Sunday, December 28, 2014

Spoiler Alert: Correcting The Imitation Game

Spoiler Alert. I am going to point out historical inaccuracies in the otherwise excellent movie, The Imitation Game, about the life of mathematician-cryptanalyst-computer scientist, Alan Turing, who was the most important person in the Ultra Project at Bletchley Park in UK during WW II that led to the important breaking of the German Enigma machine's codemaking system, and, perhaps more importantly, the father of the successful effort to use this intelligence without revealing that they had done so with their bombe computer.  Indeed, it was only in 2012 that the two papers written by Turing with I.J. Good ("Jack") on how this Bayesian statistical method worked were finally declassified. So, if you have not seen the movie and do not want anything given away, do not read this any further.

Let me note, that while I shall be focusing on distortions and elisions, the movie is much closer to being historically accurate than A Beautiful Mind about John Nash.  Of course, they both appeal to a broader audience by emphasizing problems faced by the protagonists, Turing's tragic position as a genuine gay martyr (the logo for Apple, Inc. is a partly bitten apple, memorializing the cyanide-laced apple Turing bit when he committed suicide in 1954 following his chemical castration for being convicted of "indecency" in 1951).  Nash suffered from being poorly treated for his paranoid schizophrenia.

Despite its deviations from reality, The Imitation Game is an excellent movie (as is A Beautiful Mind), fully worthy of the 5 Golden Globe nominations it has received, including for Best Picture, Best Actor (Benedict Cumberbatch playing Turing), and Best Actress (Keira Knightly playing Joan Clarke).  So, on to correcting it.

What is most wrong with the overall plot?

The main problem is overstating and misrepresenting Turing's role in what happened at Bletchley Park, including his relations with his co-workers, although he was indeed massively socially inept.  He is presented as coming in and being completely alienated from all of them, with one of them, chess champion Hugh Alexander, the leader of the Hut 8 group he was in, and angry with him.  The others are all  making unproductive efforts at cracking the Enigma code in real time, whereas Turing supposedly has the idea of building a machine to beat the machine.  All the others are shown as viewing this as rank insanity as well as by those above them.  Turing is seen as going around all of them to appeal successfully to Winston Churchill, who initially provides funding and support and eventually puts Turing in charge over Alexander and the others.  They eventually come around to helping and supporting him, although pretty much all the crucial insights are seen to be by him.

This is mistaken on several grounds.  The first is that those running the Government Code and Cypher School (GC&CS) operation already knew that they wanted to build a machine to beat Enigma, this having been decided when GC&CS got their enigma machine from Polish intelligence in July, 1939, with the Poles already having developed a primitive machine to decode it.  Turing was consciously hired to improve this machine, based on his famous writings on computers already out there, and not only was he not reviled and ignored, he was initially in charge of the operations at Hut 8, remaining so until late 1942, when he was replaced by Hugh Alexander, his deputy up to that point, who had been running the day to day operations anyway.  Needless to say, showing it the movie's way heightens the drama, not to mention the daring heroism of Turing.

Who is Most Misrepresented?

John Cairncross, the Soviet  spy in Hut 8.  He was not in Hut 8, but in Hut 6 and worked as a translater and linguist, not as a cryptanalyst. His widow claims he did not even know Alan Turing, although that is not certain.  But he certainly did not work with him.

It is true that Cairncross was a Soviet spy, as is recounted in his memoir from 1993, The Enigma Spy.  He is even more  famous for having been allegedly the Fifth Man of the "Magnificent Five" Soviet spies in Britain, led by Kim Philby, all of them idealistic communists recruited at Cambridge University in the 1930s.  He was the last caught and may have provided more useful intelligence over his career than any of them, serving on a Treasury committee that oversaw the Defence budget in the late 1940s.  At Bletchley Park he sent apparently sent useful information to the Soviets during the Battle of Kursk, and after the war he reported to them on British plans to build nuclear weapons.

Besides that he did not work with Turing and was not in the same hut, and was not a cryptanalyst. Another error is that he is presented as having been consciously put in there by MI6 Director, Sir Stewart Menzies, to feed selected information to the Soviets.  There is no historical evidence of this, and that he operated for a long time after the war without being caught is the best evidence that this is not historically accurate.  However, in the movie, it made Turing's role even more dramatic in that he is put into the position of having to determine what got fed to Cairncross to feed to the Soviets, although in fact they may not even have known each other, the place being highly compartmentalized.   I cannot leave this without noting that Cairncross had a moderately well known development economist brother, Alec Cairncross.

Intermediately misrepresented.

Hugh Alexander has already been mentioned, although in general he comes off pretty well, even if the precise facts of his relationship with Turing are off, with him replacing Turing as Hut 8 head rather than the other way around.  Also, when Turing made his appeal to Churchill for more assistance in October, 1941, it was jointly with Alexander and two  of his chess champion friends, both of them operating as cryptanalysts out of Hut 6, Sir Stuart Milner-Barry and Gordon Welchman.  In fact, it was Milner-Barry who carried the letter to Churchill and who would in 1992 lead the effort to prevent Bletchley Park from being torn down (it is now a museum).  (Another chess champion friend of Alexander's there was Harry Golombeck, a famous chess commentator after the war.) While in Hut 6 (which dealt with German army and air force enigmas, while 8 focused on naval ones, apparently more challenging), Welchman worked with Turing on building the bombe, and was the person who suggested the speed increasing diagonal wiring that is presented in the movie as being suggested by Hugh Alexander.  However, neither Milner-Barry nor Welchman a(nor Golombeck) appear in the film, even if their fellow Hut 6 member, Cairncross does.  A curious detail is that Alexander, Milner-Barry, and Golombeck were all participating in a chess olympiad in Buenos Aires when WW II broke out in 1939, immediately leaving for England to join GC&CS.

As for Hugh Alexander, he would join the successor to the GC&CS, the General Communications Headquarters (HCGQ) in 1949 as its Director of Cryptanalysis, which he did for 20 years until 1969.  This removed him from publicly playing chess, although he reportedly defeated later world champion, the Soviet Mikhail Botvinnick, in a game played by radio. GCHQ is the UK equivalent of the US's NSA.

I.J. ("Jack") Good has already been mentioned above as working closely with Turing.  This continued after the war, and Good was also working in the same computer science lab at the University of Manchester that Turing was when he was arrested in 1951.  In short, in reality, Turing may have been intellectually closer to Good than any of the others.  However, in the movie, he is shown treating Good with little respect.  There is some truth to this in that when Good first showed up, Turing caught him sleeping on the floor (something one would expect of Turing actually) and was very angry.  But he eventually came to appreciate Good's goodness very much.  In later life Good, who also worked for awhile at GCHQ in the late 50s, would move to the US, ending up on the faculty of Virginia Tech, dying in Radford, VA in 2009 at the ripe old age of 92.

Peter Hilton is the youngest of the Hut 8 group shown in the movie, and he was indeed only 18 when he first started working there.  That is accurate.  However, it is Peter Hilton who was the one who had the dramatic insight about looking for "Heil Hitler"in the messages to decode that is attributed to Turing in one of the movie's most dramatic scenes, which happens in a tavern, in this regard replicating a similarly fallacious scene in A Beautiful Mind, when Nash supposedly discovered the Nash equilibrium in a bar, although the game described there is the Blonde in the Bar game, not precusely what Nash wrote about.  Hilton's having this idea was partly due to his knowing German very well as well as being an excellent cryptanalyst.

Hilton was a successful mathematician after the war, and I even knew him during 1962-63 after my late father hired him to join the math department at Cornell University, where Hilton stayed for  many years, even though we moved away in 1963.  I can attest that he was a very witty and interesting man,whom I enjoyed being around.  He was expert in homotopy and homology theory.

 On the latter he co-wrote a book in 1960 with Shaun Wiley, perhaps the most important missing person from the movie, and claimed by some to have been second only to Turing himself in terms of importance to the whole project of breaking the enigma code and also in Hut 8.  Wiley taught at Oxford for a few years after WW II, but became Chief Mathematician for GCHQ in 1958, although he later re-emerged to continue an academic career.  Among others, one of his PhD students was E. Christopher Zeeman, an important catastrophe theorist, and the first to apply catastrophe theory to economics in a 1974 paper on stock market crashes.  I met Wiley once in 1963 through my late father, and found him to be even wittier and livelier than his coauthor, Hilton.  He died in 2009 at age 96.

Who was not misrepresented?

Joan Clarke,the woman friend of Alan Turing.  I thought her complicated quasi-romance with Turing was made up, but on checking the internet, I found this was not so.  She is portrayed very accurately, being a wrangler in math at Cambridge, feeling discriminated against for her gender, and proposed to  by Turing.  Apparently she in fact was not fazed when he confessed his homosexuality to her, and it was he who withdrew from the engagement.  She would work at GCHQ after the war for the rest of her career and did marry a coworker.  Apparently she was a brilliant cryptanalyst, and much of what she did remains classified.

I note that my connection with this runs through my late father, J. Barkley Rosser, (Sr.), who was a PhD student of Alonzo Church, as was Alan Turing.  They never met, my father leaving Princeton in 1934 while Turing arrived in 1936, leaving in 1938.  But their work overlapped substantially, and they knew many of the same people.  However, I  did not figure out until I saw the credits at the close of the movie that I actually had met some of their associates.

Barkley Rosser

Update:  An interesting link is the British Alan Turing scrapbook where one can find how to get at his declassified papers with Good and also a link to a program where  Shaun Wiley, Joan Clarke, and Jack Good all spoke back in 1990.  Turing personally recruited Wiley, whom he had known as a grad student at Princeton.  Recent papers on the Good-Turing stuff are by Freer, Roy, and Tenenbaum and by Favoro, Nipoli, and Teh.

Another update: Lones Smith has made me aware of a claim by Rob Janoff to Holden Frith that the Apple logo had nothing to  do with Turing's apparent suicide, but has the bite in it to distinguish it from a cherry.  According to this, the widely reported story I repeated appeared after an earlier film, "Enigma," appeared in 2001, screen written by Tom Stoppard.  I note that there is a more recent theory that Turing died of an accidental ingestion of cyanide (he definitely died of cyanide poisoning), but that it was not from biting the bitten apple found by him.  As it was, he died on June 7, 1954, one day after the tenth anniversary of the D-Day landings.

10 comments:

Peter Dorman said...

Fascinating, Barkley. Thanks for all this information.

As an ex-chess player, Alexander to me is C. H. O'D Alexander. His name can often be found in books about the great players of the 30s, 40s and 50s. Because he was talented but never fully developed his skills, he often lost excellent games against other top players. All along I had no idea whatsoever that he was engaged in hush-hush cryptanalysis.

Unknown said...

Have not seen the movie. I did read the Hodges book. I think that like A Beautiful Mind (book and movie) it demonstrates that a real genius like Turing is actually beyond the understanding of us ordinary folks. Consequently the central question of their persona remains an enigma.

rosserjb@jmu.edu said...

It is stated in the movie that he was twice British national chess champion, and that is accurate.

rosserjb@jmu.edu said...

My comment above was replying to Peter Dorman's, so "he" refers to Hugh Alexander, not Alan Turing. Curiously, for all his genius, he was not a great chess player like a bunch of others at Bletchley Park, nor did he speak German, this last point noted in the film, although when it had him discovering the "Heil Hitler" he wisecracks about this being some German he does know, although, of course it was Hilton, not he, who figured out the importance of that. Turing was indeed a very difficult to understand genius, but most of the rest of this group was pretty darned brilliant too,

Anuj Gangopadhyaya said...

It's a myth that Apple's logo pays homage to Turing. There's an interview in this report (linked below) with the designer of the logo - the bite in the apple was used to ensure the logo scaled in size and still resembled an apple.

http://www.cnn.com/2011/10/06/opinion/apple-logo/

Peter Dorman said...

Just to clarify, it was no doubt an honor to win a British chess championship during this period, but Britain was not a significant chess country at the time. They hosted important tournaments, but, if I remember correctly, there were no British grandmasters until the 1960s, and no Brit legitimately contended at the highest level until Nigel Short came along. My favorite Garry Kasparov quote, when asked which player he expected to defend his world title against: "It will be Short, and it will be short."

rosserjb@jmu.edu said...

Harry Golombeck became a grandmaster, but it was after the 60s, and he was one of the less important figures at Bletchley Park. Given that Alexander beat world champion Botvinnick, he might have made it if he had not withdrawn so fully into the shadows of the GCHQ world. It should be kept in mind that during the period he was there (ending in 1969), the very existence of the agency and everything surrounding was more than classified. Very few people knew of the existence of any of this, something easily forgotten in today's world where the NSA is regular front page news. In those days, if one worked fully for one of these outfits, one simply disappeared.

rosserjb@jmu.edu said...

Oh, and Alexander died not too long after his retirement. It has been claimed that while he led a healthy life, he did not live as long as might have been expected (dead at 64) due to the stress of his long employment at GC&CS followed by GCHQ at the highest level.

StudentAlephNull said...

Your conjecture that the naval enigma was more difficult than the army and air force model is correct: it had an additional rotor.

rosserjb@jmu.edu said...

SAN,

From what I have read, the naval one had 8 rotors while the army/air force one had only 5.