This strange loop has a curious effect on causality. We get up in the morning and leave the house at 7.15 because we have to get to work by 9 o'clock. Scientifically, this is a very bizarre form of causality: the future is affecting the past. That doesn't normally occur in physics (except in very esoteric Quantum things, but let's not get distracted). In this case, science has an explanation.

What causes you to get up at 7.15 is not actually your future arrival at work. If in fact you fall under a bus and never make it to work, you still got up at 7.15. Instead of backwards causality, you have a mental model, in your brain, which is your best attempt to predict the day ahead. In that model, realised as buzzing electrons, you think that you ought to be at work by nine. That model, and its expectation of the future, exists now, or more accurately, a short time in the past.

It is that expectation that causes you to get up instead of lying in and having a well-deserved snooze. And the causality is entirely normal: from past to future by way of actions taking place in the present.

So that's all right then. Except that when you think of it, the causality is still very strange. A few electrons, buzzing in ways that are meaningless from the outside of the brain in which they reside, lead to a coherent action by a 70-kilogram lump of protein. Well, at that time in the morning it's not a very coherent lump of protein, but you understand what we mean. That's why we call this very creative piece of confusion a strange loop.

Those mental models are stories, simplified narratives that correspond in a rough-hewn way to aspects of the world that we consider to be important. Note that 'we': all mental models are infected with human biases. Our minds tell us stories about the world, and we base a great many of our actions on what those stories say. Here, the story is 'the person who arrived at work late and was fired from their job'. That story alone will lever us out of bed at an unearthly hour, even if we get on well with the boss and fondly imagine that the story doesn't apply to us. In other words, we make up our world according to the stories that we tell ourselves, and each other, about it.

We build minds in our children that way, too. The Western child is brought up on stories like the time Winnie the Pooh went to Rabbit's house, ate too much honey and got stuck in the entrance hole on the way out.7 The story tells us not to be too greedy; that terrible things will happen to us if we are. Even the child knows that Winnie the Pooh is fiction, but they understand what the story is about. It doesn't lead them to avoid pigging out on honey, and it doesn't make them worry about getting stuck in the doorway when they try to leave the room after having eaten too much dinner. The story isn't about literal interpretations. It's a metaphor, and the mind is a metaphor machine.

The power of narrativium in Roundworld is immense. Things happen because of it that you would never expect from the laws of nature. For example, the laws of nature pretty much forbid an Earthbound object suddenly leaping up into space and landing on the Moon. They don't say it's impossible, but they do imply that you could wait a very long time indeed before it happened.

Despite this, there is a machine on the Moon. Several. They all used to be down here. They are there because, centuries ago, people told each other romantic tales about the Moon. She was a goddess, who looked down on us. When full, she caused werewolves to change from humans into animals. Even then, humans were quite good at doublethink; the Moon was clearly a big silver disc, but, at the same time, she was a goddess.

Slowly those tales changed. Now the Moon was another world, and by harnessing the power of swans we could fly there in a chariot. Then (Jules Verne suggested) we could get there in a hollowed-out cylinder fired by a giant gun, located in Florida. Finally, in the 1960s, we found the right kind of swan (liquid oxygen and hydrogen) and the right kind of chariot (several million tons of metal) and we flew to the Moon. In a hollowed-out cylinder, launched from Florida. It wasn't exactly a gun. Well, actually it was in a basic physical sense; the rocket was the gun and it went along for the ride, firing burnt fuel in place of a bullet.

If we'd not told ourselves stories about the Moon, there would have been no point in going there at all. An interesting view, maybe ... but we 'knew' about the view only because we had told ourselves scientific stories about images sent back by space probes. Why did we go? Because we'd been telling ourselves that we would, one day, for several hundred years. Because we'd made it inevitable and introduced it into the 'future story' of a great many people. Because it satisfied our curiosity, and because the Moon was waiting. The Moon was a story waiting to be finished ('First human lands on the Moon!'), and we went there because the story demanded it.

When Mind evolved on Earth, a kind of narrativium evolved alongside it. Unlike the Discworld variety of narrativium, which on the Disc is just as real as iron or copper or praseodymium, our variety is purely mental. It is an imperative, but the imperative has not been reified into a thing.

However, we have the sort of mind that respond to imperatives, and to many other non-things.

And so it feels to us as if our universe runs on narrativium.

There is a curious resonance here, and 'resonance' is definitely the word. Physicists tell a story about how carbon forms in the universe. In certain stars there is a particular nuclear reaction, a

'resonance' between nearby energy levels, which gives nature a stepping-stone from lighter elements to carbon. Without that resonance, so the story goes, carbon could not have formed.

Now, the laws of physics as we currently understand them involve several 'fundamental constants', such as the speed of light, Planck's constant in quantum theory, and the charge on an electron. These numbers determine the quantitative implications of the physical laws, but any choice of constants sets up a potential universe. The way that a universe behaves depends on the actual numbers that are used in its laws. As it happens, carbon is an essential constituent of all known life. All of which leads up to a clever little story called the Anthropic Principle: that it's silly for us to ask why we live in a universe whose physical constants make that nuclear resonance possible - because if we didn't, there'd be no carbon, hence no us to ask about it.

The story of the carbon resonance can be found in many science books, because it creates a powerful impression of hidden order in the universe, and it seems to explain so much. But if we look a little more closely at this story, we find that it is a beautiful illustration of the seductive power of a compelling but false narrative. When a story seems to hang together, even consciously self-critical scientists can fail to ask the question that makes it fall apart.

Here's how the story goes. Carbon is created in red giant stars by a rather delicate process of nuclear synthesis, called the triple-alpha process. This involves the fusion of three helium nuclei.8 A helium nucleus contains two protons and two neutrons. If you fuse three helium nuclei together, you get six protons and six neutrons. That, as it happens, is a carbon nucleus.

All very well, but the odds on such a triple collision occurring inside a star are very small.

Collisions of two helium nuclei are much more common, though still relatively rare. It is extremely rare for a third helium nucleus to crash into two that are just colliding. It's like paint- balls and wizards. Every so often, a paintball will go splat! against a wizard. But you wouldn't bet a lot of money on a second paintball hitting him at the exact same moment. This means that the synthesis of carbon has to take place in a series of steps rather than all at once, and the obvious way is for two helium nuclei to fuse, and then for a third helium nucleus to fuse with the result.