Not the LibraryThing I signed up for
October 16, 2007 – 7:22 pm
I started using LibraryThing because I liked its tagging power (for browsing and searching). I also liked its less-commercial book pages, fondness for and willingness to work with libraries, and above-average books reviews from LT members, and book-based discussions (easy to find and see updates on discussions of one’s favorite books/authors). But mainly, I liked their approach to tagging. View one’s own tag cloud, author cloud, tag mirror, and community tags for books! Play with tag combinations to find very specific books! This last feature I especially like, because it combines the metadata from individual taggers to allow searches they might never have thought of trying to foster. One person might tag a book by what interests them about the author, another by something about the characters, another by the locale; the sky’s the limit. If enough people are tagging, useful and interesting information will bubble up, and the noise won’t matter — and what’s noise to one person might still be useful to another, and it’s all there.
Now, however, LibraryThing is rolling out its new Common Knowledge feature which uses fielded wiki technology. I don’t know how much effort this will take away from tagging by LT members; I intend to ignore this as much as possible (not easy, considering it’s on the book and author pages). Other people have written about issues with specific fields (e.g., binary thinking above does not work for some authors/collaborations). I know some people like it, and I know some of the issues may be smoothed out. However, the idea of one person at LibraryThing getting to decide what are valid values to allow people to put in for some of these fields seems like a step back to me. I’m afraid it will combine the worst of both worlds: rigid (”expert” knowledge determining categories/values, not necessarily so expert/authoritative in all cases, and prone to being outdated/inflexible) and error-prone (at least Wikipedia facilitates footnotes/cites/references for factual claims, and makes it easy to discuss potential changes with a discussion page per article … and offers RSS for both articles and discussion pages).
[Edit: removed scare quotes (see comments) and an unconstructive aside]
[Update! I was happy to notice in 2009 that LibraryThing had fixed the dropdown depicted above.]
9 Responses to “Not the LibraryThing I signed up for”
Hey. I saw this on Google blog search and thought I’d reply. If you consider it inappropriate, please delete it. I’d ask you through LT, but I don’t know your user name.
It is a feature, not a quote-unquote-feature. I have no problem calling this a blog, not a blog.
I mention this because it’s a nice demonstration of how categorization can be used in the place of argument. If I call you a “writer” (in quotes) I’ve simultaneously attacked and avoided having to justify myself.
The ability for categories and classifications to be “used” is a serious one. Clay Shiry calls this “voodoo ontology,” when your system forces your ideas on someone. Sometimes this is overt and intentional (eg., arguments in Congress of what race boxes to allow on the census form), sometimes the voodoo only takes hold later. It’s hard to tell which came first with the Dewey Decimal System, but it’s certainly the greatest example. Christianity gets 890 times as much space as Buddhism. Islam gets lumped in with a number of religions most Muslims regard with great suspicion. Portuguese is a special case of Spanish, women’s education a special case of education.
As you say, tagging gets beyond this. Where LCSH is binary and professes an objectivity we can only question, tagging can express nuance and embraces something like multi-subjectivity. I go on at some length in my Library of Congress talk about this very issue, mentioning the subtle but important differences between “gay,” “queer,” “LGBT” and “GLBT.” In a weird sort of irony, the “homosexual” tag–which in the LCSH world is the simple unmarked expression–is, on LibraryThing, strongly identified with ANTI-homosexual literature.
I think this problem is basic to classification. You can try to reach compromises, but, in the end, someone has to win.
In the case of gender, you should note that I raised the issue in the initial blog post, and we have been in a deep discussion about it ever since. That we haven’t resolved it says more about our deep love to discussion. (I might add that discussion has been pretty weak so far, with much more effort going into place names and such. Whether or not gender should be expanded or free-form, most users just don’t care. I would encourage you to post there, even the link to this blog.)
So, in a sense, I think it’s a step back too. It’s a step forward insofar as a more rigid system does have some advantages. By enforcing one field (or a field which can become more than one) we force users to work together.
Sometimes that’s a bad thing, but it can also be good. Wikipedia–which you mention–is an example of how it can work. Wikipedia is user-contributed, but it does not simultaneously reflect all of its users opinions. It could fork, but it mostly doesn’t. Things move toward consensus.
Sometimes the advantages of classification and standards are won at minimal expense. The community is now working out rules for places. Without a unified system, people would use any old expression they wanted. By forcing users to come to agreement, we can ensure that Texas isn’t split between “Texas,” “Tex.” and “TX.”
Incidentally, your point about discussion is well taken. In part, I think the community is small enough and engaged enough that the Common Knowledge group can do the job. But you have to be checking it. One possibility–which just came to me, so thank you for prompting this–would be to allow people to “touchstone” a work or an author in a way that specifically indicates CK is being discussed. That way we could put in the CK area a message like “The Common Knowledge relating to this author are under discussion; click here to read this discussion.”
Oddly enough, we actually DO have a comments feature, which would be an ideal place for footnotes. (Note however that Wikipedia’s interest in citation is relatiely recent.) Every edit can get a comment. We haven’t exposed this because, having built it in, we didn’t see an elegant and unobtrusive way to expose this in the user interface.
Unfortunately, all the options for exposing comments, discussion and etc. would require more space, which you’ve already come out against. Physical space–or virtual physical space–like traditional classification, is something someone has to win on.
Thanks for the post. Do you mind if I snag the image from your blog for mine?
By Tim on Oct 18, 2007
Tim, thank you for your thoughtful reply to my bit of spleen. I hope to post a longer reply, or another blog entry, tomorrow.
“I mention this because it’s a nice demonstration of how categorization can be used in the place of argument. If I call you a “writer” (in quotes) I’ve simultaneously attacked and avoided having to justify myself.”
You are right. It’s not a feature I wanted, but that doesn’t invalidate its feature-ness; obviously some people (beyond LT staff) are excited by it. I apologize, and I have edited my post to change “feature” to feature.
You’re welcome to the image.
By configures on Oct 18, 2007
Sorry. That was snarky of me to notice, but I felt it tied into the point: categories are weapons.
Thanks for the picture. I’m going to show it at a talk I’m giving at a conference in Manhattan on Monday :)
By Tim on Oct 18, 2007
Oh, no, it was a fair cop. I’d rather be called on something like that so I can correct it.
“Categories are weapons” — sounds like a good title!
I’m glad if my picture is useful for your talk. Even if it turns out to be about Users Who Don’t Get It, or something like that. :-) Really, there’s a lot I like about LT; I wouldn’t have taken time to blog about it or create that image if not.
By configures on Oct 18, 2007
“Genders and Drop-down Menus” is a more recent take on the sort of issue that sparked this post (which LibraryThing fixed, I’m happy to say). In it, Sarah Dopp explains the influence issue of binary gender categorization. As I’ve said in letters to the ACM before they fixed their member survey (thanks!), my college Honors Department, and others, a poll (or drop-down menu) which only allows “male” and “female” defines people who don’t fit into binary gender categories out of existence, and reinforces rigid binary gender thinking which harms real human beings. An “Other”, “Complicated”, or “Prefer not to say” option (or some mix of these, or all on one line) would encourage participation from people who just don’t fit so neatly, and would help remind people that it’s a big world with lots of kind of people in it.
By configures on Jan 30, 2010