NookReview 012

Update: Diane Duane has written an interesting piece in response, that I will cover in full when I have time.

I was going to bring this up in my review of the Nook Reader app, but realized that doing so would be putting the blame in the wrong place.

When I was reading the Young Wizards series by Diane Duane on the Nook Reader, I ran into this particularly egregious typo in the first chapter of So You Want to Be a Wizard. “The reader is invited to examine the next Jew chapters…”

And it was not the last. Characters might enjoy meals of “arroz con polio”, or instead of arch-adversary the Lone Power go up against “the Tone Power”. Occasionally entire words would simply be missing.

Up to this point, I hadn’t bought a lot of e-books from either Amazon or Barnes & Noble. Apart from DRM-free Baen, my commercial purchases had been from eReader and Fictionwise, and I never discovered a typo in any of those books. I had thought that some of the other complaints about e-book quality I’d heard elsewhere (including Joanna’s famous rant about hurting authors back in the Amazon-vs.-Macmillan days, or her more recent examination of four Kobo books for errors) might have been exaggerated. But now I buy my first complete series from Barnes & Noble and discover multiple errors in every book.

And it’s not B&N’s fault either. I downloaded the sample chapters of So You Want to Be a Wizard from Amazon and found the “Jew” typo there as well. The error was not introduced by the e-book vendor, but was present in the e-book file provided by the publisher. In fact, as I understand it, e-book stores can’t change anything in the e-books they sell without the publisher’s consent—so even if they wanted to fix the typos themselves, they couldn’t without the publisher’s explicit say-so.

Why are publishers providing these flawed, faulty scans? It doesn’t make sense. The whole point of a publisher is to be able to take an author’s book and polish it so there aren’t any mistakes, because selling a typo-riddled paper book would reflect poorly on both author and publisher. To that end, publishers employ hundreds of people in structures organized for the sole purpose of scrutinizing submitted manuscripts for mistakes—and writers get proofs and advance copies to eliminate errors in printed books before they get printed.

Why on earth are publishers not sending their e-books back through those editing structures on the way out? If they don’t have enough people, they could hire more. Even running an e-book past one person would be an improvement, because it’s hard to believe that this kind of error could go uncaught if scanned by living eyes.

I don’t have any way of knowing what percentage of e-books actually have this kind of error. These, and all the others that people complain about, might just be anecdotal evidence and the vast majority of e-books, that people don’t complain about, might still be all right. But looking at it another way, any typo-filled e-book is too many. We don’t tolerate any typos in printed books, so why should we tolerate them in e-books?

Are they just not taking the e-book market seriously yet because it’s still small compared to the paper market? I suppose that makes a kind of sense. I’m not going to fall into the conspiracy-theory trap of assuming publishers are trying to harm the e-book market to promote the print market, because that doesn’t make sense—and it also makes more sense not to ascribe to malice what can be adequately explained by incompetence.

But publishers really ought to get their act together. Releasing e-books full of typos makes them look bad both to authors and consumers, and publishers depend on the one for their source material and the other for their money. Authors might decide to take their books elsewhere if the publishers prove incapable of doing a good job with them—including publishing them themselves on Amazon or Barnes & Noble where they can keep more of the royalties.

And consumers might decide to stop buying poor-quality books. Or they might decide to go to peer-to-peer where the books will not only be free, but they’ll at least sometimes be higher quality too. Since many of those who contribute e-book scans to peer-to-peer do care about how efforts associated with them look, they’ll take the time to edit.

The sad thing is that the current infrastructure of Internet-connected readers would make it possible to get these errors fixed quickly and cheaply, if publishers and e-booksellers could be sold on the idea. After all, they’re already sending data back upstream if only to let the Amazon servers keep track of where we left off reading. Why not take a cue from Distributed Proofing and Wikipedia and let us submit corrections?

I know that if it didn’t take too much effort, I’d be happy to correct errors I came across for free, and I’ll bet a lot of other people would, too. If it were a simple matter of just tapping a couple of keys to change that capital “J” into a lower-case “f” and click “submit”, why wouldn’t I want to?

There could be someone at the e-book store or publisher whose job would be to review the corrections, and perhaps he would also be able to call up the original scan or photograph of the printed page and compare it to make sure that the fix was correct. Once the fix was applied, a corrected version could be made available in the store and pushed out to all the people who’d already bought it, (To save on constant updates, they might be generated just on a weekly or monthly basis.)

Eric Raymond said that “to many eyes, all bugs are shallow.” All typos could be shallow, too. It would take less manpower at the publisher, wouldn’t be that much work for people reading the books, and would result in better quality books overall. Of course, it’s probably never going to happen—and as long as publishers continue to let poor-quality e-books slip out, that’s a pity.


  1. At this point there is really no rational course for the ebook reading consumer except to completely boycott the large traditional publishers.

    In truth the publishers have so little regard for us, the readers, that they won’t even hire a handful of english majors fresh out of college to proof read their ebook conversions. It’s cheaper to just run the book through an automated scanner then hit it with OCR and a quick spell check to pick up non-words but leave false word substitutions behind.

    The situation is so bad right now with DRM, Region Restrictions, Poor Quality, Agency Pricing and Device Lockdown that I would advise people to buy books directly from authors or from small publishers who respect the reader.

    If you absolutely must read an ebook from a major publisher then you learn how to find it for free and send $5 payment direct to the author.

  2. My jaw just dropped: “The reader is invited to examine the next Jew chapters…”

    I find that there are more typos and incorrect words in newly published books across medium. Clearly editing is no longer the priority it should be.

    Do I think your idea that readers send in corrections and they’re reviewed on some kind of regular cycle is a good idea? Yes. Do I think it’s likely to happen? The cynic in me says no. Here’s hoping I’m wrong.

  3. Based on what I’ve been told by pro authors, the publishers farm out the scan-and-OCR to third party companies… and yes, they pretty much release them as-is.

    I’d also like to see some sort of peer-based correction system, if publishers won’t get off their butts to do it themselves, but the heck if I know how to set something like that up. It would be better for the pubs to do the correction, and maybe, when the ebook market starts to reach a significant size of their market, they will. But only if they see consumers avoiding their books over concern for OCR errors… and so far, the pubs aren’t seeing that. (Avoiding over concern for high prices, yes… accuracy, no.)

  4. Editors are still editing, but they are correcting files used to produce print pages. I think it’s clear fiction publishers will need to move to a single-source publishing system, just like information providers did a decade or more ago, when the web became the preferred delivery mechanism. Converting pint books via OCR or even from PDF does not produce good, reliable results. Novels need to become data, stored in XML or SGML repositories and corrected once when errors are found. Then they can be output in Mobi, or ePub, or pay-layout format as needed.

  5. I’ve worked with the Project Gutenberg Distributed Proofreaders project ( off and on since about 2003, and *those* books are thoroughly proofread for any discrepancies between the scanned text and the electronic text. If I worked for a publishing company, I’d be pretty embarrassed that a crew of volunteers is better at producing error-free ebooks than my company, which presumably employs at least a few professional proofreaders (probably on a freelance basis, but still…).

    I read a lot of ebooks, and I often find the quality lacking. It’s not just scanning errors, either; it’s silly things like a non-fiction book not having links in the table of contents, so you can’t skip ahead to a different chapter if you want. I end up bookmarking chapters I pass while I’m skimming, so that I can go back to them later.

  6. …oh, and I really like Karen’s idea (her post wasn’t showing yet when I was composing mine) about XML or SGML. I went to a presentation once by some guys who were doing exactly that for a university (Rensselaer Polytechnic Institute, IIRC): Everything from the course catalog to marketing brochures was dropped into an XML document that could then be output to the university’s website, or to DTP software for print production, or whatever. It was brilliant.

  7. I’m sorry, but I’m damned if I’m going to pay $9.99 or more for the privilege of doing a Big Publishing House’s editing job for them. If they want me to pay lots of money for an ebook, they’re going to have to prove to me that they’re adding value to compensate for the DRM-crippled books they’re selling me.

  8. The problem with mentioning quality problems in customer reviews is that you’re limited to doing so on sites that actually provide a spot for customer reviews (which leaves out Sony) and don’t remove/refuse to post critical ones (which leaves out B&N and Kobo), which basically limits you to Amazon, and (a) not everybody checks Amazon for reviews prior to buying from somewhere else and (b) some problems are store-specific (I know of cases where both Sony and Kobo aren’t just selling error-riddled copies, they’re selling error-riddled copies of *a different book than you think you’re buying*, while Amazon at least sells the correct file).

  9. It’s not necessarily always the publishers fault (at least not totally).

    I’ve got six books from Kobo with errors (words run together, completely missing italics, etc.) that the publisher has sent them new files for twice yet they’ve failed to replace the bad files. When first discovered I did some checking and the errors only occur at Kobo. The same books from elsewhere are fine. This has been going on since September and they haven’t done anything except their “support” has marked the ticket as closed. I’ve been in touch with Michael Tamblyn their VP for content a couple of times throughout this (including after Joanna’s post here and he’s done nothing either from what I can tell. Yet after all this time the books are still being offered for sale.

  10. It seems self evident that it is indeed the Publishers who are at fault. They are the ones supplying the files to the vendor. They are the ones accepting 75% and giving 25% to the author. Whatever the reasons they have no excuse except a carelessness and shoddy attitude to customers.

  11. Publishers don’t just get it wrong in ebooks either. The Faber Finds series of warmed-over backlist titles, for instance, has the most atrocious typos, obviously based on poorly edited scans. But Gutenberg is not innocent either. Take a look at most of their ePub versions of older poetry titles. The result is usually unreadable.

  12. Yet what you are seeing in publisher-provided ebooks, you also see in many, if not the majority or more, of indie ebooks, even those that are backlist files provided by the authors at places like Smashwords. I’m currently reading an ebook by a New York Times best-selling author that is one of his older books and one that he has made available at Smashwords. It is riddled with formatting and editorial errors. The formatting errors I can live with, but the editorial errors are distracting. I’m getting closer to just deleting the book. I certainly won’t buy any more of this author’s ebooks.

    The problem I find with the complaints is not that they aren’t justified — they are — but rather how one-sided they are. The complaints are always directed at the traditional publishers, rarely at the authors, especially at the self-publishing authors. Yet the problem is not confined to the traditional publishers.

    Just as the traditional publishers are cutting back on the editorial process in an attempt to increase their profit margin, so, too, do many authors who are responsible for the editing of their books now that they are bypassing the traditional publishers. And many of these self-publishing authors are selling their ebooks for the same price as traditional publishers. This is a problem that will not be resolved by simply directing one’s ire at the traditional publishers.

  13. >The complaints are always directed at the traditional publishers, rarely at the authors, especially at the self-publishing authors. Yet the problem is not confined to the traditional publishers.

    Very true! I think we can safely say the problem is most often the “publisher’s” fault (except when the digital seller declines to upload corrected files), but the definition of who the publisher is shifts with the situation. If you are uploading a file onto Amazon for publication as a Kindle book, you are acting as a publisher and you need to be sure the book is as error free as possible. This is true whether you are publishing the book for the first time or merely providing a digital copy of a backlist title.

  14. >aren’t just selling error-riddled copies, they’re selling error-riddled copies of *a different book than you think you’re buying*

    I’ve been sold the wrong book by Barnes & Noble (though they refunded my money after a brief argument). The wrong book problem is rare, but the error-riddled copy problem is common. I’d say 95% of the e-books I buy contain multiple errors. Most of the errors seem OCR related, but not all. For example, in a Mercedes Lackey book, I ran across an obvious search/replace error (the word “agreed.” stuck in after every mention of a character name for about 70 pages of the book). A basic spell-check would fix many of the problems I see.

    This issue is even more discouraging than the DRM issue because it could be so easily resolved.

  15. “For example, in a Mercedes Lackey book, I ran across an obvious search/replace error (the word “agreed.” stuck in after every mention of a character name for about 70 pages of the book).”

    Ha! This reminds me of an ebook I read recently in which every instance of “ADD” had been replaced with “AD/HD.” Unfortunately, the find-and-replacer hadn’t matched the case, so every instance of “additional” was “ad/hditional.”

    I think every ebook would benefit from a final proofread as part of the production process. This seems like very basic quality control to me, and I don’t understand why it seems to be done so infrequently… which is pretty much what all of us seem to be saying here. 🙂 But this has been a common complaint since the early days of ebooks, and it just seems to be getting worse as more books are published electronically. My impression is that the big publishers, who know better, don’t care, and the small publishers & self-publishers might care, but might not have the resources to have it done professionally or to do it themselves.

  16. I work at an online medical publishers in the formatting dept, we got taken over a couple of years ago by one of the big players, and now they’ve fully subsumed our tools and business, 70% of us are being laid off in the next few months. apparently all the editorial and production jobs are going to be handled cheaper and in more volume by outsourcing. What they fail to take into consideration is that a lot of it is outsourced already, and most of editorial and production (my dept) work is catching all the errors that the outsourcers miss.

    Simple reason? they’re paid by volume and speed. Which means automation, (spell checking and grammar as well as formatting) *not* double checking the layout (we have a four-line list of basic, easily caught errors, 20% of all files come back with these) and shoving stuff through if it looks vaguely okay. At arms’ length.

    And it’s always more expensive to fix the errors *after* publication than before.

    Let’s hope they’ve got the foresight to ramp up the customer complaints dept.

  17. (In passing.)

    The impression I have is that between Amazon’s Walmart-style price pressure on publishers and the small number of sales likely for a reprint edition of a fairly old book, the money just isn’t there to redo all the copyediting. If the ebook production wasn’t done upfront, at the first edition, then it’s only going to be done as Diane Duane is actually doing it–as a labor of love.

  18. and yet the Big Publishers try to tell us that their editorial input is what justifies the additional $$$ they’re charging for electronic files. Yeah, right. Just like file prices will go down once the paperback is released. sure.

  19. Indeed becca – and not only that but haven’t I read a lot in recent years about publishers who won’t market any more and insist the author does most of the marketing ? Sounds like a interesting strategy .. abandon editorial checks, abandon marketing and charge the same percentage. It’s no wonder they are riding the back of a dinosaur.

  20. “Amazon’s Walmart-style price pressure on publishers”

    Tell us more please ..

    And while you’re at it tell us why, if “the money just isn’t there to redo all the copyediting” they have the bloody cheek to persist in charging full price for a defective product ?

  21. @Randolph, that doesn’t explain errors in newly issued stuff. A brand new Harper Collins release I bought a few months ago had enough errors in it to fill eleven pages in Word. Now after I notified the editor they fixed it (took about two months), but they also continued to sell it with the errors too (I suppose at that point with it being a new release they felt they had to???). I paid the same for that book as someone buying a print copy (more really, because I could have gotten the print version at a discount).

    I’m sure it can be hard for publishers too. Everyone want their favorite authors backlist out NOW and so publishers are trying to come up with cost effective ways to get those backlists out fairly quickly however current methods are often unsatisfactory.

  22. I don’t know what’s wrong with the new books–it sounds like the publishers are simply cheaping out. As to backlist books and pricing, though, Walmart, er, Amazon wants to be able to set the prices of ebooks–they expect the publishers to absorb the cuts that Amazon is pushing on them. The result of this with Walmart has been vast amounts of poor-quality, in some cases dangerous products, and a huge transfer of manufacturing work to China: this is the only way to meet Walmart’s pricing demands. I don’t see how good editorial and design work can be outsourced to China or India, so, who knows? Ultimately, if no-one is willing to pay for good-quality books, there will be far fewer good-quality books.

  23. I agree with this article. I pre-ordered Melissa Marr’s “Darkest Mercy” (HarperCollins) from B&N and was very excited to read it on its release date in February. It is the final book in her series and the first time I have purchased a book from this series as an ebook instead of buying the hardback.

    To my grave disappointment, this ebook was so full of formatting errors that it significantly interfered with my enjoyment of the story. There were strange line breaks in the middle of paragraphs or words that made no sense. I stopped counting these line breaks when I reached 50.

    I called B&N customer service to tell them about the errors. I was told as a seller they are not allowed to correct formatting issues when customers bring them to their attention because to correct the file they receive from the publisher is a violation of copyright. The best they could do was refund my $9.99, which I told them to do. For just 20 cents more I could have purchased the hardcover from Amazon and no doubt would have received a nicely formatted version!

    Silly me: I had thought that buying the ebook would have meant getting the same quality as the print book in terms of being well-formatted. I hear some people complain when the ebook they downloaded from a pirated source contains typos and poor formatting. I was unsympathetic because, people, you get what you pay for.

    But here I purchased from a legitimate source that came to the vendor direct from a major publishing house… and it’s as rough and unpolished as any scanned book. This is ridiculous. Whether it’s an ebook or print book, publishers should make sure their customers receive a quality book.

  24. Fascinating. I am reading this article in 2014 after being concerned with exactlly the same problem in eBooks purchased for my new Nook reader. Three years after the article I’m sad to report that the situation is, if anything, worse.

    Barnes & Noble offer a great number of free eBooks for the Nook. Perhaps 90% of these have been obtained from the Google Library Project, and one must therefore assume, were scanned under Google’s direction.

    Scanned, but apparently not prrof-read even once. The books are riddled with scan errors and OCR artifacts to the point where — after downloading a sampling of more than 80 books, I found 90% of them to be unreadable. Whole lines and pages of random characters have been insterted, and whole words, lines, and pages of text omitted or scrambled.

    In the science books — Einstein’s original papers on Relativity Theory stand our — the equations have been rendered into complete gibberish by these errors.

    The bulk of the blame for this has to fall on the Google Library Project, the efforts of which not only fail to meet its stated goal of preserving these books and making them easily accessible on-line references, but are, in fact, actively undermining that goal. Google is not “preserving” these texts, is it DESTROYING them through negligence.

    Barnes & Noble, however, must share some of the blame here. They contracted with Google to be able to provide these texts, to beef up the number of free eBooks they were offering (presumably to compete with Amazon’s Kindle). However, Barnes & Noble has apparently eschewed any quality control responsibility for the free eBooks is it offering. The result is that they are, in effect, perpetrating a sham. Providing a half-million unreadable eTexts for its customers is no more use than artificially inflating their list of available books by simply making up thousand of bogus titles.

    Anyone who loves books, and accurate information, should be deeply concerned about this.

The TeleRead community values your civil and thoughtful comments. We use a cache, so expect a delay. Problems? E-mail