Sunday, November 29, 2009

It's the End of the World: Crowdsourcing in Literature Actually Works!

Ok, a little context before my comments.

On September 15, Dan Brown's latest book, The Lost Symbol, was launched with much fanfare in the United States and around the world. Because of the secrecy involved with the contents of the book, no advance copies were released to foreign publishers so that they could have it translated and launched at the same time in their markets.

As reported here, Swedish publishers decided to assign the job to multiple translators in order to limit piracy and to prevent impatient fans from buying the English version of the book, by expediting the publishing of the Swedish translation.

Well... they did it. On October 21, 2009 — only 36 days after the launch of the English version of the book — Albert Bonniers Förlag released the book in Swedish. In that period, they were able to translate, edit, format, print, and distribute 300,000 copies of a 614-page book.

And how did they do it? This article in Swedish (I read it using Google Translate) narrates the details of the adventure.  But for our purposes, what matters is that seven translators worked on this project. Their names are Leo Andersson, Tove Janson Borglund, Ola Klingberg, Lennart Olofsson, Peter Samuelsson, Gösta Svenn, Helena Sjöstrand Sven. From what I could see in AdLibris, the Swedish online bookstore, all of them are very experienced translators.

As one review says: "Another positive aspect: the translation is actually quite okay. Even here, I have put a sadly, because it would have been preferable if the insane circumstances surrounding the translation into Swedish - seven translators, a few mere weeks - had left its mark in the text."

What do I think about this? I think that this must have been a very exciting project, as it epitomizes the power of collaboration.

The publisher needed to have the book out fast (I saw the English version of the book exhibited very prominently at the Stockholm airport both times I was there before the launch of the Swedish version) in order not to lose 150,000 sales as the publisher of Harry Potter did because of delayed translations. Time-to-market was the critical element in protecting its investment and maximizing its return.

And before you say the Q word, I actually believe that several translators working together might deliver better quality than one working alone.

Special thanks for my friend Anne-Marie Colliander Lind, from Common Sense Advisory, who helped me collect some of the data for this posting.


  1. Uhm, As I hear from colleagues in the publishing industry, this happens almost every time a book is translated into Italian. This also allows Italian publishing industry to offer the tightest timelines and lowest rates on the market. As for the quality of translations... crouwdsourcing doesn't always mean good translations, especially with some languages and tight timelines, due to a constant lack of good proofreaders...

  2. Ummm. They picked experienced, fully identifiable translators to work together as a team, as opposed to letting anyone anonymously chip in if they wanted. That's not crowdsourcing, IMO; it's project management (I hope!). Otherwise, we'd be saying that most if not all serious LSPs "crowdsource" at one time or another. Hardly fair.

    IMO, crowdsourcing is what happened with the Harry Potter books when they came out in English only and the fans took it upon themselves to translate them: first and foremost, no selection process of translators. Anyone could chip in. No proofreading nor quality check afterwards, either.

  3. The Da Vinci Code is actually closer to a technical manual than literature. The reason is that in both texts, there is no such thing as style or nuances, content is everything--how that content is transmitted: doesn't matter (stylistically speaking). The writing, and therefore the translation, could be split and done by several people. That's not the case with literature, where the text is more 'dense' and the content is actually crafted and not necessarily the most important thing. A consistency in the style is absolutely necessary, even if the translation is done by several people. That can be achieved in a collaborative translation (with a small team of translators), but not with crowdsourcing and certainly not against the clock. Granted, more and more writers are now more like Dan Brown than like Hemingway or Dos Passos.

  4. @ Claudia:

    I was thinking comething along the same lines.

    There's a big difference between works that have "what happens" as main selling point/appeal, and books where the rendering of what happens is as or more important than the events themselves... not to mention, the target audiences of these are vastly different from one another.

  5. I'm with Andrea on this one. If I read correctly, we have 7 experienced translators who are not working for free. That's not a crowd and that's hardly crowdsourcing.

    Regarding the volume: 614 pages, 600 words per page (my guess) done in 25 days (my guess again, but it leaves 11 days for proofing and publishing).

    Daily wordcount: 614*600/7/25 = 2105 words/day. That's not even a rush job for the individual translators.

    Please correct my figures if you have a copy of the English book.

    Kind regards,


  6. Anonymous3:25 PM

    You seem to be totally mistaken about the meaning of the word "crowdsourcing", which is very surprising and disappointing given that -according to the info under your picture- you are the CEO of a translation agency. We are talking about a team of translators. The fact that they worked under an unreasonable deadline doesn't make their word method "crowdsourcing". I often work with another 5-6 translators on large projects when the deadline is tight. Is that crowdsourcing too? Absolutely not. Perhaps it would help if you looked up the definition of crowdsourcing. Try googling it. Or read the ATA chronicle, there was a good article about crowdsourcing a couple of months ago. Seriously, read it before publishing such rubbish on the internet.

  7. This kind of shared translation has been done for ages in the IT localization world in the same circumstances, i.e. when there is a very tight deadline for the release of a product and a huge amount of words need to be translated. As other readers said, this is project management, not crowdsourcing. And there must be some kind of proofreading to help unify the style: otherwise the quality of the final product, and particularly for literature even if it is for Dan Brown's standards, which are not very demanding to say the least, would be severely flawed.

  8. Greetings

    It seems fair to say that it was not crowdsourcing at all. I think we need to be extremely careful with the terminology used. The translators were chosen and are qualified translators. I also assume that the translators were paid for this job.

    Also, I just about to finish the English version, The version I have has 509 pages. I did a count on a page which is around 425 words, so the English version should be around 220000 words, so It seems fair to say that with 7 translators, it would have taken 15 to complete the translation, leaving 21 days for proofing and manufacturing.

    Kind regards


  9. Anonymous1:32 PM

    Another interesting aspect of the review you refer to above is that she thought it was rather boring and that language isn't the most important thing in a Dan Brown book.