Speaking Volumes: Douglas Hofstadter's Godel, Escher, Bach: An Eternal Golden Braid

Speaking Volumes: Douglas Hofstadter's Godel, Escher, Bach: An Eternal Golden Braid.

Eighteen years old, and I should have been revising for my A levels. A trip to the bookshop left me with a 750-page tome of pure distraction that severely interfered with my swotting. Apart from being exhil- arating and challenging, the book gave me an excuse I had been after. It also gave me a career.

At 16, I had gone to Abingdon College of Further Education to do A levels in art, computer science and human biology. One of the lecturers said this was a poor combination which might make it difficult to get into university. But these were the only three subjects that interested me, and I found it almost impossible to study something I was not interested in. So I stuck with this choice, but I needed a better excuse. By the end of the first year, my future on the art course was looking insecure: a picture I had drawn of Margaret Thatcher and Ronald Reagan in a compromising position with a cruise missile had attracted negative reviews, and I had to start thinking about switching to another A level. I did not much care what degree course I did, I just wanted to move to a big city. I switched from art to maths, cramming the first year over the summer holiday. My new set of subjects meant I could apply for a place on a computer science degree course. In the interviews I attended, my initial choice of A-level subjects and my motivations were questioned. I still did not have a good excuse - but the book that interfered with my revision gave me one. It was Douglas Hofstadter's Godel, Escher, Bach: An Eternal Golden Braid (1979). I had never before read such an imaginative and entertaining ride through such a span of subjects. Hofstadter draws on material from biology, mathematical logic, physics, psychology, and linguistics, with recurring references to the art of Escher, the music of Bach, and the maths of Godel.

This book introduced me to the possibility of studying the mechanisms of human thought. I already knew from my human biology classes that not much was known of how the brain works, but the possibility of using computer models to offer explanations was new to me. I was captivated. Now, although a little late, I had an excuse for my choice of A-levels and a decent reason for doing a degree.

While I should have been reminding myself of the anatomy of the ear, the design of airline-ticket-booking computer networks, or the tedium of parametric calculus, I was ploughing through Hofstadter's book, trying out the ideas on friends.

Arriving at Leeds University, I was disappointed to discover that it was recommended reading only for a final-year course module: I would have to spend two years studying core computer science modules before I could start on artificial intelligence.

I sat it out, and at the end of my final year I liked AI so much, and was so horrified at the thought of getting my hair cut and wearing a suit, that I decided to do postgraduate studies. I moved to Sussex, did an MA and then a DPhil in cognitive science. I got hired as a postdoc, and then as a lecturer. Now, after ten years at Sussex, I have had my hair cut, bought a suit, and I am off to sample the delights of working at MIT. I have not read the book again; I suspect I would just pick holes in it and I doubt it would work the same magic on me. But Hofstadter's enthusiasm did rub off on me back in 1984, and for that I am very grateful.

Dave Cliff was lecturer in computer science and artificial intelligence, University of Sussex. He is currently an associate professor of computer science and engineering, Artificial Intelligence Laboratory, Massachusetts Institute of Technology, United States.

He is the author of A History of Broadcasting in the United Kingdom.