Against original research

[Cross-posted at Cliopatria.]

I know. Writing about Wikipedia is so 2006. And yes, finding errors in Wikipedia articles is not exactly difficult. But I have a bee in my bonnet which needs releasing.

Wikipedia's page on the Blitz has a section entitled 'Commencement on September 6'. This is how it currently reads (sans hyperlinks and superscripts):

There is a misconception that the Blitz started on September 7, 1940. Bombs began dropping the night of September 6 and continued for the full day of the 7th and on into the morning of the 8th. Saturday 7th was the first full day and has officially and erroneously become known as the day the Blitz started. Hermann Göring launched bombers and the first bombs caused damage the night of September 6.

Quoted in the The Manchester Guardian is Göring's communiqué:

Attacks of our Air Force on objectives of special military and economical value in London, which began during the night of September 6, were continued during the day and night of September 7 with exceptionally strong forces using bombs of the heaviest caliber.

A witness recalled the evening of Friday September 6, 1940:

My name is John Davey. I was born on December 27th 1924 in South Moltom [sic - Molton] Road, Custom House, West Ham, and a couple of miles from the Royal Docks. In September 1940, on the Friday evening of the weekend the docks were first blitzed, I was sitting with my friend in his house. At about 7 p.m. there was a series of explosions and the shattering of glass. We ran into the road and saw at the end a flame that shot into the sky, seeming to light up the whole area. My friend and I and lots of others ran towards the fire.
—BBC, WW2 People's War

The first damage to property on September 7 was recorded at eight minutes past midnight, a grocer’s shop at 43 Southwark Park Road, SE16.

It has long been the accepted, but erroneous, view that the London Blitz lasted 57 consecutive nights starting on September 7 1940 and ending November 1. In actuality September 6 makes 57 nights and not September 7. The historian AJP Taylor wrote of such an error:

… it is the fault of previous legends which have been repeated by historians without examination. These legends have a long life.

This is really quite silly. Yes, it's true that the accepted date of 7 September 1940 as the start of the London Blitz is a bit misleading, since there was a non-trivial amount of bombing before that date (e.g. see here). Judging from contemporary press accounts, 7 September certainly seemed to mark an important change in German bombing strategy, but more one of quantity than quality -- almost more an inflection point than a turning point. In retrospect we tend not to see it that way, which is fine. But we could recognise that -- leaving aside the eventual reification involved in the name 'the Blitz' itself -- the 'start of the Blitz' was less clearly defined then than it seems now.

But this is not what the Wikipedia article is talking about. Instead it chooses an equally precise date for the start of the Blitz, 6 September, and says that this is more accurate than 7 September. Somehow, it seems, every historian since 1940 and every witness who wrote about it at the time or later has somehow forgotten that 'Bombs began dropping the night of September 6 and continued for the full day of the 7th and on into the morning of the 8th'. The citations for this are just two. One is from the Manchester Guardian (9 September 1940, 2), a reprint of a Luftwaffe communiqué vaguely claiming attacks on London beginning on the night of 6 September and 'continued during the day and night of September 7'. This is a source which should be used with caution: did the Guardian quote the communiqué accurately? Did the Luftwaffe communiqué tell the truth? What sort of 'attacks' were carried out on the night of 6 September, lone raiders or formation raids? Then there is an account from a Blitz eyewitness taken from the BBC's WW2 People's War site. He says that the first raids on the Royal Docks took place on the Friday evening of that weekend, that is 6 September. Well, he was living in West Ham at the time, so he would know, wouldn't he? But this is account written down over sixty years later. Whether it was Friday night or Saturday night seems like something which one could be mistaken about after all that time. And doesn't this one account need to be balanced against others?

The only other evidence given is that the first bomb damage recorded in London on 7 September occurred shortly after midnight. This fits in with the narrative here of continue bombing throughout 7 September. But was it continuous? Only on a naive reading. The source given for this is an excellent spreadsheet published by the Guardian. Note that it only gives data for the one day, so we can't compare it to a typical pre-Blitz night, or what happened on the supposed first day of the Blitz, 6 September. But even so it shows that the raid (which is certainly known to historians) in the early hours of 7 September was only moderate at best. Around 50 bombs were recorded up to around 2.30am, with only another 20 or so falling in the next twelve hours (with gaps of up to two hours between them), and about 15 in the two and a half hours after that. That takes us to just before 5pm which is when the bombing really starts to escalate: there are nearly 760 bombs recorded in the 7 hours until midnight (and it didn't stop then, either: that's just when the Guardian's spreadsheet does). So if you are going to start counting individual bombfalls, there was definitely a big quantitative change on the evening of 7 September.

The icing on this cake is the totally pointless quote from A .J. P. Taylor's The Origins of the Second World War. Yes, historians do sometimes repeat 'previous legends [...] without examination'! So what? This is not evidence that it has happened here. And note the artful use of this quote: Taylor has written 'of such an error'. A careless reader might think that Taylor is talking about this 'error', when of course he's not (he's on about pre-war diplomacy).

And then there's the smug, self-congratulatory tone of both the article and its Talk page (current revision):

Isn’t it rather exciting to be in on the correction of a widespread, albeit small, but important corner of recorded history?

I swore I would never again but the Sept 7 inaccuracy was too important to leave uncorrected; it did my ego some good too.

Once again an historical inaccuracy, this time an error in arithmetic, has been perpetuated for decades.1

The thing is, this is not the Wikipedia way. One of the cardinal rules of Wikipedia is No original research:

Wikipedia does not publish original research. The term "original research" refers to material—such as facts, allegations, ideas, and stories—not already published by reliable sources. It also refers to any analysis or synthesis of published material to advance a position not advanced by the sources.

The Blitz article -- at least this section of it, which has been there for nearly three months now -- fails to uphold this principle. The idea that the Blitz began on 6 September seems to have appeared out of thin air: it is not already published by reliable sources (whether they be right or wrong). Nobody says this, and even if they did, it would be wrong in this case to choose them over everybody else who chooses 7 September. There's nothing wrong with original research -- I'm quite fond of it myself -- but it's not what an encyclopedia is for. And if you're going to do it, do it right.

Of course, this being Wikipedia I should just roll up my sleeves and go to work editing the article myself. That's also the Wikipedia way. But I've tried that before and come up against stubborn editors who refuse to let go of their misconceptions. Wikipedia has its own processes for judging original research disputes. If I had much faith in the system I'd use it. And this is only the most egregious example (at least to my mind). So instead I'm going to post this here and hope that the 'Romanian plane' effect will work to my advantage.


  1. It seems another 'source' for this insistence on 6 September as the start of the Blitz is a paper by Peter Stansky, author of a book called The First Day of the Blitz. But nowhere does Stansky say the Blitz started on 6 September -- he actually says 'The Germans selected September 7, "Black Saturday," to start the Blitz' -- instead, judging from the Talk page the Wikipedians seem to be counting backwards from Stansky saying the Blitz lasted 57 nights. Or something -- the numbers don't match what Stansky says and at this point I've given up trying to understand where this whole idea came from. 

Creative Commons License
This work, unless otherwise expressly stated, is licensed under a Creative Commons Attribution-Noncommercial-Share Alike 3.0 License. Terms and conditions beyond the scope of this license may be available at airminded.org.

18 thoughts on “Against original research

  1. This is something of an endemic problem on Wikipedia - people finding something generally stated that doesn't quite seem to be the case, deciding to correct it, and banging on about having done so at great length. Which isn't so bad when they're right, but as you've noticed can be exceptionally tedious when they get the wrong end of the stick...

    Looking at the discussion, it does seem to be very much one person's pet topic that no-one else felt up to arguing over. I'll try and rework it into something more reasonable tonight...

  2. ...well, one mystery cleared up; Stansky was originally cited to support the 7th and they just changed the number whilst leaving the cite in place!

    I've removed the section (for now, at least) and I'll try to add something a bit more reasonable about the dates in the near future.

  3. Ah good old Wikipedia. I stopped reading them for detail a long time ago.

    However, the RAF's AHB Narrative, written by T C G James and still considered the work on the Battle, cites 7 September as the important date.

    It says that:

    'The German operation on this day have generally and rightly been regarded as the beginning of the attack on London. The reasons are obvious enough. Not only were were virtually all operations against the United Kingdom concentrated against the capital...but they were followed up by the largest concentrated night attack the Germans had thus far launched...'

    On the other hand it describes the night attacks, which had been occuring from 25/26 August when Birmingham and Coventry were attacked, as a relative failure based on the forces involved.

  4. I find the cautions one must apply to Wikipedia are also most useful when applied to historians and Official Histories. :)

    Mildly more seriously, after reading Brett's well argued case, it occurs to me that Wiki provides a service to people interested in (or learning about) history in that it shows history is mutable and debated, and that one can make a contribution without being an authority from an institution. That's as well as showing the wide variety of people and 'realities' on the internet, incidentally good training for the wide variety of people in academia and their realities.

    Obviously that ducks the 'original research' issue on Wiki specifically, but it might direct people to realising that any source (including those blasted Official Histories) has to be one interpretation, not necessarily a definitive one, even if literally carved in stone.

    And the numeric obsession of humans is literally unnatural. If we didn't use numbers as absolutes the way we do, it would be obvious that the difference between the 6th and 7th is of no great significance, any more than stating the first day of spring must occur on on a particular calendar day.

  5. The problem (well, a problem) with Wikipedia is that everyone wants to get in their two penn'orth on these articles about big historical topics, and they usually do this by splitting hairs over something that isn't terribly important in the first place. So someone adds a section to the Winston Churchill article saying that he liked broccoli. Then someone else edits it saying no, he liked carrots. Then the first editor defends the broccoli-loving WSC interpretation at twice the original length. Soon, the article is devoting more space to Churchill's vegetable tastes than it does to his role as Prime Minister; and no-one in the escalating pissing contest ever pauses to think whether it really matters all that much whether he liked broccoli or carrots anyway.

  6. An excellent bit of work, Brett. I entirely agree with you about the difficulties of tackling Wikipedia - or more precisely Wikipedians - and have myself sworn off writing or correcting anything the site publishes in those areas I actually know something about; it just gets too time consuming and frustrating.

    For what it's worth, I did, however, have some success in removing another increasingly pernicious myth from the site: the notion that Gavrilo Princip, assassin of Archduke Franz Ferdinand, was able to commit the crime that plunged the world into the Great War because he had stopped at a delicatessen to eat a sandwich just as the Archduke's car pulled up right in front of him.

    The one great advantage of WP is that it is possible to address this sort of thing directly, so long as you're willing, at least in theory, to engage in an edit war to keep your version up for the rest of your life. This particular myth - and it is a myth; there's no mention of it in the trial transcripts and Bosnians did not even eat sandwiches in 1914 - seems to spring from an incautiously dramatised BBC Schools documentary, and I've had a lot less success (well, none, actually) in getting the Beeb to address it.

    Like the problem of the date on which the Blitz "started," the tale of Gavrilo Princip's sandwich is completely unimportant in itself, yet a worrying example of the way myth becomes history if it circulates widely enough.

    http://en.wikipedia.org/wiki/Talk%3AGavrilo_Princip#Gavrilo_Princip.27s_sandwich

  7. EyeSerene

    Thank you Brett for picking up this issue. The disputed section has been removed from the article pending resolution. You're absolutely right that Wikipedia prohibits so-called "original research" (writing based on personal knowledge that isn't backed up by a reputable source), and it does seem at first glance that may be what's happened here. Fortunately it's easily settled by answering the question "what do the preponderance of reputable sources say?" The article must reflect that, although a brief mention that a significant amount of bombing took place before this date may also be worth including. Contradictory sources are a perennial problem, but that's where subject expertise such as yours is very helpful in deciding the proper weight to give to information.

    Unfortunately the open nature of Wikipedia means that such things do happen and can slip under the radar, but as you also point out they can be corrected. It's a shame that in trying to contribute you seem to have run across difficult editors, but most users aren't crusaders and there are check-and-balance mechanisms to deal with partisan and tendentious editing. Of course, to access these mechanisms one has to know they exist and how to use them... a constant problem for new and inexperienced users. Many fall into the trap mentioned by Mike Dash of edit-warring (making repeated tit-for-tat changes), which is specifically forbidden and results in editors being blocked and articles locked. The best place to get more eyes on problematic military-related content is to post at the talk page of the Military history WikiProject (disclosure: I'm one of the editors involved in helping out there).

  8. Post author

    It seems that the squeaky wheel does indeed get the grease! I can't say I'm glad to see that my experience is familiar to other historians. There have been other attempts to come up with alternatives which are more 'expert' friendly, like Knol and Citizendium. But they're nothing like as pervasive or comprehensive as Wikipedia, and Knol goes too far the other way in letting the experts ride their own hobby horses. (Here's Citizendium on the Blitz, for comparison.) Overall Wikipedia is surprisingly good, but it has to be used with caution.

    And good catch on Princip's sandwich, Mike!

    Andrew and EyeSerene:

    Thank you for swooping in so quickly to fix the Blitz article: it looks much better now. And I appreciate an editors' point of view -- I don't think everyone is cut out for the job! I wonder what you think of my strategy here? I thought that writing a blog post gave me enough space to make my case at length, whereas I might not be able to do that on a Talk page. Or is that not really necessary, in general? Should I have just made the edits, briefly noted my reasons on the Talk page and then waited to see what happened?

  9. EyeSerene

    Glad to be of help (though Andrew did all the work). Despite some - I hope comparatively rare - appearances to the contrary, many of us take Wikipedia's credibility very seriously indeed and are grateful for any assistance in raising article quality.

    Your strategy is hard to fault - after all, it's worked! However, if you'd decided to approach this on Wikipedia the best way would be as you've suggested - to make the changes, note the reasons on the talk page, and wait. There's no restriction on the length of talk-page posts as long as they're relevant to improving the article, though with one (cynical) eye on third parties getting involved if dispute resolution becomes necessary, it can help to bear tl;dr in mind.

    The major potential drawback of the above method is that you do end up in a one-on-one with another editor who appears to be guarding their preferred content. This is the point where many new contributors make a perfectly understandable decision that life's too short and leave in disgust. It always saddens me when that happens because we've lost another potentially good editor and, if only we can get the message out, there are many ways to deal with things long before they go out of control (from asking for third opinions to contacting an administrator to opening a general request for comments to various sourcing/neutrality/user conduct noticeboards to getting the parent WikiProject involved to... the list goes on!) The issue you've raised with Blitz article is, thankfully, an easy one to resolve simply by applying site policy ("no original research" in this case).

    Incidentally, on the wider issue of academia/experts and Wikipedia, which isn't an easy relationship, you might be interested in this essay on a university/Wikipedia collaboration we ran a couple of years ago - while not directly relevant to this issue I think jbmurray sums up well some of the issues we face. My follow-up here may also be of interest.

  10. I take a certain perverse pleasure in checking out the "Discussion" page of the more romantic and colourful Wikipedia pages. It can be almost like flipping over a rotten log in the forest....
    Which is kind of an unfair analogy, since the monomaniacs waging endless war for their theories that "Etruscan really is so an Anatolian language," or "ancient astronauts really did build Tiwanaku as a port city on the Pacific " are probably Asperger's victims finding a social outlet in edit wars on Wikipedia. I'm also singling out examples where a particularly persistent crank has ended up expunged.
    Which is what makes it perverse. As frustrating as it is to see a useful article hijacked by someone's crazy theories, its eventual return to anodyne usefulness implies that some person who really needs some kind of human contact has been re-isolated. And, of course, I am now projecting exactly the wrong kind of shallowly affective pity that makes the lives of the genuinely mentally ill just a little worse than it already is.

    So what I'm saying is that some heroic therapist should reach out to all these people and help them with their problems. I offer my emotional support, and would be prepared to donate , oh, say, 10 bucks CAD towards the effort so that I don't have to feel too guilty.

  11. Chris Williams

    Allan, please may I quote your "Winston and carrots" view of wikipedia to several thousand students? Attributed of course.

  12. Post author

    EyeSerene:

    Thanks for those links. I do like the idea of getting students to contribute to Wikipedia; it would be a good way of teaching first years things like research skills, how to use sources and how to avoid plagiarism.

    Erik:

    I must admit I enjoy reading a good Talk page stoush. Or even a bad one.

  13. Simon Harley

    Wikipedia has its good and bad points. One of the worst things about it is actually it can be a little bit too similar to academia. Take for example its Featured Article procedure, supposedly the epitome of Wikipedia peer reviews. A couple of years ago an article about the winner of the Victoria Cross was promoted. Ostensibly, it was well-referenced, based on a number of Victoria Cross books. The only problem was that all these "popular history" sources completely misrepresented the facts. One had to go into slightly more specialist (in this case naval history) books to get all the relevant information, which cast an entirely different light on the article (i.e. the Victoria Cross probably wasn't deserved). Unfortunately, none of this was addressed at the Featured Article nomination, which effectively just gave the article a spell-check and rubber stamped it. Alas, it slipped under my radar at the time, and I got some peeved comments when I started updating the article based on the available evidence.

    I've read enough doctoral theses in my line of expertise to know that Ph.D. students get away with writing absolute crap in their work, and they can get away with it because no-one else either knows what they're talking about or they simply won't call them on it (presumably the former). At least with Wikipedia rectifying errors of fact is a simple matter of pulling out a source, typing a few words and in an instant its available to the world.

    At least until someone decides to revert it for whatever reason ...

  14. Simon Harley

    *"a winner of the Victoria Cross", I ought to have written. Definite articles on the brain ...

  15. I've been corrected by my military friends before now that such medals are 'awarded' not 'won'. Checking the background, I also learn that strictly speaking, but rarely observed, the VC is a 'military decoration' rather than a medal. Most interesting.

  16. Post author

    Simon:

    I've read enough doctoral theses in my line of expertise to know that Ph.D. students get away with writing absolute crap in their work, and they can get away with it because no-one else either knows what they're talking about or they simply won't call them on it (presumably the former).

    Yes, I would hope it's the former...

    At least with Wikipedia rectifying errors of fact is a simple matter of pulling out a source, typing a few words and in an instant its available to the world.

    ... though by the same token most of the rubbish written in PhD theses never gets read by more than about 3 people, so is in that sense is less toxic :)

  17. Pingback:

Leave a Reply

Your email address will not be published. Required fields are marked *

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>