Sunday, 23 March 2014

Blind Spot – Gordon Rugg with Joseph d’Agnese ****

I read this book with a mix of responses. One was fascination, the other frustration. The fascination came from the topic, which we catch a glimpse of in the subtitle ‘why we fail to see the solution right in front of us.’ Gordon Rugg, with the help of journalist d’Agnese, gives us a remarkable analysis of how we all – including experts – make errors in our decisions and research. But not limiting himself to saying what’s going wrong, Rugg also provides a method to root out the errors in the ‘Verifier method’, a suite of techniques to pull apart the way we approach, assimilate and make use of information to come to a decision.
The frustration is that the method is seen through a veil of vagueness. We are constantly hearing about this Verifier toolkit, but only get sideways glimpses of what it entails. I would have loved an appendix with a brief description of the contents of the toolkit and a couple of the tools explained in more detail. I appreciate that Rugg and his colleagues probably want to keep the toolkit proprietary (and the danger of having a journalist as a co-author is that they will tend to weed out the detail and weave their stories on people instead). But the book would have been better with a bit more depth.
Having said that, it’s pretty good as it is. The authors take what could be a rather dry topic and give it some life. We see how Rugg started with knowledge elicitation techniques – used, amongst other things, in the attempt to construct expert systems that are designed to provide an accessible bank of expertise. This first requires experts, who often don’t know how they do what they do, to initiate the system builders into their methods and knowledge. From there we move onto looking at the way errors occur in even the most detailed academic study and the gradual realisation that it would be possible to build a series of techniques that would help uncover these errors or, even better, prevent them happening in the future.
There are two major case studies to explore this – the (probably) medieval Voynich manuscript, which for more than 100 years has proved a mystery to all those who have tried to crack its strange script, and the nature of autism. In both cases, making use of the early version of the Verifier method uncovers gaps in expert understanding. While it doesn’t enable Rugg and his colleagues to actually solve the problems, it does provide impressive pointers to where there are currently failings and what should be done next.
All in all, the book will appeal to a very wide market. Whether you are in business (interestingly, the early writing style, before it settles down, is rather like a business book with numbered ‘lessons’ like ‘Experts often don’t know what they know’ in pull quotes) or any branch of academia… or just interested in the nature of knowledge, understanding and human error, there’s a lot here to get your teeth into.
Paperback (May 2014):  
Hardback:   
Review by Brian Clegg

No comments:

Post a Comment