Robin Gandy (1919-1995)
Robin Gandy (1919-1995) |
|
 Robin Gandy in 1953 | Robin Gandy's death on 20 November 1995 has ended the strongest living link with Alan Turing, with whom he was all of intimate friend, student and colleague. He inherited all Turing's mathematical books and papers; and thereafter also carried forward part of Turing's intellectual tradition; more precisely he took on the subject that Turing lost interest in, by becoming a pre-eminent British figure in the revival and renewal of mathematical logic. His own many students through the 60s, 70s and 80s were conscious of being Turing 'grandsons'. This unique status was a glory but perhaps also a burden for him, for there was something in the legacy that he could never finally discharge. At the moment of his (quite sudden) death he was still working on the definitive edition of On Computable Numbers for the remaining volume of the Collected Works. |
Mathematical logic is not well represented on the Web as yet, and I cannot point to a site that would detail or explain his contributions. However Wilfrid Hodges (NO genetic relation!), [W.Hodges@qmw.ac.uk] is building a collection of material. There were many biographical contributions at a July 1996 conference in his honour.
At Turing's death in 1954 Robin Gandy was a lecturer in applied mathematics at the then Leicester University College. But after a stint at Leeds, Max Newman brought him, as he had earlier brought Turing, to Manchester; thus after 1961 he was fully engaged in his true vocation, creating joint schools of mathematics and philosophy. Professor there after 1967, he then left for a Readership at Oxford in 1969, and found his home in the then sparkling new and modernistic graduate college at Oxford, Wolfson College. In 1993, I think, he gave an interview to the Wolfson College student magazine, Romulus. (I have corrected spelling errors and obvious false transcriptions.)
Romulus: What brought you to mathematics?Robin was conscripted and his Cambridge study was deferred for six years. The interview went on to describe his meeting Turing in 1944 at the MI6 establishment, Hanslope Park. That's how how he came into my book. And that's why I came to know him, because he was my most important personal source. At this point a picture, taken in November 1986, is worth a thousand words:Gandy: It was when I was somewhere between ten and eleven --- I remember exactly how it happened, I suddenly discovered that if you added up consecutive odd numbers you always got a perfect square. I thought that was an absolutely magic fact. I really felt quite faint with pleasure and went around for the next few days thinking that life was quite wonderful... Eventually I got an exhibition in Maths at King's [Cambridge] and then ... I also became very political in the first year of the war and that took up a lot of time.
Romulus: Had you always been interested in politics?
Gandy: At school I was rather unpolitical but then after a while at Cambridge... I realised that if one wanted to do anything the Communist Party were the only people. There was the Socialist Club, but that was run by the Communists. So I joined the Communist Party and became one of the people who ran the Socialist Club.... Eric Hobsbawm was there in the first year; after the war he became a close friend. When I joined the Party the King's College cell consisted of two people including myself! The days of the romantic intellectual Marxists were gone. It was less modish. I left Cambridge and the Party but was a fellow traveller for a while, but as the years went by it became clear to me that communist rule was bellicose, tyrannical, inefficient and corrupt.
*Steve Huggett, Robin Gandy, me, Ken Barrow, Peter Chadwick |
The occasion? It was after the first performance of the play based on my book, Hugh Whitemore's Breaking the Code, at the Haymarket Theatre, in London's West End. I had nervously invited Robin to come, highly conscious that he might feel the whole thing a disgrace to Alan Turing's memory. But he was as generous and positive as he had been to me all along since the day in February 1977 when I marched into his office and announced my ambition to write Alan Turing's life. After the final curtain we went round to the stage door, and it appeared that he knew the way; he explained to me that a gentleman admirer had taken him backstage in 1936! For Robin somehow the thick throng round Derek Jacobi parted as if for royalty, while Clifford Williams, the director, was pointing and stage-whispering "He was Alan Turing's lover!".
Which he wasn't, let me make it quite clear.
Anyway, there he was afterwards in the London house of my own dear friends Peter Chadwick and Wesley Gryk, who hosted a late evening party for first-nighters, and you can almost see the way he talked. Sometimes it was very hard to get a word in; he had a most annoying way of guessing what you were going to say, and starting to answer it --- beginning "I see! I see!" --- before you had finished. You might also think he just made a joke of everything. But he more than anyone between 1977 and 1983 read everything I put before him in draft, taking every word seriously. Given time, he showed tireless generosity, bringing out the best in other people, which was why he was so much appreciated by his many students and why he will be greatly missed by the logic and Wolfson College communities. The book I was writing differed utterly from anything he would have written, but he never tried to dictate its form. He was open to every topic --- science, literature, politics, sex. I felt in him a contact with a lost British culture of the past, not just the link with Alan Turing. Besides --- it wasn't every day that I met someone who could remark "Burgess, yes you could see he was a rogue, but Maclean, one wouldn't have known..."
1986 was his formal retirement year, but I saw him often thereafter. The last time was on 1 November 1995, at a lunch arranged by Donald Michie at Balliol College. He was in full form in dialogue with Roger Penrose. As I left to do my tutorial, I thanked him for lending me his Morphogenesis volume of the Turing Collected Works. "Didn't find Turing's notes very clear", he said; that was the last I heard him say. I was going to ask him to my new house, ostensibly to discuss my Dictionary of National Biography commission, really to return his many dinner invitations. And I might have shown him this Website. It wasn't to be, but I can imagine so easily what it would have been like, me feeling an idiot as he peered owlishly at the Netscape screen and muttered, puffing as he worked up to a laugh that would have audible across the road; at the end leaving me feeling he hadn't really listened --- only to hear him another time make some very sharp remark about it.
Although his personality filled any room he was in, he was modest about his own work, and it strikes me now how little he made of it in our lengthy discussions of the 1948-1954 period when he knew Alan Turing best. He said more in the interview:
Romulus: What was Turing like to work with?Gandy: He on the whole let me just get on with it but when I did come up with something he would criticise it very acutely. He once reduced me to tears by criticising one of the things I'd written. Quite rightly, I left Cambridge without even really deciding what I was going to write a thesis about....
To begin with I would really have liked to have been a theoretical physicist on the fundamental side. But it was all such a mess. I realised quite rightly that I simply wasn't tough enough mathematically to cope. All the integrals diverged but the people who were good at it managed to get round that. Then I became very interested in the philosophy of physics, and very keen on Eddington. I realised that if I was going to do the philosophy of physics then the thing was to find an appropriate logical setting for it all. Turing suggested quite sensibly that a theory of types was the thing. You can have types for different physical quantities. There's a type for charges and one for masses and so on. There was a lot of talk in the late forties about the meaning of indistinguishability and the difference between Einstein-Bose statistics and Fermi-Dirac statistics. Turing said, why don't you think about indistinguishability. You have a type of individuals and you can think what does it mean to say that they are indistinguishable. And I think that it was really within a fortnight or so that I realised that the logical operations were more or less invariant and that this might lead to a proof of the independence of the axiom of choice. I was very sad to hear that it had been done in 1922! However the business of constructing the right kind of models made a nice contribution to my doctoral thesis. After that I became rather less ambitious in the philosophy of physics and concentrated on the theory of types.
 Robin Gandy in about 1970 |
As it happens, my own research work has been precisely on the problem of eliminating those divergences, using Roger Penrose's twistor theory, that Robin modestly claimed he was not tough enough for. More than once he said to me, "I must get you sometime to explain the Twistor Programme to me," but we never got round to it. Again, it surprises me now that he allowed me such latitude in formulating the structure of "the logical" and "the physical" in Turing's work; he never tried to impose any line, even though it was an area so close to his own motivations.
In 1989 Roger Penrose came out with his far more concrete and radical suggestions relating logic and physics, with a hypothesis that implied Robin amongst others had been for forty years knocking the nail squarely on its side; and here again Robin responded particularly constructively, though with disagreement, in the discussions that ensued. He felt, he told me once, that he'd enjoyed a security in his background, which left him free to be objective, without the need to defend territory; others, he appreciated, had not started with such good fortune. |
More about how he saw things:
Romulus: What do you think of the state of logic now?Gandy: Once many mathematicians believed that logic would provide a secure foundation for all mathematics: a system of notation and rules for its manipulation which would be generally acceptable, consistent, complete and universal. By the 1930s it became clear that this heroic project must fail in each of these respects. Mathematical logic broke into pieces. The study of the various failures led to more or less separate lines of enquiry: non-classical intuitionistic and constructive logics: the theory of proofs and of decidability (or computability): the study of particular models and new axioms for set theory; model theory. Each of these lines has led to deep and beautiful results. But why should ordinary mathematicians be interested in the reasons for the failure of an utopian project? By and large work in mathematical logic seems to be on the fringe. At the moment there is just one notable exception --- model theory (which is very active in Oxford). There purely logical results about restricted languages lead to or help with the solution of particular algebraic problems.
Romulus: What about computer science? Do you think that it has lived up to all you hoped for it?
Gandy: This is one of the things that I wanted to say. The philosophy of computer science is one place where logicians can really say something. Dana Scott and others after all made fundamental contributions to the way that one should think about computing. These ideas came partly from category theory and partly from straight logic, and naturally enough partly from recursion theory. If I was younger I would have switched part of my attention to conceptual problems in computer science..... In all branches of logic and category theory conceptual analysis is an important --- perhaps the most important --- activity. One tries to make a vague idea precise, or to dig out what a huge mass of disparate situations have in common. Success (for example Turing's analysis of the process of computation) is deeply rewarding. I, alas, get too side-tracked and frustrated by the details to be wholly successful at it.
After my book was published he had a mildly annoying way of producing further good Turing lines that I wished he had told me. Did he know any more about Alan Turing's death? --- it seems so strange that he visited just ten days before what he himself never doubted was suicide. I don't believe so: I think that if anything, he failed to hear the subtle, allusive, elliptical remarks that Alan would drop. His forte was much more in jollying Alan along. Was Alan ever in love with him? Again, I think not. He did tell me later that just once Alan had asked for sex, but that he'd said "Oh, Alan, that's not really my cup of tea."
 Another picture taken at that party |
What was his cup of tea? My book mentioned his motorbike leathers, Knightsbridge Barracks, and teasingly that his interests were "more uniformly distributed" than Turing's. My impression is that the leathers and the uniforms were mostly associated with men, but Robin nevertheless always conveyed a rather Bloomsbury-ish bisexual ideal, flirting with the whole world, not identifying with a minority. Alan Turing was much more a committed in-your-face Queer (the 1990s allow me to bring back the word of the 1950s), but Robin rather took the line that he should diversify, as I reflected in a story I took from him (page 485): Alan asking him, faux-naif, "Is that what you call a pretty girl?".
A further letter from Alan to Robin, written in 1953 or 1954, has now come to light. Complaining that Robin hadn't been in touch, Alan asked "have you got yourself a boy or something?", adding in one of his throwaway postscripts "I'm getting slightly hetero, but it's fearfully dull". The blunt question was perhaps a dig at Robin's attitude. Was Alan really trying to become "slightly hetero"? I doubt it: not if he found the idea "fearfully dull"; though I suspect he discovered the new fun of safe flirting as a gay man with Lyn Newman. |
Robin somehow escaped trauma: enjoying not just a secure life, but a charmed life; whether in teenage affairs with older men, touring Albania in 1938, being a Communist student in MI6, having Turing as supervisor, roaring round as a Fifties leatherboy, or making a career out of a very isolated and difficult research field, he emerged unscathed. In the interview he said:
I don't think it's a exaggeration to say that I rather assumed that I wasn't going to survive the war. After all one knew plenty of people who hadn't survive the First. There seemed to be no reason to assume that the Second war would be all that different. I'm saying that because it's fair to say that I didn't worry too much about what was going to happen to me in the future.
He did survive. The war and other diversions meant that his career only took off when he was in his forties, but he still came through on top. But, like Alan Turing, Robin had his chaotic side. That party in 1986 went well; plenty of wine went through those plastic cups; the taxi was called too late and then lost its way; Robin and Stephen Huggett missed the last bus back to Oxford and, I heard later, sat in Paddington Station until early morning.
 Wolfson College |
I remember Robin giving a Wolfson Lecture, in 1986 I think. Physicists always use transparencies for lectures; logicians don't; but for once Robin was trying the unfamiliar technology. His audience was spellbound when, his line of writing being off the screen, he vainly tugged at the image rather than at the transparency. A perfect muddle of logic and physics, that endeared him to all. Of course, he continued with brilliant clear speech. |
I can't be the only one to have seen, in glitches of time like this, the shadow of the would-have-been Grand Old Boy (only seven years his senior) of the 1980s. Did the shadow haunt his mind? Who knows.
*The other people caught by chance in the photograph: My friend Peter Chadwick was the host. Stephen Huggett was and is my collaborator in twistor theory. Ken Barrow I knew from work in the Lesbian and Gay Archive project around 1981-83. He was also a biographer, but of film stars. Ken died of AIDS a few years later.
 Alan Turing Home Page |
 Bibliography |
Last updated 10 July 1997.
Feedback, new links, to me at
andrew.hodges@wadh.ox.ac.uk