Keeping one jump ahead
Donald Michie
COLOSSUS: BLETCHLEY PARK’S GREATEST SECRET by Paul Gannon Atlantic, £25, pp. 562, ISBN 1843543303 ✆ £20 (plus £2.45 p&p) 0870 429 6655 Over half a century has passed since the Allied victory in Europe in the second world war. In significant measure it was expedited by the interception and decryption at Bletchley Park of the ‘Fish’ cipher traffic over communication links between Hitler’s Berlin headquarters and the army groups in North Africa, Italy and Western Europe. The veil of secrecy over these operations has only recently been lifted, and Colossus by Paul Gannon is the first to have the full story.
I joined BP in the early summer of 1942, and later worked with the high-speed electronic Colossus machines. These enabled our group to expand the daily breaking of enciphered traffic to round-the-clock mass production.
Upon the German surrender I was given the leaden job of condensing the work of many hands into two detailed volumes on how our various tricks were worked. History-writing traditionally allows inferences, conjectures, stories, jokes, and even the occasional grinding of an axe. This was not history. It was an exercise in embalming — a necessary exercise but with no assurance that the embalmed specimen would ever be seen again.
In the mid-1970s small leaks began to occur, fertilising a crop of popular and semipopular books of varying remoteness from reality. After 20 years of this I paid a number of visits over a two-year period to the Government Communications Headquarters (GCHQ), pleading for declassification. On 29 September 2000 the more significant of the two, the General Report on Tunny, was released to the Public Records Office. Gannon’s new book is the first to take advantage of this 505-page report and of a wealth of other key materials that he has brought to light.
In refreshing contrast to many earlier publications, his account of wartime interception and decryption is deeply researched. It is also masterly in its breadth and sweep, with an interesting assessment of the role of the Colossus enterprise in Britain’s post-war computing. But although Gannon makes a glancing reference, he has skimped the whole story. To understand how, we need to grasp a fundamental distinction between 1) The patterns of the rotary wheels used by enemy machines to encipher traffic over given links, and 2) the starting-points, or settings, to which these wheels were set for enciphering each individual transmission.
In February 1944 when the first Colossus electronic code-breaking computer began operation, it was never imagined that it could do more than hugely speed up the finding of wheel-settings, leaving it to the excruciatingly slow and chancy efforts of cryptographers to break the patterns by hand. Yet by this stage of the war the Germans were changing the patterns every day on every link as well as changing the settings for every new message. Gains in the speed of finding these latter were thus nullified. To find settings you must first have the patterns. Yet to break just one set of patterns by hand could take days.
In April 1944 I made to my co-worker Jack Good a suggestion simple enough for the two of us to validate in a couple of hours’ experimentation. It was a way of rejigging Colossus for breaking wheelpatterns, a role for the machine not previously entertained. We reported it at once to our section head Max Newman, and a ‘crash programme’ was authorised at War Cabinet level to have a working Colossus of the new design in time for D-day. At the end of hostilities nine new-design Colossi were operational and 63 million characters of highgrade German messages had been decrypted. Using precious Colossus time for the lowly task of wheel-setting became something to be avoided wherever possible.
The three individuals who could have briefed Gannon on this are the three whom Max Newman first appointed when he founded the ‘machine section’, namely Jack Good, David Rees and myself, all of whom are still alive. The author acknowledges that he did not consult any of us. It is puzzling that a writer of evident diligence and scholarship could commit such an oversight. But I nevertheless commend the book to both the professional and the general reader.
Previous page
