Tim Spalding on Social cataloguing

What it is, and what it means for libraries?
Tim Spalding founder of LibraryThing

Introduces self as a failed academic, worked in publishing, started LibraryThing.

Warning: Library Science being practiced without a degree

Started as a personal project, now a company. 850,000 members who catalogue their personal libraries - so far 44million books. Available in 12+ languages. (Not Māori but would be open to that - translations done by members.)

Social cataloguing is "what I say it means" because he invented it! It's what emerges when personal catalogue goes social. It's becoming increasingly important to libraries. Used in LibraryThing, Shelfari, GoodReads; Visual Bookshelf, BooksWeRead.

Ladder of social cataloguing:
- started as personal cataloguing and grew from there
- users climb the ladder
- climbing the ladder is more altruism, more cooperation, more social. But participating is primarily for self. There's some application to libraries but it's different there.

Live demonstration of adding "History of New Zealand" by Michael King to his bookshelf. Mixture of tags - "new zealand", "history", "lianza", "interesting". Bookshelf with ratings. Can add from Amazon or many other bookstores or even libraries - 10 libraries in New Zealand contribute data. Can view libraries by list, cover, tag (list or cloud); author cloud or portraits. Statistics on language, number of characters, places. Reviews and ratings. Members' profiles - social networking component but LibraryThing is more about content than people, reflected in focus on users' names rather than user icons.

23,000 people adding Twilight. All doing it for themselves but as a result there are now 1200 reviews people can read; tags are added, recommendations are generated ("Will I like it?" - it correctly predicts he won't like Twilight. :-) ) Can follow a feed of new recommendations. There's also the "unsuggester" - trying to be entertaining around books.

Example of Neuromancer - library of congress has bizarre subject headings; LibraryThing has "cyberpunk" and you can click through to read more cyberpunk. "Chicklit" is sorted by how many people have called it that; cf Library of Congress "love stories" which is just either/or, no sorting. Idea of prototypes - a robin is a really good example of a bird, a penguin is a kind of okay example of a bird...

Non-romance readers think romance readers read romance, but they don't - they read contemporary romance, trashy romance, regency romance, lesbian romance, paranormal romance....

"If you're using terms like "social capital" you've already passed some kind of brain test" so not worried about vandalism....

"magic" is problematic - Harry Potter mixed in with academic ones.
"leather" even more so

Can do tagmashes to get tagmash "France", "WWII", "fiction"

"chicklit" is now an LCSH but not geographically subdivided and will never have a "zombie" subdivision.

Tags: glbt vs lgbt "But those are the same thing!" -- but no: the books are actually different. The terms that people use encode all sorts of stuff. Many things labelled "homosexuality" actually mean "anti-homosexuality".

More than 1.5million covers added (including Albanian, Serbian editions of Harry Potter). When you upload it for yourself, everyone gets the benefit.

Social networking based on books you have in common. "Even if I don't want to be his buddy, checking out his library will be very interesting to me. Social networking for people who don't want to talk to each other."

Most popular group is Librarians who LibraryThing. Conversations about books on groups are tied into the books' own records.

LibraryThing Local - showing us map of bookstores and libraries in Portland, Maine. Can connect to local LT members; find events at bookstores, libraries. Add a photo of our libraries to these pages!

Example of wife's books - members have combined all the editions (other languages, etc) FRBR-style. Members have combined "Mark Twain" and "Samuel Clemens".

"Common Knowledge" - awards, quotes, characters and places in the story, blurbers - all sorts of things not captured in typical metadata.

Series pages - eg Star Wars series. Plus "related series". Much more information than any library has. Collated by people who know about it - the books' fans.

How many books does George Washington occur in? How many books take place in Washington, D.C., or in Hell?

LT "member" Thomas Jefferson, Marie Antoinette. No New Zealanders at the moment. Based on eg auction house records. Done by the group "I See Dead People's Books". Nice to be able to search Thomas Jefferson's library - couldn't do it before; now can see how you overlap with these people. Most popular book among all legacy libraries is Don Quixote; #2 is Complete Shakespeare.

Highest rung of ladder is altruism - flash mob cataloguing where volunteers go to library and catalogue their books in a mob in a day.

Six free ways to use LibraryThing:
1 Make sure you're in LT Local
2 Make an account
3 Libraries of Early New Zealand
4 Flash-mob catalogue your local historical society, church, health centre...
5 "Community library" to create a shared local library with LibraryThing Groups (eg two churches, a historical library, and a couple of people in town).
6 Grab our free data: common knowledge data, frbrised data etc.

One un-free way:
1 LibraryThing for Libraries eg at Seattle Public Library showing other editions and translations; similar books; tags; reviews. Four or five NZ libraries are using it.

What does social cataloguing mean for library cataloguing?
The end of the world! No!

Defends the value of structured metadata but that shouldn't be all we have.

LCSH - A book has 3-6 subjects - why? because that's how many we can fit on a card.
Subjects are equally valid because of... the card.
Subjects never change because of... the card.
Only librarians get to add subjects because of... the card.
Users don't get a say in how books are classified because of... the card.

In the digital world, none of this matters. In libraries these ideas have still persisted.

The physical library was human. The first wave of technology was dehumanising but social cataloguing can rehumanise the library. Everyone can help. (We don't need to let them do everything but they can help!) Local matters again. cf Māori Subject Headings - sometimes local communities need headers other communities don't have.

A note of caution before joining the exciting world of web2.0 - join the exciting world of web1.0! Library catalogues aren't web1.0. Often you can't link to library catalogue records; they're all session-based. Why why why? People need to be able to bookmark and share. And catalogues aren't indexed in search engines! Why?????

Go with the grain of the internet, not against it. We're not in competition with the internet. We should be open. Libraries are going the wrong way. LibraryThing gets twice as much traffic as WorldCat. Dogster gets as much traffic as WorldCat.

Be part of the conversation. Trust people: put your stuff online and risk that people might find the "wrong thing" or tag it the "wrong way".

Choose solutions that favour all this. He thinks open source is the way to go. He doesn't think open source is necessarily better, but it can be.

Social cataloguing can be a last chance to join web 1.0. Before we start struggling with ebooks struggle with the fact that people can't find our books on Google! It's an opportunity to reinvigorate library technology. To reconsider some LIS thinking and improve systems. (Had a LT project to replace Dewey. Turns out to be hard and didn't work. But it's cool to try!) Chance to embrace best traditions of librarianship: radical openness, public spirit, focus, connection to the local and social. Why would we lend books but hold back metadata?

Q: Could libraries organise own flash mobs and [? get stuff on web?]
A: Absolutely! Thinks flash mobs are good for things on the periphery, stuff that's never been exposed eg churches, historical society. So many books exist in private holdings!

Q: What proportion of books on LibraryThing do people catalogue themselves rather than pulling data in?
A: Not sure but probably a small percentage. Zines, comics, etc are the main things.