Wednesday December 13, 2006

8 million books described in LibraryThing

Over at LibraryThing, they announced that they had passed 8 million books today. Now I like and admire LibraryThing and the people behind it quite a lot, but I found the descriptions of those 8 million books a bit disingenious. LibraryThing says, "LibraryThing has between eight and fifty times as many books as the World's Biggest Bookstore, in Toronto, Ontario, if laid end-to-end, LibraryThing's books would stretch 1,075 miles, LibraryThing hold 58,580 books by J. K. Rowling, LibraryThing has 80 times as many books as you have hairs on your head," etc.

The problem is, LibraryThing itself actually does not have a single book. The place where they are correct is when they say, "If LibraryThing were a library, it would be the 14th larget library in the United States." That conditional is important. I may be splitting hairs, but this language question brings up an interesting question - just what is a library? If you have descriptions of books, but no actual books, are you a library? But what if your constituent members have personal libraries? Is the aggregate description of the collected holdings of physically extant libraries, personal or institutional, a library itself?

I think not. Nor are you a bookstore, since in neither case can the institution known as LibraryThing actually provide a book to anyone in and of itself . And that ability to provide resources is the key. Require payment for
providing the end product, you're a store; don't require payment, but seek subsidies for
providing resources to some defined group of people, you are a library. Don't require payment from anybody, seek subsidies or define users, and you are some sort of barter system likely based on personal, rather than institutional, collections.

LibraryThing doesn't
fall nicely into any of those categories, though to be fair it nicely enables the buying of books from several vendors (now including 2 local stores), the location of local libraries holding copies via WorldCat, and looking at popular swap sites like BookMooch. It plays some sort of very interesting middle role that greases the discovery and serendipitous browsing of titles, especially fiction, before any of those other transactions begin and its ability to play that role well grows more potent with every new entry.

What I think I'm driving at, however, is that having book data itself, while really really interesting and very potentially useful for stores, libraries, and swaps, doesn't itself constitute a library. Libraries (book libraries anyway) and book stores, by my definitions, are not just about the collections of physical books they have, though that is the first principle. They also  have an economic dimension that revolves around the transactions involved in both the procurement and providing of book resources. And I don't think you can get away from that; not easily anyway.

This isn't to suggest that because LibraryThing is not a library or a store there is a problem. Far from it. Tim Spaulding et al. have done something really unusual in creating something novel.  LibraryThing seems to be to be exactly what its name suggests, something for which we don't quite have a word, or at least a good one.  If I were to describe it, I would call it at its smallest a personalized version of OCLC (which is based on institutional holdings), but I believe that doesn't do a good job of conveying it to nonlibrarians nor does it fully encompass the scope of what is happening. Perhaps in Geman we could string together a series of words to make a new one that perfectly described LibraryThing, but I'm at a loss in English. Anyone have a good idea for the class of objects that LibraryThing (and OCLC I believe) belongs to? Other than the best resource around to study personalized book reading (not book buying, that's Amazon) - which makes me wonder just when will the first thesis based on  LibraryThing data arrive? A one word name. Neologisms welcome.