When Google Gets It Wrong
My wife is reading a crime story I got for Christmas and read over the holidays, Christian von Ditfurth, Mann ohne Makel. It’s sleuth, Josef Maria Stachelmann, is a historian of the Third Reich. Wonderful read, if you know German. Anyway, my wife asked me about the Hossbach Protocol that Stachelmann is supposed to give a talk about. My memory failed me, so I took the easy way out with Google. Bad idea.
The first two hits on Google led to web sites that seek to appear legitimate, but which are in fact sites that deny the Holocaust and consider the Nuremberg Trial a travesty of justice. How did Google mess this up? Have some Nazi would-be academics learned search engine optimization (SEO)? Or was this blind luck? I’m not sure how Google’s search engine works, but the results here certainly point to the limitations of algorithms that rely on the syntactic relevance of a site. Also, while no one is linking to the articles about the Hossbach Protocol directly, there are many links to the main sites on which the articles appear. (You can determine who is linking to a site by typing link:www.name-of-site.com into the Google search box, unless the site is using the nofollow attribute in its links.) In other words, the sites appear to be popular and therefore relevant in Google’s eyes. In fact, Google has blessed both sites with respectable, if not overwhelming page ranks (PR). The first one Historical Revisionism, comes in at a PR 4, and the second one, Institute for Historical Review, at PR 5 on a scale of 0 to 10.
Now I could stop with this warning about the limitations of Google search results, but perhaps there is more to be learned here. Perhaps I should also issue a plea to historians to both learn SEO and write for general audiences on the web. Like it or not, Google is the first place many people turn for answers, and anyone seeking one on the Hossbach Protocol can be easily led astray. Actually, historians might not even need to learn SEO. Wikipedia already has a high page rank and its pages turn up regularly at or near the top of Google search results. Perhaps all that is needed is more and better Wikipedia articles. The Hossbach Protocol doesn’t show up in Wikipedia. If it had, the search results would have been different.
Wikipedia brings up another twist. Typically, when one uses one term in Wikipedia that is more commonly known by another, Wikipedia will at least offer alternative results. (It’s better than Google that way. Google can only offer spelling alternatives.) In this case, though, the more typical American name for this document did not show up in the search results. Only after I typed Hossbach Memorandum did I find what I was looking for. I then typed this term into Google and came up with much more satisfactory results. Only one of the right-wing links came up on the first page, and this time near the bottom.
This final result brings me back to Wikipedia and SEO. We need to enter all possible variations of terms in Wikipedia articles so that they show up in search results. (Sure, I should have entered “Hossbach Memorandum” right from the start, but I translated directly and that was that. As the first set of search results shows, others have done so too.) We also need to do the same thing with web articles and blog posts. It won’t do to leave the field open to the bad guys, simply because the world of SEO isn’t part of our training and does not make or break historical careers. I don’t know if Deborah Lipstadt does any SEO, but her three-year-old blog combats holocaust denial and has a PR 6. More established historians need to follow her example in their respective fields.
Stoneman’s Corner
- Book Review: GDR and Consumption January 13, 2013
- Teaching Notes: Synthesis and Process September 29, 2012
- Hist 499-005: Gender and Class in Modern Europe August 18, 2012
- Reflections after Class July 3, 2012
- Requiring Students to Use Chicago Style (or Turabian or Whatever) June 14, 2012
- Military History Conference May 13, 2012
- Syllabus for Hist 606: Themes in European History II May 3, 2012
- Command Culture by Jörg Muth April 28, 2012
- Who Should Groener’s Schlieffen Plan Matter To? March 18, 2012
- Hist 499-003: Gender and Class in Modern Europe (Required Books) March 7, 2012
I can only wish your comments reach a wide audience. I had a student that did a search for an easy explanation of Darwin’s work and she came back with something that sounded strange, so I checked. Turned out the guy had a PhD from a religious university; his diss – from which she got her info on Darwinism – argued that dinosaurs and humans co-existed, and thus disproved the extinction of the dinosaurs at 65 MYA as well as human evolution. Yeah, they gave him a PhD for that. Just makes you feel warm and fuzzy, eh?
It is a very important case to think about the ways of gaining historical knowledge on the internet. The key role of promotion (SEO, but not only) is a fact – without this, credible and substantive knowledge can not be used. It is the same situation when there is a popular “historical” movie with non-historical content. It can be even more dangerous (by producting or promoting historical myths) than hundred of underground books, which are being read by nobody.When learning and finding knowledge about history is based on independent searching through the directory of not checked resources, in which the most credible is the most popular (Page Rank system) the relation of history and promotion should be a part of preparing the internet presentation of historical project. The second case is how to block the dangerous materials and ideas – it is obvious that they can be removed truly, but should become a margin – not, as you have described – a first source of knowledge. Can we remove this effect by hacking it? For example, the Hitler’s Mein Kampf is widely popular on the internet (among Poles too, which for me is mostly an effect curiosity than element of sympathy for neonazi ideology) – it can be easy found in the P2P networks, mailing lists etc in the PDF format. What can we do to prevent this? None. We can only hack its meaning and promotion, use its own ways of communication to educate about everythign which was started by this book. This comments can be added to the historical content of the book and put freely on the intrenet. One who is searching for Mein Kampf (because of the curiosity) now could find this book with good historical comments and suggestions – no with only a few words about history of this source, how it is often in the underground editions.And – the last question – is about search mechanisms as Wikia (http://www.dw-world.de/dw/article/0,2144,3048800,00.html). When people more than script decide which resources is more important and credible, how we can prevent promoting historical myths? They are sometimes very popular and when search engine is based on the democratic system, it is a strong chance they will be promoted further.
I would never advocate hacking a questionable site. There is supposed to be such a thing as free speech, after all. Rather than block such sites, I would rather see us get better at communicating the truth.
It’s certainly not the place of search engines to judge historical truth, and denier sites are likely to be of interest to anyone whose object of study is holocaust denial rather than the holocaust itself. So I think Mark’s advice that the good stuff needs to be better optimized is spot on.It would be easy for me to get all Nietzschean and say that people should be suspicious of everything they read and make up their own minds about whether to accept it as true, but in practice they don’t. It can be very difficult to spot the dodgy stuff. Even the Cliopatria blogroll once mistakenly linked to a holocaust denier blog, and when I pointed out their mistake they went the other way and also deleted a legitimate history blog set up to expose deniers.I mentioned in this post that it often seems safer to not talk about the holocaust in case you inadvertently say something that helps the deniers, but not saying anything is worse, especially if it leaves search engine results full of deniers.
There’s certainly something to be said for academics learning SEO. People wanting to gain attention for for their business– or in this case, recruit/spread the word about their cause and their skewed version of history– are so much more likely to feel the need to truly WORK a system to gain attention and marketshare (or recruits to the cause). True academia tends not to think along those lines; credible publication itself is the reward and the means to the end. It tends to stop there.
@ms I mean not hacking a concrete site but hacking a message, hacking a way of distribution of such materials. It is a question if we can use this communication but change the content, re-edit it for the educational or moral aims.
Its that kind of blind output from Google Searches that made me a fan of Thomson Scientific’s WebPlus search engine. While you were away over the holidays I posted a walk-through for using that search engine at WebPlus Beta Comes Online. That post is currently the most popular post on my blog.I repeated your search there and got different results. At that point, two clicks took me to the results filtered down to just those from academic domains.Google is great when I need to find the cheapest 17″ LCD monitor or movie times, but for any real research, I turn to WebPlus.
@ m-w: I’m trying to imagine such a scenario and I’m coming up short. Are you proposing some kind of set-up that acknowledges and links to such sites, but then offers critical commentary? I personally would like to ignore them and see to it that better results come up on Google. But perhaps you have something else in mind?@ techfun: I responded to your Thompson piece on the thread about this post on BlogCatalog. Other readers will find some interesting comments there too.
@ms I can describe it using the case of Mein Kampf, which is widely accessible on the Net, but very often without any historical comment. There is no chance to block distribution of this book on the internet, so maybe it is better to prepare critical educational brochure in PDF about it (with of course elements of the source text) and put on the web, P2P (with the same name of the file), so internet users can reach it instead of non-critical versions. Do you know google bomb idea? It could be a very similar action. One is searching and expecting something, but gets it with some modification.
Of course, we need Mein Kampf available for scholarly work. But the best way to combat Holocaust denial or indeed any other pseudohistory is to:a.) make sure that real history and real controversies (as opposed to contrived controversies) are disseminated to the public, b.) that they are disseminated well through sources like wikipedia; and c.) that we constantly expose the techniques by which Holocaust deniers and their ilk formulate and distribute their fabrications.
Falsehoods, misrepresentation and misinformation can be found all over the internet; your experience is a case in point. I do think historians and scholars should be educated in SEO, if they use the internet to disseminate information. I find it amusing, troubling, and absurd that people blindly accept information on the internet without carefully researching sources.There are those who are intent upon rewriting history and they are using this platform to do so.
The problem for me is that not enough scholars are using the internet to disseminate information, mainly because the professional rewards for doing so are nil. In fact, using new media might even carry a professional penalty in some circles.
Right. A scholar on a mission to explain his or her field to the non-academic audience is treated with suspicion.,
Google has indeed lured us too many times to banal and stupid web pages and blogs of SEO-savvy idiots. It’s tiring. When Google fails, try Yahoo.
For specialized needs, try specialized solutions. Thomson is probably a good idea too, for certain types of search.
I don’t see how Yahoo could be an adequate substitute. Thompson is pretty limited too. The point is less about Google per se and more about how we use Google.
Of course, you are speaking to an educated audience here. The fact is sadly, that most people have no interest in history at all. Those who do look information up on Google deserve applause. (If for nothing other than the attempt.)
Rereading this article, I realize that I have set up a strange loop here, since I linked “Hossbach Memorandum” to Wikipedia, and anything can happen on that site, especially with terms that tie into such emotionally charged subjects as this one.
I’m far too late in responding to Don’s comment above. I suppose looking something up on Google is progress for some people, but I’ve seen quite the opposite of progress in my teaching experience. Even we academics are guilty, as my Wikipedia link and the original motivation for this article show.