Steve Coffman, Director of FYI at the County of Los Angeles Public Library gives a European perspective on his ideas for Building Earth's Largest Library.
In March of last year I wrote an article for Searcher Magazine called “Building Earth's Largest Library.” (for full text see [1] and a follow-up piece at [2]). The basic premise of the piece was to apply the business model of Amazon.com, the bellwether of the new e-commerce revolution, to the library world. For example, what if we scrapped our limited local online public access catalogs (OPACs) that list only books in our own collections? What if, instead, we adopted a catalog like Amazon's, one that would show our patrons not only all the books we had, but also all of those we could get — either through interlibrary loan or in-print titles we could purchase for our patrons, if demand warranted it?
Suppose that when a patron searched this new catalog, they received a list of all the books available along with some indication of how long it would take to get them — just as they do in Amazon. And what if, like Amazon, we provided our patrons with rich bibliographic records showing cover art, tables of contents, synopses, excerpts, author biographies, reviews, etc., instead of forcing them to make do with the skimpy little, uninformative MARC records that make up our current catalogs? And suppose we allowed our patrons to order any book they wanted out of this catalog right on the Web, and have it delivered to their local library or directly to their home or office, from anywhere in the world, just like Amazon?
If we could accomplish all of that, we would have created the single largest library that has ever existed on the face of the earth — a library containing over 43 million titles (assuming a catalog equivalent to a combination of OCLC's Worldcat database plus the current Bowker Books In Print), almost ten times the collection of the 4.5 million titles listed in Amazon's own catalog of Earth's Biggest Bookstore. Such a library would make us the center of book information on the Web, and prove, once and for all, that when it comes to books, nobody, but nobody can beat the library.
Or at least that is what I said in the article. And judging from the large and generally enthusiastic response the piece has received in the United States, there are many who love the idea and can't wait to get started building it.
However, the vote was not unanimous. A number of people — particularly from your side of the Atlantic — suggested that a more accurate title for the article would have been “Building America's Largest Library” because while the model I laid out might have worked alright in the U.S., it would do little to help a patron in the UK, Germany, France, Italy, or almost anywhere outside the U.S. Others complained that it was unrealistic to assume that libraries from different countries could cooperate sufficiently to develop a single union catalog of their works, when experience in Europe had shown that they could not even agree on a common set of cataloging conventions. Some of you pointed out that interlibrary lending policies and procedures differed from country to country in Europe, and that even if you could make a catalog that could show a patron in Germany titles held by libraries in Italy or the U.S., it still would not be easy to request them. Others noted that I had totally ignored the language problem. Finally, several suggested that Amazon itself was a uniquely American creation, and questioned whether European customers would really want to adopt the same approach.
To each of these charges, save one, I plead guilty. I did write the article from a decidedly U.S. perspective. I built the model around tools like OCLC's WorldCat and Bowker Books In Print, that only work well in the U.S. (OCLC does offer some European coverage, but it is hardly comprehensive). And even though, I called it Earth's Largest Library I'm afraid the model I proposed for the U.S. might prove of little use to patrons in Frankfurt, Bath, Venice, Orleans, or Warsaw.
Having made that confession, however, I still disagree with those who claim that Earth's Largest Library or the Amazon model in general somehow could not apply to Europe. After all, Amazon itself has created two very popular sites directed exclusively at European audiences in the UK [3] and Germany [4]. And as far as I can tell, every major European online bookstore that I have visited from Bertelsmann [5] to Waterstones [6] and WH Smith Online [7] in the UK and FNAC in France [8] seems to have closely copied the Amazon model in designing their sites. They each strive to offer as broad a selection of titles as possible, including both books they hold in inventory and titles they can order from elsewhere, and each includes as much detail on each title as they can get, including cover art, tables of contents, synopses, reviews, customer comments, and anything else they can get their hands on.
Based on this evidence, the features pioneered by Amazon would seem as important to European customers as to those in the U.S. So the real question is not whether European patrons would appreciate a library catalog fashioned after the Amazon model, but rather how to create such a catalog within the complex environment of the European community. Clearly, Europe does present some special difficulties, but I don't think the problems are insurmountable, particularly in light of the strong tradition of cooperation among many European libraries. And if Europe's librarians could pull it off, you could establish very real benefits for libraries and library patrons all over Europe.
So, how would we build Earth's Largest Library in Europe? Let's look at the pieces and see how we might fit them together.
First off, let's not call it Earth's Largest Library, that would only encourage somebody from South America or Africa or Asia to write in and complain that it didn't apply to them. Instead, let's call it Europe's Largest Library or ELL, for short. ELL would aim to provide library patrons all over Europe with a single catalog where they could find information about books, videos, periodicals, and other intellectual property, available (1) at their local library, (2) at other libraries throughout Europe either via interlibrary loan or by visits to those libraries, or (3) available for purchase (“in-print” books) by the library should patron demand warrant.
ELL's new catalog would be a big, rich; wonderful catalog with user-friendly information supplementing the skimpy MARC records, enough information to help patrons choose what they want. The interfaces would make the catalog easy to use. They would correct misspellings automatically, recommend books based on previous borrowing habits, allow users to recommend books to friends, provide lists of books appropriate for particular age levels and lists of books that have won awards, and provide all the many other amenities developed by our friends on the commercial side.
Amazon originally built their catalog around Baker & Talylor's Title Source II database, which included “enhanced bibliographic records” with cover art, tables of contents, etc. for selected titles, and inventory records of books available from other selected suppliers. Amazon then added in reviews, customer comments, the recommendation engine, and other features, to produce their single union catalog of over 4.5 million records. In the original Earth's Largest Library article, I suggested that libraries could achieve a similar effect by creating a union catalog that combined a source of current in print bibliographic information like the Baker & Taylor database or Bowker's Books In Print, plus a catalog of books held by other libraries available through interlibrary loan, such as the OCLC WorldCat database.
Of course, neither the U.S.-focused Books In Print nor its equivalents, nor the OCLC database would do much good in Europe. However, similar sources of information exist for most European countries. Why not take advantage of them?
We could get current bibliographic data from any of the various “In Print” catalogs such as British Books in Print, Spain's Libros en Venta, Germany's Verzeichnis lieferbarer Bücher for Germany, and the like. In many countries, we could license large retail and wholesale catalogs already in existence to produce current bibliographic records, for example, the Book Data database, or the Waterstones, WH Smith, or Amazon.uk catalogs for the UK, and the Bertelsmann catalog for the UK and much of the rest of Europe. And if we wait a while longer, we might someday acquire detailed records on current titles directly from the publishers themselves, as Amazon already does. Many publishers are working hard to come up with an international standard for bibliographic information that would allow them to easily transfer catalog and book data directly to wholesalers, retailers, and libraries via EDI or electronic data interchange. And if these efforts bear fruit --- and there is every reason to suspect they will --- publishers may soon be providing much of the basic bibliographic information in our catalogs. For further information on European initiatives in this area see [9].
Creating a union catalog of works held by other libraries throughout Europe would be a little more difficult. There really is no direct equivalent of the OCLC WorldCat database for Europe as a whole. To extend the Earth's Largest Library concept to Europe, we would have to build one or something very similar from the ground up.
As a start, we should combine existing national union catalogs, bibliographies, and other tools which each country has developed to keep track of their national bibliographies and to help locate titles for interlibrary loan. I know — just from the brief investigation I have done — that these sources differ from country to country. Some countries have developed comprehensive online union catalogs; in other cases, even two or three major sources will still not cover everything. And, of course, formats differ for various bibliographic records (although not as seriously as they might, thanks to International Standard Bibliographic Description).
Probably the biggest problem is the lack of complete holdings information in some of these databases, including omission of some libraries. To plug these gaps, we would have to design a way for libraries to add their local holdings to the catalog — probably by matching their local OPAC records against the titles in the national catalogs. But these differences are not insurmountable. With a little thought and hard work, we could build a source that would serve as a WorldCat for Europe. It wouldn't be perfect, of course, but then, neither is OCLC.
When operational, we would have a single source that would allow the patron to easily determine what books were available in his library had and what books were available to him at other libraries elsewhere in the country and throughout Europe. For a good example of how you might structure something like this, have a look at the Bertelsmann's BOL.com site [10] that offers an online catalog that currently covers books in seven European countries and in six different languages. The site operates off a single database and transaction system with a consistent user interface across all of the countries that makes the catalog easy to deal with, even if you are not fluent in the language. You can only search one country at a time, but you can switch from country to country just by clicking a link.
It is not hard to imagine a similar interface for the ELL catalog. Patrons could access the catalog at their local library or on the Web through their library's Web site. They would start by searching for books in their own library and their own country, receiving a list similar to what you find in Bertlesmann or Amazon., e.g. “This title is available on the shelf at your library,” “This title is available through interlibrary loan and ships within 24-48 hours,” “This title is available through interlibrary loan and ships in 1-2 weeks,” and so forth. Patrons who did not find what they sought in their own country could simply click a link on the Web site to check the catalog in another country, just as they can at the Bertelsmann site.
Once we have all the book listings in a single catalog, how can we move the real books around? As part of the basic research for this piece, I wrote to librarians all across Europe and asked them how a patron would go about getting a book not in their local library collection. I learned that each country does things a little differently. The UK has a large centralized national supply system under the BLDSC. Italy has a national union catalog with decentralized supply. Sweden has separate networks for public and research libraries, each with its own set of finding tools and requesting procedures. Belgium also has separate catalogs for public and research libraries, but coupled to a national ordering system and a standardized fee schedule; some Belgian libraries have begun experimenting with direct patron-initiated requests. ILL systems differ from country to country. Possibly there are some very good reasons for the differences, but even if not, it would be the height of folly to assume that you could easily persuade any country to give up its own ILL system and adopt another.
What we need then, is a catalog with an interlibrary loan “mechanism” flexible enough to accommodate all of the different ILL models currently used by libraries throughout Europe — plus some yet to be devised. We need an ELL catalog that can be customized to fit the needs of each library.
Once we have built the “national union catalog” for Europe with holdings information for all participating libraries, everything else about that catalog — from the way availability information is displayed for the patron, to the way ILL requests are taken, to the way they are routed among libraries — could all be customizable and configurable to fit the needs of different countries and different libraries. For example, a public library in Sweden might configure the catalog so it would first try to borrow from Swedish public libraries, then from Swedish academic libraries, then from other Nordic countries, then perhaps from the BDLSC, and finally from any country in Europe with which Sweden had an interlibrary lending agreement. On the other hand, an academic library in the UK might configure the system to first try the BLDSC, and if unavailable, then other academic libraries on the Continent, other European public libraries, and so forth.
Somewhere along line somebody has to pay for these loans, so we would also need a flexible accounting system that allowed libraries to keep track of how much they owed whom, how much others owed them, and to pass some or all of those costs along to the patron if they so chose. None of this is easy, but none of it is rocket science either. In fact all the features outlined here have already long been available in the OCLC interlibrary loan system, and many others.
Of course, all of the complexities of interlibrary loan would be totally transparent to the patron. Few patrons care about library routing processes or where a book comes from. Most only want to know whether the title is available and how long it will take to get it. If the library permitted, patrons could click on a button and fill in their library card identification to request a book online, or perhaps the site would advise them to come in to the library to place their order. Of course, some patrons might actually want to travel to the owning library to use the book rather then requesting it on loan — either because the book was not available on loan in the first place (reference titles, special collections, etc.), or for personal convenience. Whatever the case, the patron should be able to click a button on the bibliographic record to find out what libraries held the item (complete with opening hours, maps and driving directions) and any special conditions for getting access to it. If the British Library held the title, for example, you would want to let the patron know that they would have to apply for a Reader's Ticket to use the item in the building.
The system we propose here would result in a pretty sizeable increase in demand for interlibrary loan, both because of improved information for patrons and easier ordering mechanisms. Of course it costs money to move books around (each ILL transaction in the United States costs around $28.00 according to the most recent statistics from the Association of Research Libraries). Some people question whether something like Earth's Largest Library would be economically viable, whether libraries — or their patrons, for that matter — could really afford to pay the increased costs of all those additional interlibrary loans requests that would arise if we ever really let patrons know what was available to them --- and allowed them to request it. However, good evidence exists that says we pay more now than we should to handle ILL requests. Many commercial wholesalers, for example, can pick, pack, and ship a book for less than $2.00 per item --- less than 10% of what it costs ARL libraries to perform the same task. The negative arguments also assume that books will remain physical items that will cost somebody something to move around. However, many people in the book trade now predict that in the near future, books will be stored as digital objects for delivery in a variety of different ways — printing on demand, downloading in whole or part for single or multiple use, etc. all depending on what the patron wants and how much he --- or perhaps his library --- wants to pay.
Terms and conditions for access to various formats may depend, at least in part, on library memberships. For an early example, see NetLibrary at http://www.netlibrary.com. NetLibrary supplies electronic version of books to university libraries on a subscription basis. Even though the library owns the title, the actual digital copy does not exist in the local collection at all; it is housed somewhere on the NetLibrary servers in Colorado. Students and faculty at a subscriber institution can access NetLibrary titles from anywhere on the Web, because their library subscribes to it. If we follow this trend out, we can see two things … first, as books become digital, the cost of delivering them to the reader will drop sharply, second, in the future the function of the catalog will not be to show what is in a local collection ---because we don't really care where the digital copy of the work is stored, but rather to show whether a particular book exists anywhere in the world, and if it does, to show the reader how to get access to it.
So, while Europe's Largest Library would probably result in increased demand for ILL and some increased costs in the short term, in the long run it would seem exactly the kind of catalog we will need to deal with the brave new world of digital publishing and digital collections.
Something else interesting happens when you build this kind of catalog and put this technology to work … it begins to affect all kinds of other functions in the library as well.
For example, if we had one great catalog with one rich bibliographic record for each title, why would we want to continue to spend hundreds of millions of dollars (or choose the currency you prefer) paying thousands of catalogers and technical services staff to add those skimpy little MARC records to our local catalogs? And why pay millions of dollars more on systems and software ---like Z39.50---trying to retrieve all those little records from our local systems? Instead, we would spend less time and money overall by working to create one, big, fat, rich bibliographic record for each title and then making that record accessible to all libraries over the Web. And if, as I have suggested, that catalog also includes records for current “books in print” not held by any library, then cataloging a book would simply be a matter of attaching the library's holding information to the appropriate bibliographic record — already waiting in the catalog. Books would be cataloged as they were purchased, eliminating cataloging backlogs. In most libraries, cataloging would become a simple part of the acquisitions process.
This doesn't mean, of course, that we would throw thousands of catalogers out of work. On the contrary, the ELL catalog would need their services more than ever, to help build the new catalog, to locate and compile the content added to the bibliographic records, to handle authority control and assign the best subject headings, and to build browsing categories in the catalog that arranged books in ways we have never had the time to arrange them before. This would require a lot of catalogers and cost a good deal of money, but in the long run, we would build a catalog like none our patrons had ever seen before, and the work our catalogers did would be available for all to use, not squandered trying to maintain skeletal records in some puny local system.
A centralized catalog would also help us spend our collection development money a lot more judiciously.
Up until now, collection development practices in libraries have not been terribly efficient. We've had to guess at what our patrons might want and because it was difficult and time consuming to get books from outside the local collection, we had buy a lot of material just in case people might want it. As a result, a high percentage of our collections sit on the shelf, while a small percentage of our books account for most of our circulation.
The catalog we design here would offer us an opportunity to change all of that. In the first place, if we included information on current “in print” titles in the catalog, we could allow patrons to play a role in collection development. We could call it “patron-centered collection development.” If a library patron found a title in the catalog the library did not own --- but which was available for purchase — they could click on a button on the bibliographic record to suggest the library purchase it. If we got one of these suggestions on a title, we might not pay too much attention to it. But if we received two or three or four patron recommendations on a title, we would consider purchasing it.
Moreover, if we had a huge catalog and could guarantee our patrons quick delivery of many titles, it would allow us to tailor our local collections more closely to patron demand. We could move away from “just in case” collection development to “just in time” collection development We could focus our collections on the items needed to meet immediate demand ...such as material needed for school assignments, material unique to our community and not readily available elsewhere, more copies of most frequently circulated books, and more books in high demand subject areas. Then we would rely on the interlibrary loan system to deliver less frequently requested titles. Of course, the catalog would also show us what books people borrowed heavily from outside libraries and suggest what we should add to our own collection.
I'm not suggesting that we design our collections totally around patron demand … otherwise you might end up with thousands of copies of the latest best seller, and nothing on organic chemistry. However, this sort of system would allow us to focus our collections more closely on the needs of patrons, and help insure that fewer of the titles we purchased sat idly on the shelf gathering dust.
Finally, if we had all of our bibliographic records in one catalog, why would we need to develop and maintain any more than one circulation system? After all, Amazon handles millions and millions of sales from millions and millions of customers from all over the world with a single transaction system. So why spend millions and millions of pounds, deutchmarks, francs, dollars, and lire on thousands of expensive local circulation systems that don't work very well anyway.
Better to make one good system that everybody could use and make it available over the Internet on a subscription basis. A great example of this is the BookSense program which the American Booksellers Association is launching this Spring [10]. It has a single catalog and transaction system like Amazon's, but is available to any independent bookseller on a subscription basis. So far over one thousand bookstores have signed up for the program. Libraries could do something very similar with a centralized catalog and circulation system.
The system should include at least circulation and acquisitions functions, but might also include other features commonly found in integrated library systems as well. Your patron database and holdings records could be maintained locally in the interest of privacy ... but the software that acted upon that data would come from the Net.
We could also toss out most all of the local hardware and software we now use to run out local systems as well. About the only thing we would need are terminals or PCs linked to the Net. Without the hardware and software to maintain, we would need far fewer local systems staff than we currently employ.
Of course, all of this would be optional. A library could still maintain their local circulation system if they wanted to and just tie the ELL catalog to it. But in the long run, I predict that most libraries would opt for the Web-based circulation system to take advantage of the great cost savings. Let me show you what I mean. According to the most recent Library Journal survey, the current library automation market is worth about $475 million dollars per year — and that just represents the money paid to the automation vendors, it does not include the millions of additional dollars spent on hardware, systems staff, telecommunications costs, and other things needed to keep local systems up and running. If we added that on, there is little doubt that we would be dealing with a total figure many times the figure that LJ quotes.
Now let's contrast that figure with what it costs some well-known and heavily trafficked Internet sites to maintain their catalogs and transaction systems. According to Amazon's Annual Report in 1997 (I used the 1997 report because at that time Amazon was still primarily a bookstore with costs more directly comparable to ours), Amazon spent $12.4 million on their catalog and circulation systems and that included all their hardware, software, content, and telecommunications charges plus the salaries of all the staff needed to develop and maintain it. During that same period they handled about 20-25 million sales, far more than the circulation of any single library, but still well below the 1.6 billion circulations of all U.S. libraries, to say nothing of volume in the rest of the world. So, just for good measure I thought we could add in the cost of the Ebay catalog which handles about 350 million transactions at $11.4 million per year, and the Yahoo sites which costs $49.9 million per year to run and gets about 61 billion pages views per year … and who knows how many billions of transactions.
Now lets assume that a Web-based automation system with sufficient capacity to handle the traffic from most of libraries in the U.S. and Europe would cost as much as all three of these Internet sites put together — or a grand total of 73.2 million dollars per year. That figure is still less than 20% of the $475 million libraries are spending on library automation software alone. Assuming we divide those costs equally among the 9700 libraries in the U.S., we would get an annual cost of $7,546.39 per library. If we factor the European libraries into the equation, the cost per library would drop even lower. Now, I know these figures are a little loose, but they are the best we can do right now without actually sitting down and carefully costing out the whole system, but even if things costs 2 or 3 or even 10 times as much as we've estimated here, we would still save money and get a far better catalog and circulation system.
If these calculations are right, there is little doubt that the Amazon model has the potential to improve the efficiency and reduce the costs of a number of key library operations. But the real winners with Europe's Largest Library would be our library patrons themselves. The scholars, students, business people, hobbyists, mothers, fathers, children, and millions upon millions of other users who pay our salaries and rely upon us to get them the information they need. Where now their library catalogs show them only what is available in local collections, ELL would allow them to select from several hundred million titles available in libraries throughout Europe. Where today we force them to puzzle over cryptic MARC records that do little more than identify the book, the ELL catalog would provide them with rich bibliographic entries that told them something about the book and its author, offer reviews, and supply enough information about the title to help them determine if they wanted to read it or not. And whereas, today, most of our patrons must come down to the library to use a catalog or check out a book, this catalog would be available on the Web 24 hours per day, 365 days a year, for access by our patrons from their homes, offices, day care centers, dorm rooms, or any other place with Internet access.
Finally, with a potential inventory of several hundred million titles Europe's Largest Library would be the single largest source of book information on the Web, far surpassing the modest collections of even the largest of the online bookstores. It would be a living flesh and blood incarnation of the ideals of Universal Bibliographic Control and Universal Availably of Publications, which we librarians have worked so hard to achieve for so many years. It would be the place people turned to first when they needed information on books. And it would prove once and for all that for access to the world's collected works on literature, science, history, fiction, or almost subject area known to mankind, nobody, but nobody, could beat the library — Europe's Largest Library.
Now, I'm something of an optimist, but even I recognize the many potential problems and obstacles that we would have to overcome to create anything as radical as this proposal. But what if, in the end, we couldn't solve these problems and overcome these obstacles. What will happen then?
The answer is nothing. The catalog will still be created. It only makes sense. As we discussed earlier in this piece, as books move over to digital format, the whole concept of a local library catalog quickly becomes an anachronism. Companies like Amazon and other online booksellers have certainly shown us what kind of bibliographic records our patrons prefer. Somebody soon will make the large and rich catalog along the lines we have described here. It may not have all the holdings of all the libraries in Europe, but it will have a lot of books, and those books will be described in lengthy, detailed records that will make our patrons happy.
The only difference will be that it won't be us that makes that catalog. It will be Amazon or Bertelsmann, or perhaps a library wholesaler like Baker & Taylor, or a library automation company like Innovative Interfaces. And whatever commercial company does the job we didn't do, will be the one we buy our library catalog information from someday, just as we now buy our periodical cataloging and indexing from commercial sources. That's all right, I guess. After all, we will still have the catalog. The problem is that we will cede all control over what is in it and how it is made. And the proud tradition of library cataloging which began in Europe with Panizzi and the British Library Catalog will disappear with a whimper, not a bang.
Of course, the future does not have to look like that. We can take the initiative and redesign our library catalog to provide our patrons with what they want — while honoring the best elements of traditional bibliographic control. But if we are not equal to this task, others stand willing to do it for us, though they may not share all our values. The choice is clear. The game is ours to lose.
Steve Coffman
Director, FYI
County of Los Angeles Public Library
12350 Imperial Highway
Norwalk, CA 90650
USA
Steve Coffman is the Director of FYI, the fee-based business information service
of the County of Los Angles Public Library. Steve as been with the service ever
since it was first launched back in 1989, and over the years, he and the staff
at FYI have developed some pretty innovative approaches to providing reference
and research services in public libraries-even if they have to say so themselves.
Some of these innovations include:
Steve and his crew are currently working with a number of libraries on a major project to develop live reference services over the Web, using Web Contact Center software such as that featured on the Lands End site - see <http://www.landsend.com/>. The first of these reference applications should begin to appear on the Web sometime in late Spring of 2000.
When Steve is not busy trying to keep these balls in the air, he writes occasional pieces for the library trades, including "Building Earth's Largest Library," "Reference as Others Do it" and the highly controversial "What If You Ran Your Library Like a Bookstore". He is the Editor, of ALA's Internet Plus Directory of Express Library Services, (a directory of libraries offering services on the Internet), and serves as impresario and program developer for the Southern California Online Users Group. Steve got his MLS from UCLA back in 1985, and has a B.A. in foreign languages from San Francisco State University. He lives with his wife Susan and daughter Kirsten, along with a couple of roadrunners, rattlesnakes, coyotes, and other varmints in the sage-covered hills north of Los Angeles.
For citation purposes:
Coffman, S.,"Building Europe's Largest Library", Exploit Interactive issue 4, January 2000
<URL: http://www.exploit-lib.org/issue4/ell/>
[HTML Validation] - [Accessibility check]
|
Issue Home | Editorial | Features | Regular Columns | News and Events | Et cetera | ||
|
| ||
| Go to Top |
A UKOLN Service. Contact Us. Copyright © 1999 |
Last Updated: 10 January 2000 |
