Clegg Dice World

Science and Life in a Random Universe
1. Auflage 2013
ISBN: 978-1-84831-564-8
Verlag: Icon Books
Format: EPUB
Kopierschutz: 6 - ePub Watermark

E-Book, Englisch, 288 Seiten

LONGLISTED FOR THE 2014 WINTON ROYAL SOCIETY PRIZE FOR SCIENCE BOOKS As troubling as we pattern-seeking humans may find it, modern science has repeatedly shown us that randomness is the underlying heartbeat of nature. In Dice World, acclaimed science writer Brian Clegg takes readers on an incredible trip around our random universe, uncovering the truths and lies behind probability and statistics, explaining how chaotic intervention is behind every great success in business, and demonstrating the possibilities quantum mechanics has given us for creating unbreakable ciphers and undergoing teleportation. He explores how the 'clockwork universe' imagined by Newton, in which everything could be predicted given enough data, was disproved bit by bit, to be supplanted by chaos theory and quantum physics. Clegg reveals a world in which not only is accurate forecasting often impossible but probability is the only way for us to understand the fundamental nature of things. Forget the clockwork universe. Welcome to Dice World, a unique portrait of a startlingly complex cosmos, from the bizarre microscopic world of the quantum to the unfathomable mechanics of planetary movements, where very little is as it seems...

Brian Clegg is a popular science writer whose Dice World and A Brief History of Infinity were both longlisted for the Royal Society Prize for Science Books. He has written for publications including Nature, The Times and BBC Focus.

Autoren/Hrsg.

Clegg, Brian

Leseproben

CHAPTER 1

Improbable world

The world is a complicated and messy place, especially when you consider the complexities we add to it with our carefully constructed environment. Take a really simple act that most of us perform every day without giving it a thought – switching on an electric light. This is clearly not something we are genetically programmed to deal with from birth. Human beings are pretty well identical to the creatures that evolved to survive on the savannah after their ancestors stopped living in trees over 100,000 years ago. Once you get beyond basic bodily functions and activities, the vast majority of our time in the modern world is spent doing things that the human body did not evolve to do. All the rest of our activities and experiences are relatively newly learned. We live unnatural lives.

It’s certainly true that there weren’t many light switches 100,000 years ago. So we all have to learn how to turn the light on – and for most of us (until we venture across the Atlantic and find that they incomprehensibly mount their switches the wrong way up on the wall) it is a natural-seeming, easy act. We flick the switch and the light comes on. No real thought involved. It’s trivial.

But imagine that you had to program a robot from scratch to switch on the light in your living room. You would need to specify exactly where the switch was located. This would involve providing detail of where each wall was, which wall the switch was on, at what height it was located and at what distance it was from the wall’s edge. Alternatively you would need to show your robot exactly what the switch looked like from every possible angle, so the robot could search for it visually. You would also need to specify where and in which direction to apply pressure to the switch, how much pressure to use (it would be embarrassing if the robot snapped the thing off) and when to stop pressing.

What seemed trivial turns out to be anything but a simple task. But more to the point, if you now moved that robot into the hallway and asked it to carry out the same job there, you would have to start all over again. There might be a totally different design of switch with dissimilar physical characteristics. It’s highly unlikely this new switch would be in the same place on the wall in the hall as the switch is in the lounge. Set the robot in action without reprogramming it and you would probably end up with a hole punched in the plaster.

As human beings, we simply can’t afford the time and effort to do the equivalent of re-programming our brains each time we encounter a different light switch. And so we deal with patterns. We don’t learn exactly what each light switch that we encounter is like. Instead we have a broad pattern in mind which specifies ‘This is how you switch on a light using a wall switch’. It enables us to recognise the switch in a broad range of styles and then just to do it – press the switch, get the light. Until some clever designer comes up with a switch that works when you speak to it or touch the lamp itself – and then you have to start the discovery process all over again.

Finding patterns

Of course, we didn’t evolve an ability to recognise patterns to cope with light switches. But exactly the same flexibility of pattern-matching enables us to spot a predator – or a familiar friendly face – even if we have never been in a particular exact circumstance before, and so to take appropriate action. We work with patterns that give us the ability to reduce the almost infinite set of possible deductions from our sensory inputs to a manageable set we can work with, using the mental shorthand that enables us to just ‘flick the switch’, ‘run from the tiger’ or ‘see and say “Hi” to Nic.’

We are so good at this pattern-matching that we can achieve it even when we have a surprisingly low amount of information on which to make a judgement – in this we are usually a lot better at filling in the gaps than computers are. This is why the ‘CAPTCHA’ system, used by websites to ensure that people are taking part rather than software programs, makes use of distorted text with characters that are twisted or run into each other. This is a visual input that a human can usually interpret, but a piece of software struggles with turning into useful data.

Take the three partial pieces of text below:

No one would be challenged to see that the top word reads ‘BANK’, even though a sizable percentage of the text is missing. We find it trivial to fill in the gaps. In the second example, a whole 50 per cent of the text has been chopped off, but there is still enough there to be sure what the word is. It is only the positioning of the final chop, introducing ambiguity with two possible interpretations of ‘BANK’ or ‘RANK’, that finally beats our superb ability to take a partial pattern and reconstruct the whole.

Much of the time this human ability to detect patterns is a real plus. It means that we can work with limited data – and in the real world the set of data that we have available is almost always incomplete. But the danger we face is that the pattern-constructing and -matching systems of our brains are so powerful that we imagine patterns when there is nothing there.

This is a good survival principle. It’s better to be sufficiently sensitive that you occasionally see a predator where there isn’t one, rather than risk missing a killer that is lurking in the bushes. So we create bogeymen out of shadows and misinterpret all kinds of evidence. We see faces in the shadows, in the clouds, or even in the burn marks on a slice of toast. Pure randomness with no pattern is something we find difficult to relate to – our brains expect to see patterns and they do.

The patterns of science

This pattern-matching isn’t just about our low-level, immediate, day-to-day interaction with the environment around us (important though that is). It is also the basis of science. It’s strange, in a way, that many of us struggle with science because all the scientific method does is to take the basic mechanism we all use to understand the world without even thinking about it, and formalise that mechanism into a process.

In science we are looking for patterns and rules to explain what the universe and its components do and how they do it. It’s as simple as that. The mechanisms modern scientists use may get heavy-duty and scarily mathematical, but the basic principle is still one of looking for patterns. What scientists do is arguably just a simple and rather beautiful formalisation of our natural approach to exploring the unknown.

We start off in a state of ignorance. We gather enough data to be able to formulate a hypothesis about what’s going on. Then we test that hypothesis – a kind of predictive pattern – against subsequent observations; if it continues to work, we can build on it. If it fails us, we have to start all over again. That’s the scientific method. It should be how we naturally interact with the world too, but all too often, once we get a hypothesis, we get fond of it. We can’t let it go despite plenty of evidence to the contrary. And that’s when science slips into superstition.

To have any hope of making a scientific approach work, we have to expect some degree of consistency of behaviour from the universe. Take something we think of as a constant, a fixed point of certainty – the speed of light. If this varied from day to day or second to second with no logical reason for that variation, and no way of ever anticipating what the speed will be today, then we could never make use of the speed of light, as astronomers do all the time, to help us understand the universe. Given how much of our exploration of the universe is dependent on light and its speed, this would be totally devastating for cosmology. In fact, without a degree of consistency, the whole concept of science would collapse. We would live in a universe that might as well be magical. It is impossible to draw any hypotheses if every time you do an experiment you get totally different results.

This doesn’t mean that there won’t be circumstances when the speed of light does vary. We know that it is different in a vacuum from when it is passing through a substance – it is slower in air, still slower in water and so on. There are even substances called Bose Einstein condensates that can effectively bring light to a standstill. This is because photons of light don’t pass through matter unaffected but interact with electrons, being absorbed and re-emitted, slowing down their progress. But this isn’t a problem for science, because these variations are predictable. I know that the speed of light is different when it’s going through space than from when it’s going through glass. But for the same medium under the same conditions, I expect to get the same result.

I chose the speed of light intentionally because there is even a theory (a perfectly reasonable theory, though not one with a lot of support at the moment) that the speed of light has not stayed the same over time. According to this theory, over the billions of years of existence of the universe, the speed of light has changed very slightly. If this is true, while it would modify some of our conclusions about exactly what was happening long ago in galaxies far, far away (as they say), it too wouldn’t be a huge problem for science, because it is something we could predict and consider the influence of over time.

The randomness confusion

There is, however, one aspect of dealing...