Search This Blog

Sunday 23 December 2012

The Bletchley Park D-Day Pigeon Code Caper

The subjects journalists take most seriously are the ones they are trained to understand – in particular crime, politics and showbiz gossip. A high-profile report was published last week into why the BBC dropped one particular news story that happened to combine all three of these elements (and which, when it was eventually taken up by a rival broadcaster, proved to be the news sensation of the year). Since the Pollard report ruled out any kind of conspiracy or cover-up, the only conclusion is that the BBC editorial staff dropped the story because they had less than 100% confidence in the reliability of their sources. It’s reassuring to know the BBC has such tremendous journalistic integrity... but this only applies to stories they consider “serious”. When it comes to the more unusual and offbeat stories of interest to Forteans, virtually the opposite is true. Here the concern is not to dig into sources, examine conflicting viewpoints and present the balanced facts, but simply to tell a cosy little narrative with a well-constructed plot and stereotyped characters. The result is often closer to a TV soap opera than serious investigative journalism.

Last Sunday the BBC website ran a story entitled “Has World War II carrier pigeon message been cracked?” The article begins “An encrypted World War II message found in a fireplace strapped to the remains of a dead carrier pigeon may have been cracked by a Canadian enthusiast,” and then goes on to inform us that “the message—which attracted world-wide media attention—was put in the hands of Britain's top codebreakers at GCHQ at the beginning of November, but they have been unable to unlock the puzzle. They remain convinced the message is impossible to decrypt...”

Actually this story is the third episode of our cosy little soap opera. It started on 2 November, with the revelation that 74-year-old David Martin had discovered the dead pigeon, with its mysterious message (right), when he was renovating his house at Bletchingley in Surrey. Early reports suggested the never-delivered message had been en route to Alan Turing’s world-renowned cryptographic centre at Bletchley Park, and that it had been despatched on or soon after D-Day – the Allied landing in Normandy that was arguably the most famous single event of the Second World War. Although pigeons were commonly used to carry messages when there was a need to maintain radio silence, experts said there wasn’t a single other instance on record of these messages being encrypted. There was no reason why they should be, because once in the air a pigeon is a pretty secure means of communication. So this must have been an important message indeed.

When I first read the story, I took it completely at face value – as I’m sure many other people did. It was only when I read the write-up in last month’s Fortean Times (FT296), which was much more detailed than the BBC report, that I started to think about it more carefully. The most striking fact that had been omitted by the BBC (not just in that first report, but in all subsequent coverage) was that the dead pigeon was found thirty years ago, in 1982 – “74-year-old” David Martin was merely 44-year old David Martin at the time. The fact that it took the collective brainpower of Britain’s media three full decades to come up with a newsworthy angle on the story gives one pause for thought.

The idea that a WW2 carrier pigeon message should be found forty years after it was sent isn’t far-fetched in itself. That it should be the only known instance of an encrypted message is a bit of a coincidence... but after all, coincidences do happen. But there are other coincidences. It was sent on D-Day, and it was destined for Bletchley Park. “So what?” you say. “I’ve heard of D-Day, and I’ve heard of Bletchley Park.” But that’s the whole point! As memories of WW2 fade from public consciousness, D-Day is one of the few events, and Bletchley Park one of the few locations, that still have any meaning for the average visitor to the BBC website. To top it all (as the FT article pointed out) there is the coincidence of names between “Bletchingley” and “Bletchley” – a subliminal clue that would be far more at home in a work of fiction than in serious journalism!

The second instalment of the story came three weeks later on 23 November: “WWII pigeon message stumps GCHQ decoders”. This had to happen, because the hackneyed plot template we’re using runs “secret code is found ... experts say it is uncrackable ... code is cracked by amateur sleuth”. The story demands that a plodding, unimaginative, taxpayer-money-wasting Inspector Lestrade must profess himself stumped before our modern-day Sherlock Holmes can burst onto the scene. But if you read the article carefully, the reason GCHQ said the message was undecipherable wasn’t just that they were helpfully conforming to the stereotype of the stuffed-shirt bureaucrat that our story demands. It was because the code has all the appearance of having been encrypted using a “one-time pad”, which no less an authority than Wikipedia describes as “a type of encryption which has been proven to be impossible to crack if used correctly”.

So we come to that inevitable final headline, “Has World War II carrier pigeon message been cracked?”. When a Canadian amateur historian said he’d decoded the pigeon message, the BBC journalists were faced with two alternatives. They could investigate the story properly—and run the risk of it vanishing in a puff of rational argument—or they could run the story with a question mark at the end. They took the latter approach, which is a lot simpler and a lot more entertaining.

If the Canadian solution is to be believed, the message doesn’t use a one-time pad or any other type of cipher. It uses standard military codes that had been around since the First World War – “relying heavily on acronyms”, as the article says. The purpose of the code is not deception but compression. The 136 characters of the coded message, rather than encrypting a Top Secret message of 136 characters, are simply a shorthand way of writing a much longer message of a hundred-odd words.

But how certain is all this? An internet search reveals many attempts to debunk the Canadian solution, both on the grounds that it’s not how things were done in WW2, and that the resulting message contains no information of tactical or strategic value. This blog post, with the comments after it, provides a good summary of what's wrong with the Canadian “solution”, as well as serious and well-informed research towards a true understanding of the cipher used. What is remarkable is that the BBC and other mainstream media have shown no interest either in this debunking of the Canadian work, or in any of the more level-headed but less spectacular research on the subject.

9 comments:

Gordon332 said...

Great article, Many thanks Andrew for filling in some blanks. It puzzled me why it was that the image of the Code page as published in the press carried no crease marks or staining after 40 years in a really dirty environment albeit contained within a capsule.

Do you by any chance have the full image of the page? A high res version would be super.

Andrew May said...

Thanks for the feedback. I'm afraid I just did a Google image search for "pigeon code" and pinched the image I thought was best (from itv.com), which I then enlarged and sharpened slightly. I wanted to make the letters as clear as possible, whereas most of the media seem to have gone to the opposite extreme, by "cleaning" the image so much that some of the letters are very faint. Amusingly, the Canadian solution relies on a misreading of several of the letters (U in different places is misread as LI, W and H)!

Gordon332 said...

Thanks Andrew, Firstly Happy New Year to you and yours, I trust it will be a year filled with interest, happiness and fulfillment.

I noticed the ITV image, far more 'real' than the sanitised versions. The BBC put out another interview and showed a semi sanitised version: http://www.bbc.co.uk/news/uk-20843647

I have done some digital forensics on the ITV version and would like to send some of the images to you for your opinion. I think you'll find them interesting, they show quite a bit of additional detail that you would not normally be able to see. What's the best way to do that?

Andrew May said...

Thanks - happy New Year! Several options come to mind. You could link up with me on Facebook (http://www.facebook.com/DrAndrewMay) or Twitter (http://twitter.com/DrAndrewMay), or you could bite the bullet and post it on your own blog, since I see you have one called Tamam Shud but haven't started using it yet. I'd never heard of the Tamam Shud case until Mike Dash did a blog post about it last year... it's one of the best real-life unsolved mysteries I've ever come across!

Gordon332 said...

Haven't awakened that blog as yet but probably will very soon. In the meantime I have sent a friend request, I have the images on my site an issue being that Facebook images don't present that well.

Unknown said...

What is funny about the story is that the dead pigeon and code were discovered in 1982. In WW2, Leo Marks was appointed codes officer of the Special Operations Executive (SOE) ) who assign poem code to agents. He was still alive in 1982. Why nobody ask him to decode it ? One Time Pad is secure but not practical to use during WW2. Random key for encryption need to be created and distributed (millions of keys need to be created and distributed, which is logistic nightmare). The poem code is less secure but more practical. That is why Leo have a hard time convincing his superior to switch to one time pad. From the way the code was written, it look like poem code. Therefore it is quite likely the code can be broken if we can find the right poem use by the agent.

Andrew May said...

Interesting comments, thanks. I find it very strange that the message wasn't taken more seriously back in 1982, because as you rightly say memories of WW2 and its specific operational practices would have been fresher then (40 years old rather than 70 years old). With a limited number of poems that might have been used, I would have thought a poem code could be cracked with a computer program these days.

Unknown said...

I believe the GCHQ don't want to devote their computing resources to break an outdated code from WW2. They should be spending tax payer money to break terrorist's code anyway. The information in this code is useful for historical reason but not a threat to national security. In his book, Leo Marks said he create unique poems for his agent so that the code would be more secure. Unfortunately, he didn't publish his poems so it would be difficult to guess what is the key for the code. Need to use cribs ( commonly use military terms in that era ) to try to find the key.

Gordon332 said...

go ki hon, i think you are exactly right about GCHQ resources but would qualify it a little. There could be aspects of this code that may still provide relevant training materials for new staff but that of course would have its limits.

Another aspect to consider is the unusual nature of the code and any additional content there may be on the page, it may be that aspects of this document could have a direct relevance to other, similarly aged, documents.