Big Bang - The Most Important Scientific Discovery of All Time and Why You Need to Know About It by Simon Singh, Fourth Estate. Hardback ISBN 0007152515, £20.
Ex-particle physicist Simon Singh is now in the literary big time. After gaining his PhD at Cambridge University and helping to build a detector for the UA2 experiment at CERN's proton-antiproton collider, he became a BBC TV producer and director. His 1996 documentary led to the highly successful book Fermat's Last Theorem, which was soon followed by The Code Book in 1999.
His latest offering goes deep into fundamental physics and back a long time in the history of science. The first three chapters of Big Bang trace the histories of cosmology, gravity and astronomy; that's already 260 pages. Chapter 4 explains how cosmology and nuclear and particle physics became entangled, and Chapter 5 goes on to show how the conflict between two opposing cosmological pictures - the expanding Big Bang universe and a steady-state universe - was finally resolved after the discovery of the cosmic microwave background radiation. With such difficult and complex subject matter, the two-page summaries at the end of each chapter are helpful.
Its main aim is to explain physics and cosmology, but the book is packed with interesting anecdotes and biographical details. Some of these figures - Einstein, Newton, Rutherford - need no introduction, but others are less familiar, such as Fritz Houtermans, one of the first scientists to consider the role of nuclear transformations in the Sun, and Henrietta Leavitt, an unpaid volunteer who went on to show how variable stars can be used as accurate gauges of astronomical distances.
I enjoyed Singh's coverage of the colourful George Gamow and the controversial Fred Hoyle, two devout non-conformists who duly ruined their own careers. Hoyle makes two major appearances: first for his role in developing steady-state cosmology, where we read the transcript of the 1950 BBC radio broadcast in which he sceptically introduced the term "Big Bang" for the rival theory; and then for his key contributions to stellar nucleosynthesis, where he was snubbed by the Nobel committee. Nevertheless, it is surely for carbon-12 that Hoyle will ultimately be best remembered in science, so his portrait caption "most famous for his steady-state model of the universe" is off-target. The book goes way back in history, for example with several pages on Eratosthenes' pioneering third-century BC measurements of the size of the Earth. However, the book reads as if most of it was written 10 years ago, when the measurement of the cosmic microwave background by the COBE satellite in 1992 confirmed the Big Bang picture. Some crucial topics and recent developments are compressed into a brief epilogue. In this postscript-like final chapter, Singh whistles past the key mechanism of inflation that powered the Big Bang, and resolves the flatness and horizon problems of the universe. He gives scant attention there to the need for the universe to be mostly made up of mysterious dark matter (a pity, because the topic is briefly introduced earlier with the colourful character of Fritz Zwicky).
Also skimmed over are the recent measurements from high-redshift supernovae, revealing the need for a mysterious repulsive force which acts against gravity and modulates the basic Hubble expansion. This "dark energy" is redolent of Einstein's cosmological constant which, as Singh does explain at length earlier in the book, was painfully conceived by the great man before he was forced to abandon it in 1931 in the face of then-overwhelming evidence. Einstein could have been right all along and this abdication premature.
Gordon Fraser.
Computers Ltd - What They REALLY Can't Do by David Harel, Oxford University Press. Paperback ISBN 0198604424, £8.99 ($14.95).
Yes, computers can't do the mathematically impossible, which is what this book is all about, but they also seem to be unable even to do spelling/grammar checking. The phrase in the preamble "You don't go looking for triangles whose angels add up to 150° or 200°" passes easily, though it is an early point in the book at which to invoke the supernatural. I counted no fewer than 24 typos before I gave up. The use of non-metric units in a science book on maths is also not quite what I expect.
These mistakes and the split infinitives apart, Computers Ltd is a gripping book. It is enthusiastically written with excellent examples. It builds pleasingly from easy stuff up to a range of difficult-to-understand NP problems. After each individual topic, I felt delighted to have been given a refreshing view of a subject with which I'm not entirely unfamiliar. Yet each following chapter surprised again with the clarity that Harel achieves in introducing the next "bad news" - the apt term he uses for impossible computing tasks. Also amazing is the book's ability to build on the examples given before. This in itself is highly useful, giving a good view of where the author will ultimately lead the reader.
There is an interesting chapter on whether or not humans can do better than computers, rightly concluding that machines still lack the embryonic ability to deal with the vast amounts of context that we humans obtain through all our senses all the time. And context is certainly needed before one can even start to discuss understanding.
All along I was wondering whether there would be something about "almost possible" solutions. The fact that some task has been proved impossible in the general case does not mean that we can't get acceptably good solutions for everyday use. And yes, there is a short chapter on these issues, but it does not go very far and is a little too purist for my tastes. That said, Computers Ltd is a very entertaining book that anyone with an interest in computers should read.
Robert Cailliau.