Sunday, 28 February 2010

From Cosmos to Chaos – Peter Coles ***

There are times when I think there’s a ‘Shoot Ourselves in the Foot’ department at Oxford University Press. It might seem the only explanation for the frequency with which they produce popular science books that are brilliant concepts but terribly executed. In fact, I think the reason is down to their choice of authors. They seem to choose authors on their academic background, rather than their ability to communicate. So what the reader often ends up with is a book that is replete with promise, but that fails to deliver in catastrophic fashion.
This book is perhaps the most dramatic example of this I have ever seen. The subject is fascinating, both in its intellectually engaging nature and its applicability. Yet the contents are a disaster for the popular science reader. Anyone attempting to read it who isn’t at least a student on a maths or physics degree course is doomed to frustration. It simply doesn’t explain enough, and what it does put across makes much too heavy use of maths.
This is hair-tearingly frustrating because the topic is wonderful. In a slim volume of 211 smallish pages, Peter Coles introduces probability theory, the importance of statistics to science and everyday life, the difference between a Bayesian and frequentist approach to statistics and observes how Bayesian statistics could be of great value in getting a better understanding of many areas. And he throws in quite a lot of cosmology, explaining the importance of statistics in the field. Any one of these would make engaging reading in its own right – together it’s a tour-de-force. In content terms this is easily a five star book.
Yet time after time the reader struggles to understand just what is going on. This isn’t helped by an unusually high number of misprints (often missing words rendering a sentence meaningless), but primarily it’s because the language is impenetrable and Coles does not shy away from scattering the page with integral calculus, instantly turning off 90% of the audience. I’ve always felt the famous advice given to Stephen Hawking that each use of an equation halves the audience was exaggerated – and it very much depends how you use those equations – but here they are much too frequent and too complex. Coles also sticks to the representations used in the ‘real’ equations where often these could be simplified by using terms that are more meaningful.
Altogether the resultant effect is huge frustration. For those who can get the point of this book – pretty well any working scientist with a reasonable grasp of maths, for example – it’s highly recommended. And I know scientists are a significant part of the audience for popular science. But a wider audience deserves to access what’s in this book.
I’d like to make a proposition to OUP. How about doing two versions of books like this, the original and one re-written for the general reader by someone who knows how to communicate science outside the scientific community. (I’d even volunteer to write the general reader versions!) Now that really would be something to celebrate. But for the moment there’s more chaos than cosmos in this book.

Hardback:  
Review by Brian Clegg

No comments:

Post a Comment