Some library systems will do as noted above and send books which are weeded in the branches to a central repository (which may have closed or open stacks and/or the tiles still in the central catalogue or removed from them) for a given period, others will immediately withdraw them completely from the system, in a book sale at the branch level or by giving them to a charitable organisation.
Weeding is done mostly for lack of space, and is driven by title number goals. Last time I checked, a few years ago the "modern" goal for "modern" librarians was to keep about 20,000 to 30,000 titles in the adult section of any small branch library and a lesser number in the children's' section. Even a small city like Ottawa (pop about 700,000) can boast millions of titles in its central (Web-based) digital catalogue. The digital catalogue entries follow extremely strict rules, AACR X (Anglo-American Cataloguing Rules version X) and give precise locations using the Dewey system.
I wrote all this to ask you a "simple" question:
How are you going to keep track of all those titles in bookstores, if you go to the model you're advocating?
Right now the typical bookstore, large or small has a small fraction of the titles found in most library systems. When I was trying to do a PhD in Library and Information Science I looked up numbers (thank you for refreshing my memory in your blog!) and was surprised at how small they were, compared to even the numbers in a small branch library. What was more surprising and interesting still was the ratio of staff to titles. I suppose that the high number is essential to do all the moving around of incoming and (regretfully) returned books but a corollary of this is that the staff in a present-day bookstore has a much bigger chance to get to know a lot of the titles than the relatively smaller staff of a library. If you wanted to find a book in a good bookstore, any bookseller or clerk could tell you for sure where it was in the store. This is why a bookstore could always get away with a topical arrangement of books, and avoid the high cost the typical library incurs in processing a book/title, describing it with AACR X and classifying it with Dewey. The cost of processing a book/title in a library system can easily double or triple the cost of the book/title to the system. And when I say double or triple I mean double or triple the list price. What bookseller can afford this? If they centralized processing Cerberus probably could, but could any other store bear this?
Then there's the "push" aspect of bookselling which you've covered so well in your blog. You get to know the customers, so you get to know what to order and you get to know what to propose to them. How are you going to do that with the library-sized title lists you're describing? There's a also a "push" phenomenon in libraries but it depends completely on the work of the reference librarian and her (or his, in some instances) masterful handling of the digital catalogue, that costly catalogue made out of book/title descriptions that cost two to three times the list price of the book/title.
The Amazon model seemed like a good one, a few years ago, as a way to avoid the classical library processing/cataloguing of titles. All you needed was to add more subject tags, and put in more visual elements than any library would have thought possible, on their digital catalogues. After all, for the visually-oriented segments of the population those book covers and colourful spines are incredible finding aids in any bookstore. Unfortunately, it all broke down in the last years, as they got in the same million-title numbers as big city libraries. The system still works fine if you're looking for popular stuff, but for anything off the beaten path you'd better go first to the library of Congress and the Web catalogues of the biggest public library systems in the US, find the ISBN and other descriptive elements there and then go back to Amazon with it.
I'm not saying that there's no solution to dealing with this kind of "title richness" problem. I'm just saying that it's an extremely interesting problem. I hope you'll be thinking about it, as well as the technical demands of POD.
no subject
Date: 11 Jan 2009 08:29 am (UTC)Weeding is done mostly for lack of space, and is driven by title number goals. Last time I checked, a few years ago the "modern" goal for "modern" librarians was to keep about 20,000 to 30,000 titles in the adult section of any small branch library and a lesser number in the children's' section. Even a small city like Ottawa (pop about 700,000) can boast millions of titles in its central (Web-based) digital catalogue. The digital catalogue entries follow extremely strict rules, AACR X (Anglo-American Cataloguing Rules version X) and give precise locations using the Dewey system.
I wrote all this to ask you a "simple" question:
How are you going to keep track of all those titles in bookstores, if you go to the model you're advocating?
Right now the typical bookstore, large or small has a small fraction of the titles found in most library systems. When I was trying to do a PhD in Library and Information Science I looked up numbers (thank you for refreshing my memory in your blog!) and was surprised at how small they were, compared to even the numbers in a small branch library. What was more surprising and interesting still was the ratio of staff to titles. I suppose that the high number is essential to do all the moving around of incoming and (regretfully) returned books but a corollary of this is that the staff in a present-day bookstore has a much bigger chance to get to know a lot of the titles than the relatively smaller staff of a library. If you wanted to find a book in a good bookstore, any bookseller or clerk could tell you for sure where it was in the store. This is why a bookstore could always get away with a topical arrangement of books, and avoid the high cost the typical library incurs in processing a book/title, describing it with AACR X and classifying it with Dewey. The cost of processing a book/title in a library system can easily double or triple the cost of the book/title to the system. And when I say double or triple I mean double or triple the list price. What bookseller can afford this? If they centralized processing Cerberus probably could, but could any other store bear this?
Then there's the "push" aspect of bookselling which you've covered so well in your blog. You get to know the customers, so you get to know what to order and you get to know what to propose to them. How are you going to do that with the library-sized title lists you're describing? There's a also a "push" phenomenon in libraries but it depends completely on the work of the reference librarian and her (or his, in some instances) masterful handling of the digital catalogue, that costly catalogue made out of book/title descriptions that cost two to three times the list price of the book/title.
The Amazon model seemed like a good one, a few years ago, as a way to avoid the classical library processing/cataloguing of titles. All you needed was to add more subject tags, and put in more visual elements than any library would have thought possible, on their digital catalogues. After all, for the visually-oriented segments of the population those book covers and colourful spines are incredible finding aids in any bookstore. Unfortunately, it all broke down in the last years, as they got in the same million-title numbers as big city libraries. The system still works fine if you're looking for popular stuff, but for anything off the beaten path you'd better go first to the library of Congress and the Web catalogues of the biggest public library systems in the US, find the ISBN and other descriptive elements there and then go back to Amazon with it.
I'm not saying that there's no solution to dealing with this kind of "title richness" problem. I'm just saying that it's an extremely interesting problem. I hope you'll be thinking about it, as well as the technical demands of POD.