Alan Turing: The Enigma The Centenary Edition

by Andrew Hodges

  It was also in 1969, as the computer came of age, that it was first noted, in articles by Mike Woodger and R. Malik, that Alan Turing had emerged from the war with a practical knowledge of electronics. This fact, quite at odds with the prevailing stereotype of ‘the logician’, as he appeared in H.H. Goldstine’s standard academic account of The Computer from Pascal to von Neumann in 1972, took some time to be assimilated; so likewise did the place of the ACE in computer history. The compilation by B. Randell of classic papers documenting the origins of digital computers (see note 5.23) relegated mention of the ACE to its bibliography, but the mini-boom of computer history did not entirely pass it by: in 1972 the original report was issued by the NPL, and it received a first serious review in 1975.

  Meanwhile the turn of the decade had also seen the purpose of Bletchley Park become mentionable, although the first outright claim for its strategic significance came only in 1974 with F.W. Winterbotham’s The Ultra Secret. This book made no mention of Alan Turing, but in the same year A. Cave Brown’s imaginative work Bodyguard of Lies contained many sentences in which the word ‘Turing’ appeared, sometimes in conjunction with words such as ‘machine’ and ‘Bombe’. The floodgates were opened. Meanwhile Jack Good and Donald Michie had published certain disclosures concerning the electronic machinery at Bletchley. Drawing these developments together, the inquiries of B. Randell, motivated partly by the question of understanding Alan Turing’s part in the origin of the computer, enjoyed some success. His revelations of the Colossus technology did in fact reflect the achievement of Newman and Flowers rather than anything directly attributable to Alan Turing, but it meant that a first serious glimpse of the gigantic scale of operations had been given. Much of this, together with other mid-1970s disclosures, was brought together in a BBC television programme, one of a series on The Secret War, broadcast early in 1977.

  Since 1969 was also the year of gay liberation, it meant that another change in how it was possible to think about Alan Turing occurred during the 1970s. This was not the result of the Wolfenden reforms – which, although delayed through the strenuous efforts of Viscount Montgomery and others, had passed into law as the Sexual Offences Act in 1967. Setting an ‘age of consent’ at twenty-one, this rationalisation and modernisation of the law ensured that the Turing crime remained a crime.* It was rather the brief renaissance of American liberalism that permitted a reversal of the concept of ‘problem’: seeing society as a problem for the individual rather than vice versa. In its way this development created the conditions for a rediscovery of Alan Turing’s life as much as did the other openings of the new decade: not just in that it was possible to say he was a homosexual, but in appreciating his pride, his stubbornness, and the moral force he brought to bear as a very private, reserved, shy man who nonetheless insisted that this was not a matter for hiding.

  Taking these developments together, the 1970s therefore made it possible for someone to see who Alan Turing was, in a way that no one in his lifetime (but he himself) could have done. It so happened that I was placed in such a position as to be struck by all these developments. The name of Alan Turing first impinged on me in the summer of 1968 – essentially a reflection of the burgeoning of computer science, for I was reading about cybernetics and Turing machines as a mathematics student. In fact I did not choose to work in this field, moving instead into mathematical physics, learning about relativity and quantum mechanics from Roger Penrose as a postgraduate student from 1972 onwards, and trying to make a contribution to the Penrose theory of twistors.*

  But in 1973 the name of Alan Turing impressed itself on me again, this time through another compartment of my life. I was then a member of a group that had formed within the London Gay Liberation Front to write a pamphlet criticising the medical model of homosexuality. One of the other members, David Hutter, had heard something of the end of the Turing story from Nick Furbank. Knowing nothing of his secret work, and believing that his death could be accounted for by the effect on his work of the hormone treatment, we included a paragraph which used this idea to illustrate our theme. Thus after twenty years there was a first public squeak of protest.

  Lurking for years at the back of my mind, with a sense that I ought to have found out more about what happened, this story suddenly leapt to the fore again on 10 February 1977. On that day, while having lunch with Roger Penrose’s research group at Oxford, there was a conversation about the famous Mind article which took me back to my earlier interest in Turing ideas – and then quite independently, another about the BBC programme on Bletchley Park, which had been broadcast the previous evening. Roger Penrose commented on its fleeting mention of Alan Turing; he had long since heard talk of him as a man ‘hounded to his death’ but just recently there had been rumours of a man who had ‘deserved an earldom’. Nothing was clear or connected, and it was to be three years before I could compose a coherent interpretation of what had happened – but it was enough for me to sense

  A war O soldiers not for itself alone,

  Far, far more stood silently waiting

  behind, now to advance in this book.

  It had to be done, and it was the right moment to start. A first step was that of collecting extant publications, much as described above. But of course it was necessary to get much closer to my subject. So I turn now to the question of original Turing papers, where my first acknowledgement of a debt must be to Mrs Turing. According to her own book she once told Alan that she was putting things by for the use of a future biographer, and he grunted a gruff assent. Certainly she took the trouble to keep letters from her son’s schooldays onwards, used them for her book, and then in 1960 deposited them as a small archive in the library of King’s College, Cambridge.* She added to these seventy-seven letters a number of ancillary items, such as certain of the correspondence that arose in writing her book. A few other items went to Sherborne School.

  Mrs Turing died at the age of ninety-four on 6 March 1976, and so never knew of me, nor indeed of what her son had done for the battle of the Atlantic. Before she died, however, she endowed an A.M. Turing Trust, to be chaired by Donald Michie, by then Professor of Machine Intelligence at Edinburgh University. It so happened that in 1977, just as I was first making enquiries, the Trustees had arranged for the deposit of all the surviving Turing papers in the King’s College archive. These papers had been kept since 1954 by Robin Gandy, himself now a distinguished mathematical logician at Oxford; but in 1977 they were being sorted and catalogued by Jeannine Alton of the Contemporary Scientific Archive Centre, also at Oxford. That these Turing trails had focussed on Oxford was coincidental, but helpful in my early efforts to grasp my subject. I should like at this point to record particular thanks to Robin Gandy, Donald Michie, and Jeannine Alton, together with other members of the A.M. Turing Trust, who have helped me to marshal support and resources for my endeavour from the start. Others, since 1977, have also played parts of vital significance for this book; but I retain a special gratitude for those who were prepared to assist me at a time when I had nothing but a shaky idea. I should add that the use I have made of this assistance is, of course, entirely my own responsibility; the interpretation I offer of material made available to me is likewise my own.

  The enlarged Turing archive in King’s College, at first sight thin, has indeed provided a backbone of documentary material. Here an acknowledgement is due to Alan Turing himself; he kept nothing in the way of form letters, or the general accretion of correspondence overtaking those rising on the academic scale. But he took care to retain markers of most of the main points in his intellectual life. Indeed he had kept the star globe and the zeta-function machine gear wheels, although these were disposed of after his death. So interested in learning and development, he certainly cared about his own past.

  So much, then, for what was already collected together in 1977. My own enquiries have since elicited further documents from a number of private and public sources, and meanwhile the growing interest in computer history has ha
d the effect of giving me the benefit of recent Turing-related work by other researchers. Yet still the documentary evidence, in itself, would not add up to a portrait of Alan Turing. Only by meeting so many people who knew him could this picture have been drawn. Here again a first acknowledgment must go to my subject himself, who left behind a fund of goodwill on which I have repeatedly drawn. But in describing this aspect of my work, the word for what I have gained is less ‘information’ than ‘experience’.

  The work of compiling this string of symbols on my typewriter has been very different from anything performed by a computer, and different too from my mathematical work in the way that it has required so much interference with other people’s lives. If this book is truly a biography – a writing of life, not a collection of facts – then it is because people have been prepared to allow that interference, and to entrust me with words and ideas that still have living force. The act of disturbing what was a strange kind of Sleeping Death had in fact involved some difficult – and often rather moving – moments. Thus Mr Arnold Murray, in conveying his recollections to me, was also shedding an albatross that had hung around his neck for twenty-five years. For he had returned to Manchester in 1954, only to be confronted by the news of Alan Turing’s death. He was made to feel that he was to blame; and being both particularly vulnerable, and entirely unaware of a larger context, accepted a profound and unmentionable guilt. His success as a musician in the 1960s, and his movement into married life, could not resolve a trauma which had to wait until 1980 for enlightenment.

  This one example must suffice to indicate why, in a number of cases, my gratitude to people who have helped me goes well beyond the terms of a conventional acknowledgment. In some cases this will be abundantly clear from the text; in others the true nature of the debt will be invisible. Nor is the following list even complete; but I should like to thank these and other contributors to my picture of Alan Turing for all that they have done:

  J. Anderson, James Atkins, Don Atkinson, Bob (once Augenfeld), Patrick Barnes, John Bates, S.G. Bauer, Donald Bayley, R. Beadon, G. Black, Victor F. Beuttell, Matthew H. Blamey, R.B. Braithwaite, R.A. Brooker, W. Byers Brown, Mary Campbell (née Wilson), V.M. Cannon Brookes, David Champernowne, A. Church, Joan (née) Clarke, F.W. Clayton, John Croft, Donald W. Davies, A.S. Douglas, Roy Duffy, D.B.G. Edwards, Ralph Elwell-Sutton, D.B. Eperson, Alex D. Fowler, T.H. Flowers, Nicholas Furbank, Robin Gandy, A. Gleason, A.E. Glennie, Harry Golombek, Jack Good, E.T. Goodwin, Hilla Greenbaum, Philip Hall, FRS, Arthur Harris, David Harris, Kenneth Harrison, Norman Heatley, Peter Hilton, F.H. Hinsley, Peter Hogg, N.E. Hoskin, H.D. Huskey, Neville Johnson, R. V. Jones, FRS, W.T. Jones, T. Kilburn, FRS, Leo Knoop, Walter H. Lee, Sir James Lighthill, FRS, R. Lockton, D.C. MacPhail, Malcolm MacPhail, Sir William Mansfield Cooper, A.V. Martin, P.B.C. Matthews, FRS, W. Mays, P.H.F. Mermagen, J.G.L. Michel, Donald Michie, Sir Stuart Milner-Barry, Rupert Morcom, Arnold Murray, D.Neild, E.A. Newman, M.H.A. Newman, FRS, John Polanyi, FRS, F.V. Price, J.W.S. Pringle, FRS, M.H.L. Pryce, FRS, David Rees, FRS, B. Richards, T. Rimmer, K. V. Roberts, Norman Routledge, David Sayre, Claude E. Shannon, Christopher Steed, Geoff Tootill, J.D. Trustram Eve, W.T. Tutte, Peter Twinn, S. Ulam, J.S. Vine, A.G.D. Watson, Mr and Mrs R.V.B. Webb, W. Gordon Welchman, A.C. Wesley, Patrick Wilkinson, J.H. Wilkinson, FRS, Cicely Williams (née Popplewell), R. Wills, Mike Woodger, Shaun Wylie.

  It would be impossible to list all those who, besides these first-hand witnesses, have helped me by answering enquiries, commenting on draft, and in many other ways. But I should like to mention:

  A.D. Childs (Sherborne School), J.E.C. Innes (Old Shirburnian Society), V. Knowles (Manchester University), Simon Lavington (Department of Computer Science, Manchester University), David Leigh (The Guardian), Julian Meldrum (Hall-Carpenter Archives, London), J.E. Taylor (National Archives, Washington), Christopher Andrew, Duncan Campbell, Martin Campbell-Kelly, Peter Chadwick, Stephen Cohen, Cy Deavours, Robin Denniston, Fisher Dilke, D. Dunnill, James Fleck, Steven Hicks, David Hutter, David Kahn, Peter Laurie, Sir Bernard Lovell, FRS, J. Maunder, Roger Penrose, FRS, Felix Pirani, Brian Randell, Jeffrey Weeks.

  One other person, however, has played a decisive role in translating my ideas into the practical form of a book. This is Piers Burnett, who was responsible for commissioning me to write this book, and has seen me through all its difficulties, reading and advising on numerous drafts. Originally my contract was with André Deutsch Ltd, of which Piers Burnett was a director, and which advanced £5500 for the book. (Certain difficulties arising in 1981 on the completion of my work, made a transfer necessary. The first British edition was published by Burnett Books, in association with the Hutchinson Publishing Group.) I have had no grant or subsidy from any other source. While this was by publishing standards a generous investment, and one which made the difference between the existence and the non-existence of this book, it did not easily cover my two years of full time work from 1978 to 1980. Everything had to be done in a Turingesque shoestring style. There may be some virtues in this constraint, but it has certainly placed me in debt to many friends and friends of friends. My archive research and interviews in North America, for instance, would have been well-nigh impossible if I had not been given so many places to stay. At home I have also demanded great patience from those (particularly Peter Chadwick and Steve Hicks) who have had to share the tension and anxiety without the satisfactions.

  Besides Piers Burnett there have been certain other people who somehow could see what I saw in this work, and who supplied a kind of moral support without which I could hardly have continued. One of these people commented very helpfully, at a critical stage, that I ought to leave at least something for others to find out! Certainly there are gaps, and trails I never followed. There will also be errors, though none I hope too serious, both of commission and omission. Perhaps, if the micro-electronic revolution brings about the end of the printed book, it will become possible to undertake continuous revision of published work. Meanwhile, a line must be drawn around an imperfect creation, in the knowledge that Turing scholarship has yet far to go. I shall transfer the documentary material I have accumulated to the Contemporary Scientific Archive Centre, for deposit at King’s College, Cambridge; and I shall likewise pass on corrections and additions that I may receive in the future. Perhaps as important, however, to the continuation of Turing studies, is the fact that as the world changes so will its perception of who Alan Turing was. Even while writing this book the word ‘computer’ has changed its social meaning: a universal machine with the scale and speed of the ACE he envisaged is now sitting on my desk, hardly bigger than my hand. The algorithms embodied in the Bombes now amount to no more than a few lines of BASIC. Personal interaction with a private computer, with its little battles over storage and displays and checking, is now an experience common to hundreds of thousands of people. There is no knowing what this may lead to, but it has changed our perception of the past. If genetic engineering makes advances comparable to those of information technology, then again, his later work may appear in quite a new light.

  Alan Turing presumably thought that eventually a machine would be capable of writing a book such as this. In his 1951 radio talk, set against the opening of the Festival of Britain, he commented that ‘It is customary … to offer a grain of comfort, in the form of a statement that some peculiarly human characteristic could never be imitated by a machine. I cannot offer any such comfort, for I believe that no such bounds can be set.’ There was an element here of épater les bourgeois, especially when he gave as example of a peculiarly human quality, that of being ‘influenced by sex appeal’. Yet he was quite serious in describing the ‘almost certain’ advent of intelligent machines as a development ‘which can give us anxiety’ of a kind much more urgent than that of the Darwinian fright of a century before ‘that we might be superseded by the pig or the rat.’

  In actual fact I should gladly have surrendered my technical problems to a machine – and a word processor would have saved
weeks of chopping and gluing – but these have not been the real difficulties in the composition of this book. More pressing, to give just one very pertinent example, has been the need to overcome the twentieth century chasm between scientific thought and human life – moreover to resist the strong view held in certain quarters that my book should actually reinforce that separation. I have had to live, and even to struggle a little, for my vision; to put this point of view I have had to act this point of view.

  One event of particular interest for me while working on these ideas was the appearance of Douglas Hofstadter’s Godel, Escher, Bach in 1979. It put my work in a Tangled Loop, for central to his book is its exploration of a topic I have brushed aside – the significance of Gödel incompleteness and Turing undecidability for the concept of Mind. I do not myself believe that these results, concerning as they do infinite, static, undisturbed logical systems, have any direct consequence for our finite, dynamic, interacting brains. Far more significant, in my view, is the limitation of human intelligence by virtue of its social embodiment – and this is a problem relegated to a marginal place in Hofstadter’s work as in so many other accounts, though I have placed it at the centre of my own. The study of Alan Turing’s life does not show us whether human intelligence is limited, or not limited, by Gödelian paradoxes. It does show intelligence thwarted and destroyed by its environment. But why, as Alan Turing might have put it, should machine intelligence be any less constrained by worldly reality? Indeed, there seems every reason to suppose that the clever machine will accommodate itself to the crazy demands of the political system in which it is embodied. In the academic sanctuary it is too easy to concentrate on infinitely more theoretical considerations.