I have been brooding on this for a while but have been nudged into expressing it by Karen Coyle's post. I can't articulate, let alone understand, the argument as well as Karen can; but I think she strikes the nail on the head when she says that we have forgotten about classification and about the importance of subject retrieval.
When, and why, did we suddenly get ashamed of classification? Most library users don't understand how classification works; a majority of librarians don't understand how classification works; but when did we get to the point where even a majority of cataloguers don't understand how classification works? Many cataloguers think that classification is all about shelf arrangement - that classification is only the number on the spine label.
Keywords just don't do it and never will. Do a keyword search for "pain" in our catalogue and you'll find some books about pain and pain relief; a number of novels which happen to have "pain" in the title; and a book in French about bread. And yet somehow, and largely I suspect because of Google and its ilk, we have come to think that's OK.
In libraries we will never be able to out-Google Google because we don't have the enormous resources that it would take, but that's not the problem. Why do we want to follow Google instead of doing something better?
Somewhere along the road we seem to have decided that users don't understand classification and therefore we shouldn't use it. Now there is absolutely no reason for users to understand classification, any more than there is for every driver to understand the mechanics of the internal combustion engine, or for every concert-goer to be a expert music theorist. Our job as cataloguers is to provide the engine, plus just enough guidance to enable our library users to get the best out of it. And anyway, people do understand classification otherwise they would never be able to find their way around the supermarket - almost every part of our public and private world is organised in some systematic way.
Classification isn't elitist, it is an absolutely essential tool for organising knowledge. We should be proud of it, not ashamed of it.
The problem is that classification is largely done badly at great expense.
ReplyDeleteMatters aren't helped by leaving classification to individuals who seem to have their own agenda — the chestnut of local flavour.
I work in a collective of over 100 academic libraries that has one LMS infrastructure and it's simple to see that the value added by local classification is basically none. This seems harsh, but the point is that as terms (and here I mean both terms from controlled vocabularies, free-text keywords and classification numbers from the various systems in use) are added, the item becomes less findable because the terms are just noise.
This isn't entirely the classifier's fault, because the usefulness of the terms is not immediately apparent — so they (to some extent at least) try to second-guess the programming logic of the OPAC and users' behaviour. If the OPAC had provided a logical interface for using terms, then we might see different patterns in the behaviours of classifiers and users.
A final thought is that we're talking about systems that have no effective crosswalks; these systems don't in fact relate to one-another at all. If we had, say, a simple way of applying extended logic to a term, not only within the system that the term comes from, but also other related systems, we could provide something a bit more useful, and a little bit stunning.
There's hope. I hope.
I gave a talk to academic library staff earlier this year on the topic of classifying web documents using Linked Open Data. The talk argued for a faceted subject classification approach to web docs. One of the points I made was that web users are accustomed to this approach, even if they aren't (fully) aware of it as faceted search forms the basis of many big-name online retail websites. If the mass market is using sophisticated subject searching/browsing techniques on a daily basis in many aspects of everyday life then I don't think libraries should be concerned about introducing similar tools.
ReplyDeleteThis is not to say that there are not lots of sophisticated techniques deployed in libraries right now. It is more to point out that the audience is already up to speed and possibly ahead of the game when it comes to faceted subject searching.
Love this post.
ReplyDeleteI had a thought recently, sort of on the back of mentions of 'serendipty' of discovery in the physical library (discussed at libcampuk11 and written up here and here): what if all or some of this perceived 'serendipty' is nothing of the kind, but is instead the result of careful classification? It looks to the reader as if finding another really useful book next to one they were looking for is chance, luck, good fortune, but actually they were near each other as a result of classification...
Not always the case, but possibly often true?
About time someone spoke up for classification. Furthermore, keywords soon become out of date - I noticed this several years ago when looking round the selves of Surrey County Library. They had to re-do their key words and re-shelve their books every so often.
ReplyDeleteYou learn to classify in the womb or, at the very least, at the moment of birth. Your first unconscious system has two classes: "this feels good" and "this is nasty/bad".
ReplyDeleteGet everyone re-designing the supermarket to have all the tomato products together etc (which I did as an exercise on a workshop ages ago) and discover that tomato is NOT a good top term - soup, bread, fresh fruit/veg, sauce, ketchup are better!
I could go on for ever! And I'm not even a librarian just a logical organiser of information.
Many thanks to all for their comments - I seem to have struck a nerve!
ReplyDelete"classification is largely done badly at great expense" - hmmm. I agree that classification is expensive, and unjustifiably so if it can't be, or isn't, used. Badly done? Sometimes, no doubt - and we may be back to the argument about lack of skills and lack of understanding amongst cataloguers. Badly done because it takes account of local practice? Well, if it is unquestioned following of local weirdnesses, the reasons for which are long forgotten, then yes, that's bad classification. If it is an attempt to take account of local needs and the nature of local collections, then that's good classification. If I have a collection with strength and depth in particular subjects, I'll want to classify it with greater detail (or everything ends up in one undifferentiated place).
The examples from retail are interesting. What I wonder is, are they allocating keywords to products (blue, shoe, ladies', size 4) or are they allocating codes which can be translated in English for an English-speaking market, but also in French or Spanish for those markets? Then you get very close to classification (code) lying underneath, linked to subject headings for output - as Dewey/LCSH already can be.
One of the great advantage of classification is that it is language-neutral, so you can change the keyword without changing the classification.
I can feel another post coming on - does anyone still use a subject index and if not, why not?
I defend local practice, as to not accord it importance in cataloguing is to ignore that you and your library is not the centre of the universe. Spoken as a Canadian who edits constantly to create a collection that makes sense to its young users.
ReplyDelete