On November 28th, 2014, Moltar, a writer at the website GamerGate.me, a site dedicated to publishing GamerGate related news and editorials published an article claiming that a top level faculty member of Harvard University had considered writing an article for the site. On December 7th, 2014, that article was published. Many people on various websites, and in the comments for that article rightfully questioned its veracity. In this article, I intend to go through and look at the method it uses to keep the author’s true identity anonymous, and what that method means.
The need for using a method of anonymization for the source is well founded. Several people that have dared to get close to either side of GamerGate have been tracked down, and have lost their jobs. While professors usually have tenure, many employees in the United States are hired at-will. So while it isn’t a breach of tenure to write about GamerGate, it is well founded to have a concern for ones family or friends being targeted. I consider it to be a valid level of protection to take so as to not get caught up in the entire mess any further than need be. The method chosen works well for it, though many people that read the essay did not understand how exactly it works. In that, the author was either unheard, or misunderstood by a significant portion of people that read the essay.
The method used is similar to something called the Barium Meal Test better known as the Canary Trap, popularized by author Tom Clancy in his novel Patriot Games. The method in which it works is by selecting a number of words throughout the document, and replacing them with synonyms. This works to keep the overall meaning of the document intact, but makes it much harder to derive the original document, and thus the original author’s writing style. This is important in order to hinder writing analysis to essentially find the writer’s ‘fingerprint’ in their style and word choice.
To use a much simpler and straightforward example, consider the following text.
The quick brown fox jumps over the lazy dog.
Using related words, we can come up with alternatives.
The fast beige fox hurdles over the lazy hound.
An alternative could be,
The nimble tan fox hops over the lazy canine.
The effect is that the original message remains intact, but the writing style and wording is different. In a traditional counterintelligence role, seeing which version becomes public would allow those with the original document to know who leaked it, by virtue of comparing the different wording to who had the copy with that exact version. Usually over two or three paragraphs, there can be thousands of permutations. The author of The Harvard Essay had the luxury of using his entire article to do essentially that, in order to cloak his identity.
What will make it significantly more difficult for those that want to figure out who he is is that his command of the English language is such that you cannot be entirely certain which words have been replaced. Assuming that the author has sufficiently covered their tracks and has left no other traces, they have somewhere around 2,700 words to choose from. If you were to do the math on how many different synonyms could be used for each of those 2700 words, I’m pretty sure it would return a number that is large enough to sufficiently confound anyone trying to do writing analysis on the overall document.
In conclusion, it is my opinion that while many of the people that read it got angry for not coming across in plain language, you misunderstood why it was written the way it was. Yes, you should doubt and question everything, but you should also make an effort to read the entire document, and do research on the references made. I read it, and from what I can glean, it is legitimate and makes a number of solid and entirely valid points. It is a solid criticism of Anita Sarkeesian as a scholar. It also serves at least in partial as a criticism of GamerGate as a movement, both of which should be heeded. You owe it to yourself to read the entire thing if you haven’t yet. As far as what the means used to cloak the writer’s identity mean, I would say that it was a good idea, but could have been a mistake given how many people failed to see the forest for the tress. Personally, the fact that they claim to be from Harvard really is beside the point. It would not shock me if they were, either.
Is there anything I missed?
Author’s Edit 6:15 pm CST, Jan 3 2015:
Regarding the confusion around the Barium Meal Test example that I used, I should explain that its more about the message being conveyed and less on the wording. In this case, being that something (the fox) is overcoming a barrier or bypassing it in some way – what the barrier is doesn’t really matter. You can theoretically make the barrier a brick wall, and it would still work to convey the message that something is jumping over or running around something else.
The descriptors are there for a few reasons, mostly maths on how many permutations it can widen to, and thus how many people this kind of thing will work against in a traditional use.
The fact that colors and speeds or anything else are attributed to them, likewise, serve the goal of increasing the number of potential different wordings of the same core message. In practice, you can change entire sentences to use entirely different wording and verbage and still retain the core message, which is the point in the article’s entire analysis.