Saturday, June 14, 2008

OCLC's non-profit status

The New York Times ran an interesting story on non-profits that act like businesses. Apparently a number of states are taking a hard look at charities that "give nothing away," or have amassed vast wealth. A lot of day-care centers are worried, as is Harvard, where the endowment tops the GDP of more than 100 counties.*

Of course, my mind went to OCLC, the Dublin, Ohio-based global library-data organization.

OCLC's core business involves maintaining a central database of cataloging records, largely created by others, which member libraries pay to access. That OCLC was a great invention can hardly be denied. Personally, I think it has become a relic and an danger to the future of libraries. Agree with me on this or not, there's no question it is highly profitable—driving a steady stream of acquisitions—and in its fee structure calls into question the core idea of the non-profit.

So, why hasn't someone take away OCLC's non-profit status?

I Googled it up, and discovered that someone DID! In 1984 Ohio state courts stripped OCLC of it's charitable status on those very grounds:
“(A)lthough OCLC’s service may greatly enhance the ability of libraries to better serve the public, OCLC essentially offers a product to charitable institutions, for a fee exceeding its cost, and, as the board concluded, is not itself a charitable organization.”
So, what happened?

It seems the Ohio legislature passed some sort of private bill removing Ohio organizations involved in "library technology development" (and starting with the letter "O"?) from the court's requirements. Well, I guess that'll do it.

UPDATE: I'm working up a presentation on why OCLC's (also unfree) Dewey Decimal System needs to be killed-off, and what distributed, open classification could replace it. I'm all ears for anti-Dewey examples. And if any bright young cataloger with no love of Dewey wants to talk to me about heading up the effort, I'd love to hear from you.

*$35 billion, doing a quick check against Wikipedia. Of course, GDP is wiggly as heck.

Labels: , , ,

22 Comments:

OpenID Luis said...

The scarier way to look at Harvard's endowment is to compare apples to apples, and look at annual growth of the endowment v. national GDP, since GDP is an annual thing. There, Harvard's endowment grew $6B last year, which is still bigger than the GDP of 47 countries.

6/14/2008 12:43 PM  
OpenID pivyca said...

Hey Tim,

There are a lot of things I like about Dewey, but I def. think there's got to be something better and more up-to-date and average-user-friendly. I was sad that your screencast the other day cut off right at the part when you were about to talk about Dewey; I'd be interested to see/hear that part of the talk. I assume you already know about the Maricopa library District in AZ? They left Dewey for a bookstore-type model last year.

6/14/2008 1:54 PM  
Blogger Melinda said...

"I'm working up a presentation on why OCLC's (also unfree) Dewey Decimal System needs to be killed-off, and what distributed, open classification could replace it"

as long as the new system isn't based on color - i'm all for it :)

6/14/2008 2:46 PM  
Anonymous Ag1865 said...

There are several alternatives to DDC. We can rule out out the LC system because although widely used it's less of a proper classification and more of shelf listing. Ranganathan's Colon Classification has the most rigorous theoretical basis and is in use in India but is probably too complex and lacking in current literary warrant. UDC was based originally on DDC but has incorporated some of Ranganathan's theory. It too is potentially complex and anything but open, costing more than I can justify for private access. That leaves the Bliss Classification http://en.wikipedia.org/wiki/Bliss_classification which might be worthy of examination. No idea how open it is but work continues towards publication of further volumes of BC2.

6/14/2008 6:38 PM  
Anonymous GeekChic said...

I catalogue - though not well as I'm primarily a sys. admin. I agree that Dewey has it's problems, but so does LC (which _is_ a classification system), BISAC (what Maricopa, AZ uses) and any other system.

The main thing that Dewey has going for it is wide-spread usage. Once you learn how it works, that knowledge transfers to all but the largest (or most independent) public libraries.

This wide-spread usage also makes change difficult. Many of the smaller and poorer libraries I've worked with couldn't reclassify items based on changes that happened in new DDC editions - let alone start using a new system. Even the better funded libraries do changes of this magnitude only rarely because of the great effort and expense involved.

All that said, I have developed custom classification systems before for corporate clients. It's a fun exercise (though I am a geek).

6/14/2008 8:46 PM  
Anonymous Circeus said...

ag, A long, long time ago, a group was created to, theoretically, discuss this very topic and the possibility of resurrecting and updating Cutter Classification for the purpose of LT: Cutter Classification, Reloaded

6/15/2008 12:06 AM  
Blogger Tim said...

Yeah, I know. I want to do it now—and have some ideas how. Mostly, I need to get some enterprising your librarian who wants to make a name for themself. It's weird that I haven't found that person yet. LT has such cachet with certain librarians, and who wouldn't jump at the chance to do something fundamentally new?

6/15/2008 12:11 AM  
Blogger undeadgoat said...

Tim -- if I get a library science degree and a position of relative power and you still haven't got anyone, I'd do it. But you'd have to wait 6-8 years, at minimum . . .

as long as the new system isn't based on color - i'm all for it :)

Although for personal collections, color can come in surprisingly handy . . . at least, it made it easier for me to find my Dear America books when I was 11. :)

6/15/2008 12:06 PM  
Anonymous John Miedema said...

Tim, the non-profit status of OCLC is a good topic to investigate. I have been thinking a great deal lately about who should own book data. I think it should be a neutral source. While no one -- profit or not-for-profit -- is entirely neutral, my current view is that Open Library is the must neutral. I am interested in your views on the matter.

6/15/2008 4:20 PM  
Blogger Tim said...

Nobody.

6/15/2008 5:05 PM  
Blogger Amanda said...

As a cataloger, I find Dewey tedious to construct, but can see that it does have a flexibility that allows it to transcend large and small collections. Still, I'm open for something new.

As a library patron, a Dewey number (or LC or whatever) is only useful if I'm looking for a specific item. It's useless for browsing unless you're familiar with the system. And for some reason most libraries I've used only slap call # ranges on the ends of their shelves, when they could also have subject listings. A few signs could make a real difference. Say: Geography-Travel-History 900-999. And why not some movable shelf tags to help identify topics within the range? This is what makes bookstores browser-friendly and this seems like a simple way to help patrons out.

That said, it would be nice if there were a more intuitive organization/classification system for both patrons and librarians. Good luck!

6/15/2008 8:42 PM  
Anonymous John Miedema said...

Okay. Point taken. But practically speaking, does it make a difference who "hosts" it? Personally, I prefer LibraryThing over OCLC because LibraryThing allows people to add/modify book data. The only thing Open Library has on top of that is that it is open source. I am not trying to be provocative for its own sake. I'm just thinking this through.

6/15/2008 8:59 PM  
Blogger Tim said...

Yeah. My base position is that the data needs to be completely free. That's one reason I want a volunteer. LibraryThing would love to encourage such a thing and be associated with it, but if it's not going to pay the bills, we can't have it take up large amounts of staff time.

6/15/2008 9:17 PM  
Anonymous PatrickD said...

Hi Tim,

last year at IFLA I spoke about the need for a free classification system. The Presentation is at slideshare (slide 18). Or you can see a slidecast (Quicktime 33MB) Looking forward to your prasentation and maybe we can mail about this topic.

6/16/2008 4:31 AM  
Blogger Jeffrey Beall said...

You might be interested in my book chapter "OCLC: A Review" which was recently published in the book "Radical Cataloging: Essays at the Front." It's available online and open access at: http://eprints.rclis.org/archive/00013701/

6/16/2008 8:02 AM  
Blogger Becky said...

Yet another reason to raise some Hell at OCLC's back door when I move to Ohio. I have been doing research on OCLC's copyright claim to the bibliographic record database, and while the catalogers I spoke to saw it as mostly a non-issue, I disagree. The fact that OCLC has any copyright on the compilation of bibliographic records makes me very, very nervous for many reasons.

I am reminded of a quote that I read about the ownership of bibliographic data: "Protect your bibliographic data: it may be your most important investment."

6/17/2008 11:40 AM  
Blogger Shannon said...

I have a lengthy response to your brief comment about DDC.

I see a lot of potential in using the open source development model to create a clear and flexible classification scheme. It seems that many of the quirks in current classification standards originate in the fact that they are essentially one person's mental map imposed on a collection of physical objects, so a free scheme that is open to community development could be the antidote to some quirkiness. Is there any reason for a classification scheme to begin with information theory and descend through philosophy-religion-mathematics, other than a subscription to Enlightenment thinking? And yet, all major classification schemes follow this pattern. Obviously, documentation and updates are also stronger when they are open to a community's efforts, also.

I have been dreaming of a classification scheme that is more closely linked to subject headings (or tags), and could even generate or be generated by subject headings. With a faceted internal structure similar to that of colon classification, where each slot has a consistent position indicated by unique punctuation and is either filled or not filled, I think this should be possible.

Faceted classification seems ideal in many respects. Each facet can be quite descriptive, so that the entire class is very nuanced, yet the class is still legible if one facet is not filled. As one commenter mentioned, a great strength of DDC is its flexibility - you can draw a number out nearly as long as you like for a highly nuanced class or truncate the number even to the first three digits and still have a meaningful class for the object. The notation within those facets is an interesting communication issue - how to convey useful information with a few meaningful symbols?

Some necessary considerations for a widely-adopted scheme:
Language independence. This is the main problem I have with BISAC.
Sufficient treatment of fiction/literature/poetry (non-non-fiction works). It's what divides the DDC fans from the LC fans.
Material description. Both the content and the container are important. It's easy enough to have a manifestation facet, and it would be an added level of utility for librarians who have to actually put objects on shelves. In fact, FRBR-type facets could be a very useful structure.

Being a student and a newcomer to the field, I don't have the depth of knowledge that some experienced catalogers do, but I think that it also means that I am less entrenched in particulars and may be able to make some useful generalizations. I hope that I have done that. Forgive me if this is all obvious or already covered in your work.

6/17/2008 12:11 PM  
Anonymous ag1865 said...

Thinking further about this I wonder if a subject classification is any longer the way to go ... and I'm old enough to be still an enthusiast for faceted schemes.

My first alternative thought was something related to the PRECIS indexing system that was devised by Derek Austin for the British National Bibliography (BNB). It developed out of work on a new faceted general classification by the Classification Research Group that never came to fruition. It was used for quite a while by the British Library until it was ditched in favour of LCSH ! The key point of PRECIS is that enabled the setting up of multiple index points based on rotating terms while preserving the context. I suspect that it was in some ways ahead of its time, relying on accurate transcription of manipulation codes that could now be masked from the indexer. Just a thought.

Going several steps further I wonder whether we might better expend effort by ignoring 19th/20th century systems for organizing knowledge and heading for a system that was devised for web-based information, namely Topic Maps: http://www.topicmaps.com/tmc/

6/22/2008 9:57 AM  
Blogger prosfilaes said...

But, Shannon, there is no non-arbitrary order to the system, so information theory-philosophy-religion-mathematics is as reasonable as any. (Of course, Dewey doesn't follow it; even ignoring the questionable connection of 000's to information theory, mathematics starts the 500s, after social sciences (300s) and languages (400s).) Our current systems, like so many other things, aren't perfect, but their inertia is so great that it's going to take a huge amount of effort to get them replaced. In a day where they're no longer anywhere near cutting edge, why bother?

6/25/2008 12:18 PM  
Blogger Shannon said...

prosfilaes, it seems problematic to say "why bother"...Current cataloging practice is unwieldy, despite alternatives that could improve our work, because we just keep accumulating workarounds rather than making what seems like a radical change. I think a new classification scheme is a worthwhile project if there are physical objects to organize, if there is a chance that an individual may visit more than one collection of physical objects with the intent of finding something, and if the metadata surrounding a physical object is shared between collections. Things have to go somewhere, and people need to know where to put them and how to find them, so why not make the solution to those problems elegant?

ag1865, could you say more about the demise of subject classification? Classification and subject access seem so intrinsically linked to me. Do topic maps necessarily move us away from using subjects, or is it simply that the relationship between subjects is less hierarchical? The fluid relationships between defined components seems like one of the strengths of faceted classification.

6/26/2008 4:08 PM  
Blogger prosfilaes said...

Let's look at two projects: Unicode and Esperanto.

Unicode is the universal character set, the thing that allows/would allow you to use any language in the world on LibraryThing. When it was created in 1990, there were only a couple competitors in the field of universal character sets, and Unicode stomped them for various reasons. At least one was the fact that it includes all the characters from previous sets, down to order in many cases. ASCII is an awful basis for a real character set, but it's the only thing that will fly in the real world.

On one level, Esperanto is a failure; it is not, and will not be the world's language. It failed to make a dent in English or French as interlanguages. On another, it's been a fluent means of communication for hundreds of thousands, probably millions, of people, and it's been the biggest artificial language on the block for a century. At its birth, it pushed aside Volapük, the then-dominant, but much inferior language. In last century, many slight improvements to Esperanto have appeared, but nobody is interested in dropping a language with hundreds of thousands of books, a wealth of translations from all literature, dozens of in-print journals in various subjects, and speakers in every part of the world, for slightly better. The biggest competitor to Esperanto, Interlingua, did so by not taking Esperanto head-on, but infiltrating science as a language that anyone could read. Note that all the successful artificial interlanguages have absorbed a huge amount vocabulary and grammar from their competitors, English, French, Spanish, and in Esperanto's case, German.

How does this apply to a better Dewey? The point is that an innovator can win by being massively better than anything established in the field. Once something like that is entrenched, merely a better solution means nothing; you've got to show some serious flaws in the original. To put it in monetary terms, if you can recatalog and move a book in 15 minutes, and that dominates the time it takes to change the cataloging system of a library, the Oklahoma State University library (2 million volumes) will take a half a million hours; at $10.00 a man-hour, that's five million dollars. Tell me what's going to be so useful about your system to justify five million dollars to move to a system that puts the OSU libraries out of step with the library systems around them. Even if you're setting up a new library system, can you justify being out of step with the systems around you and forcing your patrons to learn a new way of ordering books? It doesn't strike me that trying to make a better linear order for books is interesting except as an intellectual challenge.

6/28/2008 7:20 PM  
Blogger undeadgoat said...

profislaes, your point about expense is legitimate, but I don't think patron re-education is that big of an issue. Some people who use libraries frequently are used to where on the shelves/what call number is associated with a topic or two they frequently read books on, but for the most part, patrons do not use the library enough to become familiar with the esoterics of the system. While there are many problems with overhauling an existing library, it might be worth it for a new library, or even a library willing to "take the plunge", to instate a less arbitrary classification system.

6/30/2008 3:53 PM  

Post a Comment

<< Home