Monday, 19 July 2010


Another one of those sites that I thought sounded great, signed up for, and then forgot all about.  I think the whole idea behind LibraryThing is great, and surely every book club should be on it.

I have never looked at it from the point of view of libraries themselves.  I can see how valuable it would be for cataloguing a small library collection, but I wonder how easy it is to export the records - anybody know?  You wouldn't want to spend all that time cataloguing your holdings only for LibraryThing to change/merge/fold altogether!

Wenzeler's article was very interesting and I really like the idea of enhancing existing library catalogues with user-generated content from LibraryThing.  I believe AquaBrowser, when it comes to Cambridge, will create a shinier, happier user experience anyway though, so do we need LibraryThing?


  1. Hey, if you don't mind, here's a quick comment from the LibraryThing founder.

    LibraryThing isn't "for" libraries, but for personal collections. Sometimes a very very small library will use us as their OPAC, but these are mostly churches and other such organizations with small collections and which stand to benefit more from the overtly social nature of the site. Also sometimes a library will use it to catalog some special set of books—like new books—in order to make a widget. That too is fine, but it's not what the site is for.

    LibraryThing is, however, interesting to libraries because it showcases some concepts that will, I think, be increasingly important, especially user-contributed content. Here, for example, is part of a talk I did, on why tags are important:

    Incidentally, Aquabrowser has an option to include LibraryThing tags. I don't know if Cambridge is getting it, but the point is that tagging is actually pretty useless without lots of content. When you have 3 tags on a book, you have pretty much nothing. You need dozens, hundreds or thousands before the patterns emerge which turn tagging from shitty categorization by random people into something more like a consensus classification. LibraryThing, which has 64 million tags, provides the sort of depth necessary to make tags work as a discovery tool.

    This sort of depth is not achievable by libraries, both because they're all separate and because people aren't motivated to add tags in an OPAC—the timing is all wrong, and the payoff is weak. For example, largest library tagging project yet undertaken—PennTags—has accumulated in five years fewer tags than LibraryThing adds every day.

    Tim Spalding
    Founder, LibraryThing

  2. Hi Tim,

    Thank you so much for commenting on my blog. You've highlighted one of my major skepticisms about user-generated tags and reviews on library catalogues - if there aren't any/enough it just looks like we're trying too hard to be cool and failing miserably!

    I am aware that LibraryThing is really for small collections and think it serves this purpose very well, but as part of 23Things in Cambridge we're considering how appropriate various tools will be for our specific library. My answer = not very!

    I don't know much about Aquabrowser, beyond a vague rumour that it's coming to Cambridge and a quick check of how it works on other sites (e.g. above). It's very interesting to hear that Aquabrowser can incorporate LibraryThing tags - if this is the case it could make a huge difference to the catalogue.


  3. Thanks for not minding when some random guy jumps in to comment on your blog :)

    So, I agree with you on trying to be cool, and failing. There is *some* use for tags as a simple list-making device. Library catalogs don't make lists very well, and have no ability to share them. So tagging can be an easy way to keep track of the books you're going to use in a paper, or for a professor to keep a bibliography for a class.

    Still, I don't think they work without large numbers. I do think they'd work in Cambridge for that reason. Although Cambridge obviously has a lot of obscure books, the coverage should still be pretty good. (Even a top academic library sees far more action on the better-known academic books than the rarities.) For some examples of LibraryThing tags in academic catalogs, see our libraries here:

    I'd also say that, in an academic context, while having LOTS of tags is idea, having a decent number is still quite acceptable. People motivated to tag a book on, say, Hellenistic Ruler Cult, have already passed some sort of "are you an idiot" test. So you don't get the wild tagging you sometimes get on a YA novel.

    Take, for example, this fairly obscure 1978 book about the logistics of Alexander the Great's campaigns, in the catalog of Bowdoin College, the "Harvard of the North."

    It's been tagged about 200 times, and the tags are all good. So too are the recommendations. That's very good user-content for a book that's pretty obscure.


  4. An interesting debate, covering lots of the things that have been discussed during 23 Things here.

    My understanding is that Aquabrowser in Cambridge will allow user tagging, through creation of a separate account so it will be interesting to see that in action. I do definitely take the point about needing a critical mass. I was however quite amused by the "have already passed some sort of 'are you an idiot' test".


  5. Aquabrowser will allow tagging through the creation of a separate account? I know it's early days, and you probably don't have many details, but do you mean different to the current UL card/University Card login? I'd imagine that needing a separate account would restrict the numbers that would bother to tag.

  6. @Tim: Interesting that you think Cambridge would get enough people tagging to make it worthwhile, but considering the size of the collection, there would still surely be many items that are used so rarely that there would be no tags added? I wonder if the solution would be some kind of hybrid, mixing the "authoritative" subject headings with the user generated tags to create the tag cloud?

    While you're here, maybe you could answer the question about exporting collection records for back-up purposes - do LibraryThing or LibraryThing for Libraries facilitate this?

    @Celine: Is there an intro anywhere to what exactly is planned for Aquabrowser, or is there anyone in particular that could explain more fully what the plan is?

    @Katie: I totally agree, who'd bother to tag if they needed yet another account in order to access it? But to come back to Tim's implication that external users would be likely to be interested in tagging Cambridge's collection, how could they do that if it was restricted by the university login systems?

  7. On Aquabrowser in Cambridge - they are implementing My Discoveries (not LTFL, sorry Tim) which basically seems to do what LTFL does and allows users to tag, rate and review, and save lists.

    I think the idea of a separate account is partly to allow non-students to use the service too, so that all potential users of the catalogue and collections have access to the feature. In the blog I link to below, there's an example from Chicago University Library which you can actually try out because you don't have to be a Chicago student/staff member to get a My Discoveries account.

    Niamh - the FAQ on Aquabrowser were posted on the lib@cam service developments blog in this post, but not sure if you'll be able to access it. It's Raven password controlled and accessible to UL staff and staff in lib@cam libraries. Given our conversation the other night, that may or may not include you.

    I should add I've not been involved at all in the Aquabrowser development but obviously have a huge interest in knowing what's going on, given my day job.

  8. Thanks for the clarification Celine. You guessed right, I don't have access to that site with my Raven "permissions" (if you can call them that!) I really do need to get that fixed. I'll see if I can find my way in when I'm in work on Friday!

  9. I see - if My Discoveries allows non-Cam users to tag (and will possibly allow people with an account to tag in other catalogues, too), then it makes more sense.

    I didn't know (probably wasn't listening at the right time) that there's a lib@cam blog. Thanks for the link, Celine.

  10. MyDiscoveries includes LibraryThing tags if you opt for it. Some libraries do, and have tags. Others don't and giant tumbleweeds blow across the tag fields.

  11. Just to clarify I don't think Cambridge will have a tagging critical mass. No libraries does. As noted, PennTags has added fewer tags in five years than LibraryThing does every day. I don't see this math changing.

    I *do* think that Cambridge's collection would work with LTFL tags. We have a lot, and they aren't just on popular stuff. Obviously, the larger you collection the longer the tail. Lots of books won't have tags, but they will mostly be the obscure ones.

  12. Got into that blog now Celine, very interesting reading and I'm looking forward to seeing how it will work in practice.

    Tim, that's what I thought you meant originally and was confused by your second post. All becomes clear! I would be interested in knowing whether our implementation of Aquabrowser will incorporate LTFL tags or something else that prevents that feature looking empty from the start.