George Santayana gave us the oft-quoted axiom that those who don’t learn from their history are doomed to repeat it, and this was a message that our team took to heart this week.
With the results from last week’s testing of our data in, we decided on several changes to implement that we hoped would give us more consistent results in areas where our initial parameters may have resulted in unevenness, and found new text, specifically from Fanny, to give us results grounded in a stronger base of text from which to draw conclusions.
We also undertook research into the history of authorship attribution this week. One of the most fascinating stories we came across, and will share here, is the case of the supposed Hitler diaries, which were exposed as fakes almost immediately after their sensationalistic publication in 1983 by the German magazine, Stern. The lessons that can be derived from this story have become somewhat legendary cautionary tales of the perils of overconfident and unrigorous attribution, which no publication or scholar (including ourselves) is keen to repeat.
The story of the Hitler diaries is as follows: Stern’s editors were alerted to the existence of the supposed diaries by their publishers, who had in turn been told by one of their marquee reporters, Gerd Heidemann, whose fascination with Nazi culture (he dated Hermann Göring’s daughter) only added to his credibility in claiming to have knowledge and access to the diaries.
It was fitting that Heidemann went to Stern’s publisher before its editors. Any publication with exclusive access to Hitler’s diaries could expect huge sales. Stern had an immense financial incentive to hope the diaries were real, which they passed along to the magazine’s editors. And Heidemann and his shadow team sold the forgery hard. Heidemann claimed the diaries had been lost in a plane crash in 1945 and hidden by a member of the East German Police (G.D.R.) who now wanted to sell them. The sale, according to Heidemann, was conducted via a game of catch — diaries for cash — between a G.D.R. truck and Heidemann’s Mercedes. Heidemann refused to reveal his source, claiming that the man’s life was at risk, which sounded plausible enough.
Stern acquired the sixty volumes of the supposed diaries, spending almost ten million Deutsche marks in the process. Aware the paramount importance of authenticity of such politically sensitive material (the diaries’ claims included Hitler’s ignorance of what was happening in concentration camps) Stern tasked two historians with verifying the diaries’ authenticity – a process which took almost two years.
During this process, many red flags went ignored. First, no record of the diaries existed – Heidemann had conjured a missing bombshell from the historical record out of thin air. Second, the history book the historians used to check facts in the diaries was the same one Kujau had copied entire passages from. Third, Kujau’s forgery was so sloppy that he inscribed a gothic A as an F — the front of diaries read “FH” instead of “AH.” Fourth, the diaries were printed on paper and bound using glue that didn’t exist until after the war. Nevertheless, the magazine’s team of historians declared the diaries authentic.
Stern wanted to double-check. The diaries were locked in a Swiss bank vault, and Stern flew historians like Oxford’s Hugh Trevor-Roper in to confirm the authenticity of the diaries. Trevor-Roper, a Hitler expert, and several other historians of similar stature proclaimed the diaries to be real. Handwriting experts were also called in — they compared passages from the diaries against documents that had also been created by Kujau. Satisfied and eager to start publishing with syndicates around the world like Paris Match and Newsweek, Stern held the press conference and published the first of the diaries.
It only took a few days of public viewing before the diaries were utterly disproved to be authentic — no less than the German Federal Archives confirmed the diaries were fake. The secrecy the editors had conducted their verifications in (understandably afraid of leaks that would undercut their story and their sales) had made for a very insular and incomplete vetting process.
The hit to Stern’s credibility cost almost as much as the sales they might have made had the diaries been real. Heidemann and Kujau went to jail, two of Stern’s top editors resigned, and the magazine’s credibility plummeted from the level of The New York Times to the level of the National Enquirer, and it took years for the magazine to recover its pre-scandal credibility.
The diaries were examined before digital humanities had taken off as a discipline, and before many of the tools our team is using to analyze our texts were widely available or known. The question of whether stylometrics could have proven the diaries fakes to Stern’s editors while they vetted the material loomed large for us in considering this story. Many tools that are available to us now (not least of which is plagiarism detection software) could have quite easily revealed the diaries as fake to a digital scholar.
The story of the supposed Hitler diaries also reaffirmed the importance to us of having a rigorous methodology to fall back on when evaluating emotional and momentous texts. Had the diaries’ forgery gone undetected, the historical narrative of Nazi Germany and the Holocaust would have been modified significantly. In this way, stylometry can be seen as almost a moral responsibility for historians, journalists, and scholars like ourselves as we undertake research into any text for which authorship attribution has the power to reframe a fraught past.