by Michael Smith, MLIS Day

If there was only one question that should be answered by the Knowledge Organization field, it should be to settle the ages-old question: how does one organize their CD collection?

I recognize that digital formats like mp3 with wonderfully adequate metadata schemes like ID3v2 make this question increasingly more moot, if for no other reason than I'm allowed to spontaneously re-organize my collection with the click of a mouse. This still doesn't solve the question of my multiple hundreds of compact discs that yearn for an organization system.

Alright, so let's assume that Berger Hjørland knows what he's talking about. Then the best first step with developing a knowledge organization system is to perform a domain analysis. Thankfully, this is actually kind of trivially easy. We already know what needs to be organized -- it's my own music collection. And it's possible for me to talk with the entire user group -- in my case, my very own self, but for others it might involve all the people in the household. So, we discover what the needs of the users are, and we think about what we've got in our collection.

In Jens-Erik's Classification Theory course, we inevitably got around to discussing this pointlessly futile question. The conclusion was that there is no method that in consistent with the current outlook on classification. So, you don't classify your CDs: you just put them in an order that is useful to you. Jens-Erik suggested the LIFO method of organization: simply pile up all of your CDs, and whenever you want to listen to something, pull from the bottom of the pile, since you haven't listened to that CD in a while. When you're finished listening to a CD, put it on top of the pile. Last In, First Out.

I kind of like it, but I think it's flawed: it doesn't provide for collocation, and doesn't help us find something specific when we really want it. For my part, if I want to listen to something poppy instead of something harsh and angry, a big pile of CDs requires me to do a lot of browsing of the giant pile just to find something. So I suggest combining LIFO with something that resembles the "Trivial Pursuit Classification Scheme."¹ It involves creating for yourself a small taxonomy. I will embarrass myself and show you a small sample of my combined schedule²:

I realize that not all of those choices make logical epistemological sense. I have a feeling that many would take umbrage with my choice to make Rock a narrower term of Pop, but I made the choice that makes sense for the domain's users -- you know, me. For my collection of CDs, Pop should carry the following: DEF Any music that might possibly be heard on mainstream radio. It's popular music, not just Michael Jackson and Madonna.

Now there are six piles. I make some choices about which CD goes in which pile, and now I have six piles of CDs. Now, when I want to find a particular CD, I only have to browse (on average) one sixth of my collection instead of the whole thing, I'm constantly changing up what I'm listening to, and now I can make a particular choice about what kind of music I want to listen to.

Of course, you can always just go with the traditional "by accession" organization or "alphabetically by artist." But those seem a little too ordinary for vibrant, interesting, cosmopolitan people like us.

¹ There may be some of you that don't know about the Trivial Pursuit Classification Scheme for Home Collection Collocation (TPC for short). The idea is: using something like Dewey Decimal Classification for your personal book collection isn't very useful. So, instead, you classify your books by the Trivial Pursuit question categories: Arts and Entertainment, Sports and Leisure, &c.

² This schedule didn't exist until just now. It's hypothetical. I created it just for the purpose of this article. Seriously, I have not previously created my own personal taxonomy for my music collection.