by Clare Playforth

Classification is a big part of my job and certainly the part that I find most enjoyable and challenging. The other day I was looking in vain for a resource that would provide me with a table of ‘date letters’ that are sometimes used when classifying collected works. After a couple of discussions about this it became clear to me that there are many practices in cataloguing and classification that might seem like they are needlessly complicated and opaque. I want to explain that there is a reason for things to be done this way.

Every time we make a decision about how to classify something we are doing it with the collection and ultimately the user in mind. When someone has looked up a book on the catalogue we want them to be able to find it easily on the shelf, but when they get to that shelf we also want them to find a load of other books that are relevant on either side i.e. we want to enable browsing. We want them to arrive at the shelf and first find the general books about a topic and then to be able to walk down that stack seeing how the subject narrows and becomes more specific as the classifications are expanded. In an ideal world we would want each classmark to represent only one book – this is part of the reason for doing the reclassification projects which you can read about here.

The argument against using letters to represent numbers (like date letters) and numbers to represent letters (like cutter numbers) could be that the user might find it more difficult to understand what the class mark means but I would say actually most library users don’t care what it means – they would like to just find their book. When it comes to systems, online and physical, the best ones are the ones that work for the user and deliver the desired result without that user having to understand the nuts and bolts of how they arrived there. This is my goal when I classify a book.

Anyone still reading? OK I’ll treat you to the process I go through when classifying… First I try to really understand what the book is about. This sometimes requires a bit of research, and will usually include establishing, not only the what, but also the who, the when and the where. Next I try to find other items in our collections that are similar – this part depends on a physical item search and on how existing items have been catalogued and classified. If I am classifying a new book to be added to our collection, as opposed to reclassifying an existing item, it is necessary to establish which scheme I will use to build my classmark. We use a combination of different classification schemes here at Sussex and so based on where this book’s best friends are shelved I will determine whether the scheme will be Library of Congress Classification (LCC), an adapted version of LCC, our locally devised Sussex classification scheme, or Dewey. I’ll work out if it’s about a particular period of history or a particular geographical area. I will decide whether to use a cutter number or a verbal extension, whether the book has a title main entry, for example if it is an edited book, or whether it has an author main entry, as this will have an effect on the classification. Once I have created my classmark I will look at where this will place my item in the collection – am I happy with the books either side being the bread in my book sandwich? Is my classmark unique and does it need to be? Have I left space for other items to be inserted into the sequence in future? All of these requirements need detailed and focused analytical work and whist it seems like such a small task, the identity of our collections is realised in the classmarks and every new book we purchase needs this attention.

So whilst it does appear on the surface that classifications are often needlessly complicated and not user friendly, it is in fact the opposite that is true. Every class mark is purposeful and considered and is designed with the user in mind. We want an author’s collected works to be shelved according to their date and we also want those collected works to not be split up if they have been published in a multi volume set. This is why the classmark may not actually show that the book was published in 1976 but I would say the user doesn’t need that information from the classmark, the user mostly needs the book to be in a sensible place on the shelf. The book is described accurately in the catalogue record and the main function of a classmark is to maintain a useful shelf sequence and not to be descriptive. I appreciate that browsing can be facilitated by having an understanding of the classifications but rather than simplify the classmarks at the expense of ‘known item’ discovery it makes more sense to signpost and provide access to resources about the classification schemes for those users that want this.

In suggesting that the user doesn’t always need to gain information about the book from the classmark I’m not advocating that we stop looking critically at our classification systems. The shelf order is a product of assumed hierarchies and value judgments made by the schemes original creators from what was arguably a position of privilege. There are areas of all the schemes that are outdated and in need of revision to reflect more modern and inclusive thought and these areas should of course be acknowledged and addressed but that’s a task for another (much longer and better researched) bit of writing. I suppose what I am trying to say is that although the complexity of classification systems can seem needlessly exclusionary I think we should live with that if the complexity serves a functional purpose that ultimately benefits the user.

Clare Playforth

Twitter: @clareplayforth