COPElAND | 153
ABCD the codebreaker used the rules of Tunny addition to add A to B, B to C, and C to D.
One might think that adding the letters of an encrypted message together in this way would
scramble the message still further, but Turing showed that delta-ing in fact reveals information
that is otherwise hidden. Delta-ing is a crypto-technique that remains applicable today. It was
the basis not just of Turingery but of every Tunny-breaking algorithm used in Bletchley Park’s
soon-to-be-developed electronic computers. The entire computer-based attack on Tunny
flowed from this basic insight of Turing’s.
Turingery itself, though, was a paper-and-pencil method that was carried out by hand.
From July 1942 until June 1943 Turingery was the codebreakers’ only weapon against Tunny,
and during this period they broke an astonishing total of approximately one-and-a-half mil-
lion letters of ciphertext. But no matter how fast the hand-breakers worked, they could not
keep pace with the increasing volume of Tunny messages. The codebreakers needed a machine
to help them—and because of Tunny’s complexity, a machine much faster than Turing’s bombe
was required.
Codebreakers using Turingery had to make use of their insight—Turingery depended on
what you ‘felt in your bones’, Tutte said. Deprecatingly, he described Turingery as ‘more artistic
than mathematical’.^5 Tutte desired a Tunny-breaking method that required no use of what you
felt in your bones—the sort of method that a machine could carry out. He found it in November
1942, using an ingenious extension of delta-ing; it was his second great contribution to the
attack on Tunny.^6 But there was a snag. Tutte’s method demanded a huge amount of calculation.
If the method were to be carried out by hand, it might take as long as several hundred years to
decrypt one sizeable message.^7
machine against machine
At this point Turing’s teacher from Cambridge, Max Newman, entered the battle against Tunny
(Fig. 14.6). Back in 1935, in a lecture on mathematical logic, Newman had launched Turing on
the research that led to the universal Turing machine, and in the following year he had assisted
Turing with preparing his groundbreaking paper ‘On computable numbers’ for publication (see
Chapter 40).^8 At the end of August 1942 Newman left Cambridge for Bletchley Park to join the
codebreakers of the Research Section in their struggle against Tunny.^9
When Tutte explained his theoretical Tunny-breaking method to Newman, Newman imme-
diately suggested using electronic counters to do the necessary calculations at high speed. It was
a ‘Eureka’ moment. Newman was aware that, before the war, electronic counters had been used
to count radioactive emissions in Cambridge’s Cavendish Laboratory, and in a flash of inspira-
tion he saw that this technology could be applied to the very different problem of breaking
Tunny messages. With Turing’s assistance, Newman sold his idea to the head of codebreaking
operations at Bletchley Park, Edward ‘Jumbo’ Travis, and was put in charge of building a suit-
able machine (Fig. 14.7).
The prototype of Newman’s machine, installed in 1943, was soon dubbed ‘Heath Robinson’,
after the famous cartoonist who drew absurd devices. Smoke rose from Heath Robinson’s
innards the first time it was switched on.^10 Newman’s wonderful contraption proved the feasi-
bility of carrying out Tutte’s method by machine, but Heath Robinson was slow, and prone to
inaccuracy and breakdowns.