So here it is, Stephen Hawking's latest work, written with physicist Leonard Mlodinow. A new Hawking book is always somewhat of a media circus - even though, in my very humble opinion, there are often far more interesting and better-written science books published which don't get any attention at all. Stephen Hawking has become such a science celebrity you would think he is the only physicist who ever writes books. Actually, you'd be forgiven for thinking he is the only physicist.
The Grand Design is an attempt to explain the current state of thinking in cosmology and physics. It doesn't, predictably, offer "new answers to the ultimate questions of life", but we can hardly blame Hawking for his publisher's hype. If you're looking for deep answers to existential questions, this is not the book for you. Instead, we get a fairly readable, and surprisingly short, explanation of Hawking's view of the Universe: the idea of model-dependent realism.
Model-dependent realism is the idea that it's pointless to ask whether a particular model of the Universe is real: you can only ask if it agrees with observation. To use Hawking's example, imagine a goldfish seeing a completely distorted view of the world through the convex glass of his fish bowl. If he were a particularly clever goldfish, he might construct experiments to determine if his view were consistent under all conditions. He might then come up with a model of the outside world which, while being perfectly consistent with observations, and being able to predict the result of further observations, we would view as incorrect. In this case, the goldfish's view of the world is different to ours, but both can be seen as correct depending on whether you are in the fish bowl or outside it, because in both locations you can do experiments which will verify your own particular model.
Hawking says that there is probably no one view of the Universe, no one model, that is correct at all times, and that we might need different models depending on what we are observing. This might come as a shock to reductionists, and those searching for the equation-on-a-T-shirt which will explain everything, but it gets worse. Hawking says that an implication of Feynman's sum-over-histories is that the Universe does not have one single past history, but a possibly infinite number of them. So not only can we not rely on our perception of the history of the Universe to be correct, we can not rely on any one model of the Universe to be definitive and we cannot make predictions about its future either. In Hawking's universe the idea of a universal truth does not exist: we may never get answers to The Big Questions because there are no universal models which apply everywhere and at the same time. Everything depends on what you are observing and how you are measuring it, which may be explained through using a different model of the Universe from what you will be using tomorrow to explain other observations. If this all sounds depressing, it is in a way, but it is also deeply mysterious, and that's what makes it interesting.
