a journal of modern society & culture

Review Essay: B. Jack Copeland, Turing: Pioneer of the Information Age (Oxford University Press, 2012)

Alan Turing was an outstanding British mathematician who joined the Government Code and Cipher School (GCCS) at the renowned Bletchley Park on the first day of the Second World War. He was just 27. Before the war he had made a name for himself by introducing the concept of a ‘universal computing machine’. At Bletchley he unlocked the German Enigma cipher machine and thus became a legendary code-breaker. Immediately after the war he turned his pre-war abstract computing machine into a universal electronic computer. He became a pioneer and advocate of Artificial Intelligence.

B. Jack Copeland’s Turing: Pioneer of the Information Age was published on the centenary of Turing’s birth. Many books on Turing preceded it, including Andrew Hodges’s excellent biography Alan Turing: The Enigma, 1983; and, especially valuable for less numerate readers, David Leavitt’s The Man Who knew Too Much, 2006. All three books are recommended. Although Hodges’s The Enigma was reprinted in 2012 with a preface highlighting the topics that in 1983 attracted scant attention but are now the subject of lively debate, Copeland’s volume has the edge in topicality.

Turing was born in London in 1912, the second son of Julius Turing, an administrator in the Indian Civil Service, and his wife Ethel. Julius arranged leave from India so that his son could be born in England. Alan grew up in a privileged world of cooks, maids and holidays abroad, but it was ‘the life of a near orphan, lodging with carers, and seeing his parents only when they returned from India on leave’. For three years he attended a modest day school to learn Latin (which he never mastered) before being moved, at age nine, to Hazelhurst boarding school in Kent. Hazelhurst prepared pupils for the Common Entrance examination – the test for admission to Public Schools. Turing took a detached, if not hostile, view of Hazelhurst. He had been given a textbook on elementary biology, which opened his eyes to science, and he liked to invent things, but there was nothing on these subjects in the public schools examinations. Indeed, the staff did what they could to damp down his interest in these areas. But his mother discovered a public school that had a science master: Sherborne School in Dorset, in the west of England. Turing was accepted.

Sherborne School is one of the oldest all-male public schools, with monastic origins. Like most public schools, Sherborne was a nation in miniature. The headmaster told students: ‘In your relations with us masters and in the scale of seniority among yourselves, you have become familiar with the ideas of authority and obedience, of cooperation and loyalty, of putting the house and the school above your personal desires’. Turing was an independent-minded young man, but he seemed resigned to the prospect of incarceration there. ‘It is run on the same principle as the Gallic councils that tortured and killed the last man to arrive,’ he wrote to his mother. And in spite of the headmaster’s priorities for intellectual pursuits: ‘Classical, Modern, and Science, in that order’, it was at Sherborne that he became a mathematician and a scientist. But it was touch and go. Even his mathematics teacher, who thought him a genius, sounded a warning. ‘He spends a good deal of time on investigations in advanced mathematics to the neglect of elementary work. A good groundwork is essential in any subject.’ But Turing did well in examinations, and survived.

He eventually became school prefect, won a major mathematics prize, and obtained a scholarship to King’s College, Cambridge in 1931. There he embarked on the most difficult honours course in mathematics and soon was punching above the weight of other new entrants. He attended a series of lectures on methodology of science by Arthur Eddington, professor of astronomy, who suggested to him a topic for a dissertation should he decide to apply for a fellowship. In 1934 he passed his finals with flying colours. The following year he was elected a Fellow of King’s at the age of twenty-two. There were celebrations back at Sherborne School where the following ditty circulated:

Turing

Must have been alluring

To get made a don

So early on.

But Turing was far from alluring at King’s or later in his tragic career. He was a loner in work and not a team player. King’s was feudal in its structure but that would not be unfamiliar to someone from a public school. The freedom to smoke, drink, and work as one pleased was a joy. But the King’s scholars mostly formed a self-consciously elite group, and he was on the margins.  His work often was poorly expressed and seemed muddled.  Max Newman, a fellow of St John’s College, later noted that Turing ‘found it hard to use the work of others, preferring to work things out for himself’. But work wasn’t everything. Turing did become an accomplished oarsman and tennis player – unusual for students who were expected to be either ‘athletes’ or ‘aesthetes’.

Turing attended lectures on the foundations of mathematics and logic by Max Newman. One aspect fascinated Turing so much that it was to lead to close collaborations with Newman. Newman discussed systematic procedures in mathematics (simple methods that anyone can carry out, step by mechanical step, without the need for any creativity or insight.) The basic feature of these procedures is that a machine could do them. Turing worked on this for months without, characteristically, telling anyone. In April 1936 he surprised Newman by presenting a paper ‘On Computable Numbers’. It proposed, among other things, an idealised machine which in Copeland’s words, ‘consisted of a limitless memory – an endless paper tape – and a ‘scanner’ that moved to and fro along the tape, reading it, and in turn printing further letters and numbers on the tape. The machine’s program, and whatever data it needed for the computation, were printed on the tape before computation started. By placing different programs on the memory tape, the operator could make the machine carry out any procedure that a human computer could carry out.’ Turing dubbed the device a ‘universal computing machine.’

The remarkable paper was published in 1936. Turing was interested in building a working machine, but the electromechanical switches of the day were far too slow. Only when they were together at Bletchley did Turing and Newman learn how a fast electronic universal computing machine might be built. Turing accepted an invitation to study at Princeton University where he got to know distinguished mathematician John von Neumann. After a year, Turing decided, with the backing of von Neumann, to stay on at Princeton to work for a PhD, which he was awarded in June 1938. Turing was offered a research assistantship at Princeton’s Institute for Advanced Studies but, with war in Europe threatening, he returned to Cambridge.

The British security services were unable to cope with the volume of German wireless transmissions. The Germans were using increasingly secure versions of Enigma machines, so GCCS planned to take on more cryptanalysts. Turing was a natural recruit who worked for them part-time, until he took up residence at Bletchley Park in September 1939. He quickly became involved in the Enigma problem and chose to tackle the most difficult task of all, deciphering messages from German U-boats. This was of critical importance as Britain relied heavily on North America for food, munitions, and oil and merchant ships were easy prey to torpedoes.

The Bletchley code-breakers were organised into a system of large huts surrounding the main building.  Each hut was numbered and the numbers soon came to refer to the activities within them. Turing and his team moved into Hut 8 to deal with Naval Enigma. Turing may not have been a good team player, but this was his ‘baby’, and he led through enthusiasm. The period in Hut 8 was probably the happiest in his working life. The basic operating system of Enigma machines was well known as they had been introduced in the 1920s for use by banks: at the heart were a key board, a lamp board, and three rotating wheels. Each wheel could rotate through 26 positions (one for each letter of the alphabet) and each could be turned to a different starting point before each message. When a key was pressed, one or more wheels would turn, and a different letter would be illuminated on the lamp board. Plain text entered on the keyboard would be shown on the lamp board in encrypted form. This would be written down and sent by wireless to a receiving machine, where the process was reversed. The German military introduced yet another stage of encryption: a plug board, which greatly enhanced the security of the system. The U-boat version had other security measures built in and, later in the war, four wheels instead of three. The number of possible settings was astronomical.

The earliest successes against Enigma were made by the Polish Cipher Bureau. Just weeks before their country was invaded they passed everything they knew, together with details of a machine for speeding up decoding (called a bomba), to Britain and France. Unfortunately the Polish techniques depended on a single weakness in German operating procedures: a flaw in the method by which the sending operator used to tell the receiving operator which positions he had moved the wheels to before starting to type. Because the Germans might at any time remove this defect (which they did in May 1940), GCCS decided that a new approach was needed.

All this information was available to Turing when he took up the Enigma challenge. The old methods were unsuitable. He realised that they would have to rely increasingly on ‘cribs’, segments of plain text in messages (for example, the standard wording of a weather report) that could be matched to a segment of encrypted text. This would be done by feeding a message containing a crib through settings at which the Enigma cipher process might begin and see if any generated acceptable plain text. If none did, it would be necessary to start again matching the crib to another section of encrypted text.  It would be impossible to tackle the large volume of incoming messages using such a time-consuming process without the aid of a machine. Turing drew up the specification for such a machine, called a bombe, which was much faster and more technically advanced than the Polish device. Each bombe was made up of 30 (later 36) replica Enigma machines connected in whatever configuration seemed best for attacking a given message. The first one was delivered in the spring of 1940 and Bletchley Park suddenly became a code-breaking factory. Turing, meanwhile, became engaged to a female co-worker but broke it off, admitting to her that he was homosexual.

The bombes began to deliver answers, but productive cribs were proving hard to create. The best way to get good cribs was to capture (‘pinch’) decoded German messages. The Navy obliged: they boarded a weather ship and disabled U-boats and took their hardware and paperwork. These activities and continuous improvements to the bombes increased success: the sinking of allied ships sharply declined. The threat that Britain might be tipped into starvation had been removed.

Bletchley faced another challenge. British wireless operators were intercepting streams of messages from an unknown machine. At first the decoding was assigned to a team in the research department which made some progress despite the absence of any knowledge of the hardware involved. The Germans were using a new state-of-the-art cipher machine (codenamed ‘Tunny’), for the transmission of messages between Hitler and his top brass in Berlin, and the commanders-in-chief on the principal battle fronts. With some inspired guesses the team managed to deduce the form of cryptology and how the machine might work. Turing was brought in from Hut 8 and within a few weeks had come up with a paper, pencil and eraser method for breaking the messages. This method became known as Turingery.  Mechanising this process fell to Max Newman, Turing’s mentor and friend, and a brilliant electronics engineer, T H Flowers. Together, and with ample facilities at their disposal, they built a giant all-electronic computer called Colossus using 1,500 vacuum tubes.  It speeded up enormously the decryption of Tunny messages. At the Normandy landings in June 1944 the Allied commanders knew almost exactly the dispositions of all the German divisions. Turing’s contribution was the basis for every main algorithm used in Colossus.

Turing in 1946 was awarded the OBE for his wartime work. It was not a very high honour – two steps on the ladder below a knighthood, and two steps up from nothing. Turing described it as ‘ludicrous’ and kept the shiny medal with its red ribbon in a tin box full of screws, nails and other odds and ends. A better measure of his value came from GCHQ who offered him a large sum of money for consultancy on post-war code breaking. Towards the end of the war the National Physical Laboratory (NPL), a few miles upriver from London, formed a new mathematical research division. Its superintendent, John Womersley, had read Turing’s 1936 paper and had heard about the potential of electronic machines. In June 1945 he invited Turing to join his division. There Turing drew up designs for his stored –program electronic computer and called on Flowers and his team for help. The machine was promptly dubbed the Automatic Computing Engine (ACE). At first all went well, but top management at NPL was in disarray at the time and priorities changed at a whim. Turing became frustrated, although he was in part to blame. In late 1947 he took sabbatical leave for Cambridge for theoretical work on whether a computer could learn by experience. He never returned. Max Newman had established a Computing Machine Laboratory at the University of Manchester and Turing accepted a timely offer of a position. A Pilot Model ACE was commissioned at NPL, and the development of a marketable version was carried out. The result, DEUCE, was immensely successful.

At Manchester Newman’s team had a programmable machine up and running as Turing arrived. It had a smaller capacity than that projected for ACE, and was called the Manchester Baby, but it was the first electronic machine to run a stored program. A larger machine followed, developed by the Manchester firm, Ferranti. Turing played no part in the design of the Manchester Baby, but helped to get it working and used it to pioneer Artificial Intelligence. Other computers were being built in the United States but although designs differed all (except pure number-crunchers) embraced Turing’s 1936 idea of a universal computing machine. The same can be said of today’s laptops.

In January 1952, Turing formed a relationship with a 19-year-old man and brought him home to live. The house was burgled and in the course of their investigations the police learned of the relationship between the two men. Homosexual acts were then illegal and both were arrested for gross indecency. The case was brought to trial. Turing pleaded guilty, but he himself felt no personal guilt. He was given the choice of imprisonment or a year on parole during which he would receive female hormone ‘treatment’. He chose the latter, and took a pragmatic attitude towards his probation, even laughing about it to his friends. Newman gave him full support and Turing was able to use the Manchester computer to assist with his new studies of biological growth, a precursor to the field of Artificial Life. In the following year, with the therapy over, this work was in full flow. But he was never able to work for GCHQ again. The patriot had become a supposedly grave security risk.

In June 1954, Turing was found dead in bed. An autopsy revealed cyanide poisoning. At the inquest, the coroner returned a verdict of suicide. Copeland in his book questions the verdict pointing out that no evidence was presented to indicate that Turing intended suicide, while friends and neighbors had spoken of his recent enjoyment of life. Copeland suggests other possibilities: Turing used cyanide at home to gold plate certain objects, so an accidental death was a possibility. Less plausibly he might have been murdered because of his secret service background. Hodges in his biography accepts that Turing killed himself. No doubt the debate will continue. In 2009 the then Prime Minister, Gordon Brown issued an official apology for Turing’s trial and punishment in 1952-54, but campaigns for a full pardon have so far not been successful.

Colin Hughes is an historian of science and technology and a former Assistant Secretary at the Ministry of Defence in London. He is a regular reviewer for Logos.

Email this to someoneTweet about this on TwitterShare on Facebook