnashing my teeth and shouting "well of course!" but I've been having it since starting Alien Phenomenology. But I don't mean that in a bad way; I had the same experience when I dove into Jane Bennett's Vibrant Matter, and more recently exploring the work of philosopher Graham Harman. It's the frustration of a long-delayed recognition of kindred minds. I talked a little about that recently, and here's that same new feeling again.

So, as I'm reading Bogost and I come across statements like 

That things are is not a matter of debate. What it means that something in particular is for another thing that is: this is the question that interests me. The significance of one thing to another differs according to the perspective of both.

...I am forcibly reminded of, how, nearly fifteen years ago now, I imagined Jordan Mason sitting on the shore of a lake, and listening as the smart-dust nanotech that pervaded the entire surface of the planet Ventus tried to figure out what it was:

He could hear the song of the lake. It was deep and powerful, belying the tranquility of the surface. Thin grass grew here, but the soil beneath his feet was shallow, quickly giving way to sand. Below that... rock? He couldn’t quite make it out, though it felt like there was something else down there, a unique presence deep below the earth.

There was no indication that anything supernatural dwelt here.

He sat down, mind empty for the first time in days, and watched the water for a while. Gradually, without really trying, he began hearing the voices of the waves.

They trilled like little birds as they approached the shore. Each had its own name, but otherwise they were impossible to tell apart. They rolled humming towards Jordan, then fell silent without fanfare as they licked the sand. It was like solid music converging on him where he sat. He had never heard anything so beautiful or delicately fragile.

He didn’t even notice the failing light or the cold as he sat transfixed. His mind could not remain focussed forever, though, and after a while he made up a little game, trying to follow individual waves with both his eyes and his inner sense.

He tried to follow the eddies of a particular wave as it broke around a nearby rock, and in doing so discovered something new. It seemed like such an innocent detail at first: as the wave split, so did its voice. From one, it became many, then each tinier individuality vanished in turbulence. As they did, they cried out, not it seemed in fright, but in tones almost of... delight. Urgent delight--as if at the last second they had discovered something important they needed to tell the world.

This quote from my year-2000 novel Ventus presents a vision of the self-definition of the world becoming visible for the first time to a human being. The designers of the Ventus terraforming system imagined a technology that would dissolve into everything in the world and actively investigate it. The nanotech in and on a tree would figure out that it was a tree; a rock would know it was a rock, a hill that it was a hill. And each of these objects would be able to communicate to the human settlers of the planet what it could do for them. "I am flint, you can build a fire with me." "I am mint, you can eat me." The only problem was, this magnificent system for identifying things had to be able to invent its own categories in order to do its job; and it did that too well. When the human settlers arrived, it quickly decided what they were--but on its own terms, and using its own ontology and semantics. As far as the humans were concerned, the nanotech didn't recognize them. But something far more interesting had in fact happened: it saw them, not as they wanted to be seen--not through their filter--but as it had come to see things. 

And so the nanotech (which later generations of humans called the Winds) destroyed all the settlers' competing technologies, knocked them back to the stone age, and went about integrating them efficiently into the artificial ecosystems of Ventus.

Ventus was far more than a cautionary tale about technology run amok--in fact it wasn't really that at all. I wanted to talk about how objects see other objects; but back then, I had nobody to talk to about it. Bogost's new book is another indication that the hourglass has turned, and that these ideas are finally current.

I've since moved on to next steps--but I would recommend Alien Phenomenology because Bogost also senses the need to go from discussing OOO in the abstract, to working out what it means in practice. Alien Phenomenology is the first book I've seen that explicitly challenges its readership to employ and deploy the ideas of speculative realism. This will be no mean feat, and I've already spent five years planning how to do that for my as-yet unwritten third novel in the Ventus/Lady of Mazes series, a book I've tentatively titled The Rewilding. 

Because now that an entirely new world--new universe, in fact--lies open to us, it's time to stop pointing at it, and time to start exploring it.

And building in it.

Thirteen years ago, I began my first published novel with the following words:

...Frankenstein's monster speaks: the computer. But where are its words coming from? Is the wisdom on those cold lips our own, merely repeated at our request? Or is something else speaking? --A voice we have always dreamed of hearing?

So begins Ventus, which of course is about nanotech and terraforming; but is also about something else, for which I didn't have a name at that point. I made one up: I called the concept thalience. Thalience is what you get when you find (or deliberately create) entities that are clearly objects, but which behave in ways that are supposed to only be possible for subjects. A thalient entity is neither object nor subject, or perhaps it's both. The book explores this tension (though not without a few swordfights, battles, betrayals, and romances).

I mention this because, now that Bruce Sterling has talked about Graham Harman's 'object-oriented philosophy' in Wired, this meme appears ripe for becoming a new intellectual fashion. Perhaps it's petty, but I'd like to put a stake in the sand here.

Two terms, speculative realism and object-oriented ontology, have very recently given a name to the thing that I've been thinking and writing about for nearly two decades now (it took seven years to write Ventus). It's been unbelievably gratifying for me to discover these kindred spirits--people like Jane Bennett, Ian Bogost, Andy Clark, Timothy Morton, Graham Harman and Bruno Latour. Latour's been at this for decades, and I confess to only recently discovering him--but the others in this cadre seem to have undertaken their intellectual investigations at about the same time as myself. They are all scientists, theorists or philosophers, of course; as far as I know, I'm the first person to have explicitly built science fiction novels around these new areas of inquiry.

After Ventus, the novel where I jumped in with both feet was Lady of Mazes. In it, the anti-Ariadne, Livia kodaly, wages a one-woman war against what Quentin Meillassoux has now helpfully labeled as correlationism. Correlationism is the belief that the only reality is the object-subject pair--that all I can ever say about anything is that is like such-and-such for me. I can never say what it is in itself; Kant made that impossible. In Lady of Mazes, Livia begins with this belief; as she puts it, "reality is always mediated."

That may be true, but Livia is unsatisfied with the conclusion everybody else has drawn--a conclusion that has direct political and emotional consequences for her and her people. The artificially intelligent systems that create and sustain the consensual realities in which Livia's people live, called manifolds, do not interface with them through speech, or any normal communications medium--they do so by observing our values. At one point in the story, a manifold has become empty because all its human citizens have died. Yet the manifold still exists, because its creators built it around the value of music, and they have left a single drum beating, its tapping driven by water dripping from a rain-catchment barrel. Livia's peers want to retire the manifold and take over its resources--but late one night, she sneaks into the drummers' reality and replaces the ailing drum with a fresh one.

“Each drumbeat sounded clear and distinct.  Each one rolled out into the night, reaching nobody’s ears, but real nonetheless.  It was a tremble of air, nothing more, yet in that tremble the drummers lived.  In that tremble of air was something not of Westerhaven, not preserved by your Government or to be found in the narratives.  Call it the Song of Ometeotl, if you wish.  It remained in my ears as I stole back through the forest, and returned in secret to my home.”

...“At the time I didn’t know why I did it.  It was one of those actions that you can’t reconcile with the person you think you are.  But now I understand.  I was honoring the existence and dignity of a reality independent of my own." 

This is one of the purposes of object-oriented philosphy (or speculative realism if you prefer): to honour the existence and dignity of a reality independent of our own. For me, to have written the above words in 2003 was to expose a nerve that I thought at the time was entirely private and personal--it was to confess to a unique mania that I felt no one else would understand or sympathize with. While the critical reception to Lady of Mazes was very kind, I did get that sense: the book was good, the topic... odd. What is most odd is that now, in 2012, the issues I brought up in the book seem utterly current, even obvious. (I suppose that's one reason why The Atlantic just reviewed Lady of Mazes.)

Livia never abandons the idea that reality is always mediated, but she does abandon the idea that there is nothing real outside of the human-world correlation. She imagines the relationship I called thalience, and it sets her free. She uses her new knowledge to in turn free her people from a correlationist tyranny personified in the novel by the culture known as the Archipelago, and an idealist AI named 3340.

Messianism aside, this pair of ideas--rejection of correlationism and commitment to a necessary mediation between the things of this world--locates me rather precisely in the current landscape of speculative realist thinkers. To be exact, it puts me in substantial agreement with Graham Harman, whose new book The Quadruple Object is compatible, I guess you'd say, with Livia Kodaly's stance. (Of course Harman is doing philosophy, and I am not: my explorations are artistic, though they allow me to create some odd quasi-philosophical entities, such as artificial intelligences designed to make the cracks in correlationism obvious.)

With people like Bennett and Bogost and Morton and Harman writing about this stuff, I'm suddenly overwhelmed with ideas and new perspectives. You can see it in my recent work, in particular two recent stories, "To Hie from Far Cilenia" and "Deodand." There'll be more.

In 2003 I thought I was alone in wanting to wage what Blake called 'mental fight' for what I'd come to call the dignity of the real. Somehow (surely without my influence) an army is coalescing around the issue. 

It's great to have discovered kindred spirits.

How many planets are there within 20 lightyears of our sun? Even five years ago we couldn't have answered this question. Today, without actually having spotted any, we can give a fairly confident estimate of how many there should be, and what they should be like. Interested in finding out? Then read on.

The Studies

There's been a lot of commentary in the news in the past year or so about the Kepler mission's cataloguing of distant planets. Kepler has allowed the number of known exoplanets to balloon up past 700 at latest count. Of course, since Kepler is watching a vastly distant patch of sky, it can't tell us how many planets there are in our local neighbourhood. 

A lot has been written about the significance of Kepler's technique, which involves watching for the mini-eclipses that happen when a planet crosses the face of its star. Very little has been written about a parallel hunt that uses microlensing to accomplish a similar end. Microlensing looks for the distortions in the image of a star made by a planet's gravity. These surveys have been going on for ten years now and the results are staggering. 

For instance, did you know that by some estimates there are up to 100,000 nomad planets--planets without a home--for every star in the galaxy? In my 2002 novel Permanence I boldly proposed that there might be one or two brown dwarfs for every star, and that seems to be true; but even in my wildest dreams I couldn't have imagined there might be tens of thousands of planets Pluto-sized or larger drifting between Earth and Alpha Centauri! I still can't really believe it.

These nomads are interesting, because sufficiently large ones (many will be of super-earth size, 2 or more earth-masses) can sustain a trickle of heat from their interiors for billions of years. Though their surfaces may be frozen, they can easily support sub-surface oceans like the one thought to exist in Jupiter's moon Europa. In other words, they can support life. There should be some thousands of these worlds for every star in the galaxy.

This is just the beginning of what the microlensing survey data is showing us. There's enough data now to begin to estimate how many orbiting planets your average star has, and what kind of planets they are. And the combination of microlensing survey data and Kepler data lets us be really precise.

The Numbers

 Kepler's preliminary data seems to indicate that one third of main sequence F, G, and K stars (sunlike stars) have at least one earth-sized planet within the star's habitable zone. There are nineteen such stars within 20 lightyears of us, so this indicates, conservatively, that there are six earth-sized planets in the habitable zone of sunlike stars within 20 light years of Earth. 

These numbers don't include habitable moons of gas giants that might orbit within the zone. So the actual number could be higher by one or two.

The microlensing survey lets us be precise for the whole population of stars. Here, survey says that the average number of planets per star in our galaxy is 1.6. This leads to the number of bound planets in the galaxy being close to 200 billion, and the number of total planets (including nomads) being ten quadrillion. (There are thus trillions of nomadic super-earths, many of which will have sub-ice oceans capable of developing life.)

The microlensing survey data is so far limited to planets in the super-earth to Jupiter size range, and between .5 and 10 AU distance of their stars. Within those limits, it suggests that 17% of stars have a Jupiter-like planet; 52% have a Neptune-sized planet and 62% have a super-earth. Since the smaller the planet, the more likely it is, we can continue this trend-line to say that in all likelihood, each star will have at least a 62% chance of having an Earth-sized planet. This puts the number of Earth-sized planets in the galaxy at 60 billion or so. The absolute number within 20 light years is at least 42. There's 51 stars outside the main sequence (giants or dwarfs) within 20 light years; another study suggests that the absolute probability for all stars of having a planet within the habitable zone is about 12% (which looks highly conservative). That would add six to our local total, meaning that within 20 light years, there should be at least 12 habitable earth-sized planets. This doesn't count marginal planets, exomoons and Europan worlds. Or, of course, nomads.

To zoom in on a couple of famous local stars, we can say that it's highly unlikely that Alpha Centauri has no planets, given that it is a triple system all of whose stars could support planets. We know Alpha Centauri has no gas giants, but that's consistent with the numbers; but the odds that either Centauri A or B have at least one earth-sized planet within the habitable zone are very high. The Centauris are close to our sun in age, so their planets may still be able to support life.

Tau Ceti, a very sunlike star only 12 light years away, probably has a couple of planets. It's an older star, however, and any earth-sized planets are probably getting arthritic: their plate tectonics will be shutting down somewhere around now. They'll be more like the Barsoom of Edgar Rice Burroughs' Mars novels: ancient and dying.

There's a lot more being discovered and theorized; for instance, one new study suggests that having a Jupiter-like massive planet in your solar system doesn't protect your planet from massive impacts, but on the contrary is a actually bad for you. Another suggests that at least 12% of earth-sized planets have a moon large enough to stabilize their axial tilt (a supposed necessity for planetary habitability) and another suggests that axial tilt won't affect climate all that much anyway. The prospects for life look good around the nearest stars.

The galaxy is literally overflowing with planets, far more than can be crammed into the orbits of its stars. Many of these planets could support life. The question now is, do they?

And if so, where are our nearest neighbours?

Last summer I wrote a guest article on Charlie Stross's blog about wicked problems.  Some of the characteristics of wicked problems are:There is no definitive formulation of a wicked problem (defining wicked problems is itself a wicked problem).

1. There is no definitive formulation of a wicked problem (defining wicked problems is itself a wicked problem).
2. Wicked problems have no stopping rule.
3. Solutions to wicked problems are not true-or-false, but better or worse.
4. There is no immediate and no ultimate test of a solution to a wicked problem.
5. Every solution to a wicked problem is a "one-shot operation"; because there is no opportunity to learn by trial and error, every attempt counts significantly.
6. Wicked problems do not have an enumerable (or an exhaustively describable) set of potential solutions, nor is there a well-described set of permissible operations that may be incorporated into the plan.
7. Every wicked problem is essentially unique.
8. Every wicked problem can be considered to be a symptom of another problem.
9. The existence of a discrepancy representing a wicked problem can be explained in numerous ways. The choice of explanation determines the nature of the problem's resolution.
10. The social planner who tackles a wicked problem has no right to be wrong (planners are liable for the consequences of the actions they generate).

Now Chris Smith has introduced me to a great article on How Complex Systems Fail by Richard I. Cook, MD. It's a very similar summary, but wickedly (if I can use that word) clever and, for anybody who's actually dealt with complex systems, so utterly true. Some of Cook's observations on the failure of complex systems include:

1.  Complex systems are intrinsically hazardous systems.

3.  Catastrophe requires multiple failures - single point failures are not enough.

4.  Complex systems contain changing mixtures of failures latent within them.

and one of my personally favourites:

5. Complex systems run in degraded mode.

For any of us who watched the Fukushima fiasco last summer, some of these will have an uncanny familiarity:

7.  Post-accident attribution of accidents to a 'root cause' is fundamentally wrong.

8.  Hindsight biases post-accident assessments of human performance.

15.  Views of 'cause' limit the effectiveness of defenses against future events.

16.  Safety is a characteristic of systems and not of their components.

...and finally, 

18. Failure free operations require experience with failure.

It's a sobering list and every single item on it bears a great deal of thinking. The article as a whole is brief, but each of the items is explained in enough detail to make the ideas understandable and to provoke some thought.  Everything in here is applicable in many different contexts, from Fukushima and Chernobyl to the Eurozone meltdown, to current electoral issues and the unintended consequences of urban planning decisions anywhere in the world.  Check out the article.

...And stop thinking in terms of root causes, damnit!