Update: 5:04 p.m.: Bowker’s Andy Weissberg has just responded. Appreciated, Andy! I want to hear all sides of the debate. – D.R.
Should e-books have unique ISBNs for each electronic format—be it PDF, Mobipocket, TXT, HTML, you name it? That’s the recommendation of a Book Industry Study Group report (PDF alert)
This thorny issue has been the topic of past posts in the TeleBlog and elsewhere. Alas, the debate isn’t so academic. ISBN bloat could steal money away from small e-publishers, editors and writers. Granted, just the publishers pay directly. But given the stringent limits of small-press budgets, the others will suffer from the bloat.
Yes, I, too, worry over problems such as retailers’ in-house identifiers adding to confusion. I love the idea of trustworthy identification standards. Down with “rogue” identifiers! I just don’t want publishers of any size to pay more for ISBNs than they have to, in order to pre-empt the rogues.
eBabel-related, natch
The ISBN problem is merely one more outgrowth of the eBabel one. And remember, it isn’t just an issue of the multiplicity of format alone. What about unique identifiers for new editions or perhaps customized ones? Format-unique ISBNs will literally multiply the existing complexities and costs.
Meanwhile I’m not impressed by the argument, from Bowker’s Andy Weissberg, that $400 for ISBNs is an acceptable burden for small-press titles. Better that the cost normally be $80-$100 at the most and that the savings go the publishers and their authors and editors. ISBN bloat will literally take money away from those who should be able to put out e-books on a shoestring. Publishers and those paid by them are the ones who create the core value here, not ISBN-related companies, regardless of the useful services that Bowker and the likes offer. And who knows? Maybe some of those shoestring books can grow into best-sellers in P as well as E.
Anti-ISBN bloat sentiment from university press digital projects manger
In correspondence with me, Claire Lewis Evans, who deals with digital projects at the University of Alabama Press, has weighed in.
As she sees it, publishers should not be under pressure to offer an ISBN for each electronic format—the recommendation of the study group report. She notes that Bowker and Nielsen, “who traffic in ISBN,” paid for the report.
ePub as a solution
I can’t read minds, and I’ll welcome comment from the defenders of the following: “The practice of using a single ISBN for all digital manifestations of the same work is strongly discouraged.” I’ll e-mail BISG’s Michael Healy to see if he can’t line up a response from Michael Holdsworth, the author. Does the latter really agree with Andy Weissberg and consider $400 an appropriate burden for a small press? Or even $275? In my ideal world a small publisher doing an e-book and a trade paperback would have to buy just two ISBNs: one per medium. No need for fancy packages.
For now, I would agree with two small publishers, Elizabeth Burton and Rob Preece, that a single ISBN for an e-book is hardly a disaster for small publishers. Here’s to more money for creators and less for bookocracy! And speaking of precision, the more focus on format, the harder it becomes to deal with other metadata. Some ISDN mavens note that large publishers may have different, more elaborate needs, but reduction of format complexities can help the big guys as well.
My own hope is that eventually ePub will be the standard, without proprietary DRM to muck it up, so that budget-strapped publishers don’t have to pay for a zillion ISBNs, while book buyers at the same time can precisely and reliably identify the purchase they really have in mind (Michael Cairns is certainly right about the desirability of preempting rogue ISBNs!).
Claire is concerned about the XML-related challenges of putting out books in ePub, and she’s right, but progress is being made, as shown by the ePub capabilities of InDesign software. My forthcoming novel, The Solomon Scandals, will appear from a small literary press. And guess what. It’ll be in ePub. What’s more, other possible ePub tools exist, such as one from BookGlutton, and I know some others are on the way.
Liz Daley on the unique identifier issue
Let me end with thoughts from TeleBlog contributor Liza Daly of Threepress.org, who herself has been working with ePub apps and has a special interest in open source tools for academic publishers. Here’s her reply after I forwarded Claire’s note:
“Helping academic publishers introduce ePub into their workflow is definitely something I’m interested in (especially since ePub is synonymous with getting content into XHTML, which opens up the whole web as a distribution channel).
“I don’t know much about the business side of ISBNs but from the ePub perspective, each ePub book is required by the spec to have a unique identifier. However, there is no authority for enforcing this, and the spec states that “specific uses of URLs or ISBNs are not yet addressed”.
“It’s an open question what should go into this unique identifier field—some publishers are using ISBN-13 [link added], but I’m in discussion with one about how to handle multiple revisions of the same ebook (should they get a new identifier or not? And if so, what?) And of course for public domain or content that was never published, ISBN doesn’t apply.
“But the format, at any rate, does not require an ISBN at all; publishers can (and certainly will) generate their own IDs when ISBN doesn’t apply, or isn’t economically feasible (e.g. for content designed to be freely distributed).”
Question: I’d love to have a breakdown of how much of that $400 or $275 goes to various people, and for what services. Maybe there’s a hellva value here. I’d just love to know the stats, and see how what automation-related efficiencies might lower the costs.
I’ve already weighed in on this but can’t resist another shot. In addition to the purchase price of ISBN numbers (and that purchase price is scaled so it’s cheap for big publishers but expensive for small publshers and horrid for self-publishers), there’s the administrative burden of updating the Books in Print system (once per ISBN), and pure ineffectiveness of this system for those of use who use Fictionwise’s Multi-Format (non-DRM) system since Fictionwise decides what formats it wishes to make available. Am I supposed to provide them a blank set of ISBNs to use whenever they decide to support a new format?
Certainly requiring an additional ISBN for each format would make me reluctant to consider adding additional formats (can anyone say ‘hurdle to adopting ePub?’).
If it ain’t broke, don’t fix it. The ISBN system is definitely not optimal for eBooks, but we’ve been able to work with it. Publishers who want to register separate ISBNs per format can do so. Those who don’t, don’t have to. The next customer who tells me she’s confused will be my first.
Rob Preece
Publisher, http://www.BooksForABuck.com
We’re facing this question too; at our publishing house, we have decided to give the ebook version of a publication just one isbn, even if there is a mobipocket version and a pdf version; we count them as one. Multiple isbns for one electronic publication makes things far too complicated from an accountant’s point of view. Of course, we hope that in the near future there will be just one standard for e, eliminating the need for extra isbns.
David,
All of your points are well-taken. In addition to this response, I have asked my colleagues at Nielsen, BISG and the International ISBN Agency to work with me to formulate a response.
I can, however, at this time, directly respond to your questions pertaining to “a breakdown of how much of that $400 or $275 goes to various people, and for what services”…
1) In addition to its Books In Print databases and web-based platforms, Bowker continues to invest in new and improved systems and platforms that enable publishers to maximize the discoverability of their books and title information within thousands of bibliographic catalogues, retail channels, search engines, social networks and others used by consumers, library patrons, librarians and others to find and purchase books.
Our newest platform, MyIdentifiers.com enables publishers to freely search optimize their titles in conjunction with ISBN purchases while also enabling publishers to channel those listings to places where they sell their books. Bowker also operates faceted search technologies like Aquabrowser which help refine and deliver results information about published titles during searches in both library and retail markets and locations.
2) Through MyIdentifiers.com, and especially for self-publishers and authors, Bowker makes it easy to provision and manage title-data that we also immediately optimize for discoverability and search engine indexing via our new SEO Title Card offerings (like a free mini search optimized web site) and distribution to thousands of channels, retailers, etc. as described above. These tools and services are included with ISBN purchases.
3) Quite often, publishers provide data that is incorrect, has mis-spellings, wholesaler/distributor detail and other critical elements necessary to correctly inform the supply chain and consumers directly about key attributes associated with their titles. Bowker maintains a full-time staff dedicated to ensuring data quality on behalf of publishers.
4) Bowker also enriches publisher title data with other critical elements that enhance discoverability within bibliographic catalogues, including and not limited to reviews, citations, media-list mentions, best seller list inclusions, etc. All at no additional cost to the publisher when it comes to the inclusion of this detail within the title record in Books in Print which ultimately is what is made discoverable in the channels that publishers want their books to be contained and highlighted within.
So, basically for $40 per title as you’ve described, publishers can uniquely identify their titles across the supply chain to comply with ISO-approved industry standards, effectively market their titles (e or p).
Compare that $40 to what it would cost a publisher to buy keyword advertising on search engines, issue a press release over a major newswire, create a web site, etc., and its quite easier to rationalize.
As for needing separate ISBNs per each format, we agree with our colleagues that without doing so, the downstream implications of not being able to provide consumers with accurate identification of a product they want or need based on the reading device they use and within structured bibliographic catalogues, are major.
Again, I will be circling back with my colleagues but wanted to respond to this as quickly as possible.
Andy W.
Bowker
It is not needed. Any server by running a simple script can provide a Worldwide Unique Id, for free.
What is more this can be used for any document, not just e-books.
Plus we need Ids that are XML compliant and can be used to identify fragments of texts.
Please have a try of this simple script:
http://www.lestec.com.au/aui/auilist.cgi
The IDs always begin with an alphabetical letter, they are case insensitive (retaining the same value in upper and lower case).
The script is simple, one, and only one, copy per server is the rule — it uses the IP address and a date, simply encoded and producing IDs sequentially.
I helped design this system precisely because ISBN cannot work it is a good system for print, but not for electronics, for it is not just e-books but most files that need to be unambiguously identified.
Our idea, was to aim towards an independent electronic registry of publications, authors, translators, publishers and most important editions — so that any work can be found unambiguously.
Thanks for a very interesting piece David. This issue has also come up in the UK via Nielsen, who are taking a firm line on this issue, saying that they will exclude e-books from its database that do not comply with the International ISBN Agency standard that each different format of an electronic publication has a separate ISBN. The news coverage was followed up with a full-page advert in the trade press outlining this policy.
OK, class. Show of hands. How many here have USED Books In Print any time recently?
For myself: not in years. I turn to the Net, I turn to abebooks, I turn to ebay, I turn to the NYPL.
I search by author, I search by title.
I have NEVER searched by publisher or ISBN.
Why should ISBN even have a place in the ebook world? Can someone answer *that* question first of all? (Hey, sneer all you want, Bowker, at what you might consider an ignorant question, but you have to justify *your existence* now. Welcome to the end of your monopoly, baby.)
I work in a library and I frequently use BIP online when I’m ordering. I use it because it includes professional reviews, and that means I don’t have to chase down a bunch of back issues of journals. As I’m flipping through a catalog, if I see something that looks good/interesting/promising, I can go to BIP online and look at the reviews.
Whether it’s BIP or Amazon, 75% or more of my searching is by ISBN. If as I’m reading book reviews I see something that looks interesting, or I think a friend would like, I pass the information along to them in the form of an email usually with a subject like book rec, and the body is just an ISBN.
In searching the library catalog, if I’m checking before doing an order, I use an ISBN.
For me, and most of the people I know, ISBNs are one of the primary ways we search for and share information about books.
I love that BIP online includes professional reviews. That said, I think Andy Weissberg left something out of his loving description of all the value added things that BIP does with your ISBN. He neglected to mention that BIP online is a subscription database with a fairly hefty price tag. Those value added things they do, I’m willing to be FAR more of that cost is paid for by subscriber than by ISBN buyers.
As for #3, call me heartless, but the publisher should take care of that. Don’t clean up their data for them and they should learn pretty quick.
@De: You are a professional in a field. Therefore, niche. And if self-pubbed ebooks are being sold direct to the public, cutting out you lot, why should they have ISBNs? To help you keep your job?
>>>I love that BIP online includes professional reviews.
Some could see that as snobbery. Teleread has already done the death of pro reviews here.
I still see no justification for Bowker’s continued existence.
@Mike Cane: Scattered comments-
Niche, yes. And I admit that. Even more so because I don’t buy fiction or normal non-fiction. I probably should have said that.
But my point from yesterday, which I apparently failed at making, was that while I use BIP and the added information Bowker includes in there, *we pay for* it in the database subscription. All that additional information, that’s something that *Bowker* chooses to do and *we* choose pay for. The publisher’s *shouldn’t* be covering those extra costs.
What I realized yesterday after I posted that was that I didn’t ever comment on ISBNs and ebooks. As for that, I don’t think each format needs it’s own ISBN. A note somewhere saying ‘released in x, y, and z formats’ would be nice so that buyers would know if the format they need is even available. Of course as formats are added that information would need to be updated and that could be a nightmare, so I don’t know how practical the idea is. But a note would be nice. Ideally, of course, the multi-format idiocy will go away and it’ll be a moot point. But until that happens, no I don’t think each format needs an ISBN.
If the ebook is revised/changed, then yes, new ISBN would be appropriate just the same as for a new version of the paper book.
If the publishing world wants to settle on some form of edition specific ID other than ISBN, go for it. But consistency across ebookland would be just as nice in this as it would be in a file format. I suspect that if ebookland moves to some other ID, Bowker will either include that information right along side ISBNs or lose all relevancy as publishing moves from print to digital.
As for professional reviews:
I buy reference books and if I’m looking at spending $200-$3500 on a set of books, I want as much information as I can get about that set before buying. The BIP database has multiple reviews in one place, so I don’t have to track down the individual reviews in multiple issues of multiple journals scattered around the library in other people’s inboxes. Again, *we pay* for that in the subscription.
Multi-volume reference sets are not the kind of thing I’m going to find reviewed at most book blogs. Niche product, niche reviews, niche buyer. I don’t think that makes my need for this information any less valid.
If I’m looking for information for Reader’s Advisory, then yes absolutely I look online. Although I still like to compare those reviews with the reviews in Booklist and Library Journal. Frequently they will include a comment for teen/YA readers about levels of sex or violence. Those are useful for the senior and/or conservative adult population as well. So yes, niche here as well because I’m mostly just looking for an idea of what the book is about so I know who to suggest it to and who not to suggest it to.
ISBNs work well and have been a great asset to printworks, but issuing IDs centrally on a fee paying basis, is a lot different to getting IDs instantly.
Formal e-book publishing is one thing, perhaps ISBNs could work. However, where they cannot work is amongst a huge tide of electronic publishing that includes, articles, OCRed texts, photographs/graphics, audio-visual, historical records, in short the etc’s., of our electronic age.
My original passion was archaeology, fragments of cooking pots, discarded objects, broken fragmentary remnants of the past. Recently in Britain I met an amazing amateur field-walker. He collects surface fragments, brought up by ploughing, records their position and maps them. His achievements now being written into a paper are awe-inspiring.
With today’s technology he could do nearly all his recording with a cell phone. Ie spots a fragment, associates it with a GPS position (it has to be more accurate than the present limited system), takes a photo of it, uses the internet to get an ID, and also send the information to be stored.
This is the near future.
If he then writes his paper, it can be ID, but every reference to an object, each photograph, each location reference on a digital map, can be associated in order thanks to having a simple thing like an ID. ISBNs could never do that, they were not intended to do so.
What distinguishes an e-book, from a business letter, from an archaeological object, is not having a special number, but a different place that registers that number.
I am using this example as to placing e-books within their greater context (the communications revolution). It is not hard providing world-wide unique Ids. They can follow any number of forms. ISBNs are designed for the much slower age of print, that is their only problem
What is an e-book and what isn’t is something of an artificial division. I do not believe ISBNs have a future in this area, but they could have, if integration is allowed. That is if some means are found to make ISBNs compatible to other sources of IDs a form of mutual recognition.
And it would not be hard.
Consider the ISBN data base simply spawning a free registry for e-books, MS and other publishable material. That would serve everyone’s interest. Then paying a fee for promotion to ISBN would in fact be a promotional tool.
Andy W. Bowker, please consider this as a viable option, integration of ISBNs with other forms of IDs would serve to integrate electronic and paper books together. There are issues, but if you are interested write to me greg.schofield @ iinet.net.au, or reply to this thread.
The same goes for anyone else. It is not the mechanism of producing IDs, but what they get attached to that needs to be hammered out. Meanwhile I will try and write an article on the issue for Teleread.
Bruce Sterling. The Internet of Things.
http://www.the-internet-of-things.org/
I agree with De. If you can go out as far as adopting a totally new identifier for the books then why not simply append the ISBN of the ebook with some unique code like for each format? The first 13 digits can associate the ebook with its corresponding print edition while the appendage identifies the format