In response to a post about Metis on the ALSC blog, some commenters argued that we were mistaken in being concerned about our students understanding the way that Dewey worked, and that all that mattered essentially was that one could search for items in the online catalog, obtain the call numbers and then locate the items on the shelves. That got me thinking. Most of our work in this area has been in relation to elementary age children, but these statements seemed to go beyond that to the heart of the subject classification process itself.
As my colleague Tali pointed out, if all you need from shelf order is a code that you can locate on the shelf, what is the point of an elaborate classification system? One might as well put books in accession number order. It would certainly save a whole lot of trouble.
There must be a reason why subject order was chosen by Dewey to organize the physical collection. Amazingly enough, I found the first edition of Dewey (1876) on Project Gutenberg. I thought he might explain why he chose subject order, but in his introduction he seems concerned mainly with the internal order of his system and the use of decimals. It seems as though the prevailing system at the time was one in which there was a fixed location on the shelves for each item, which would not change as the collection grew. One of Dewey's innovations was relative location, in which items were placed in relation to each other and might move to different shelves as the collection grew. He discusses these issues, but more or less assumes the validity of subject order.
So what is the point of a shelving system arranged by subject? A.C. Foskett, whose textbook (The Subject Approach to Information) I used in library school, has this to say:
"There will be many occasions when readers will approach the collection without any particular need in mind but wishing instead to be able to select items at random. To help in this situation, our system should permit browsing; a reader should be able to follow a casual train of thought as well as a planned search.” (London: Library Association, 1996; 5th ed., pg. 26)
Later in the same book he asks, "Why is classified shelf arrangement helpful?" and answers: "There are two important reasons. The first is to satisfy the browsing function; readers like to wander round the shelves and find books which attract them." (pg. 213)
Similarly, Robert M. Losee of the UNC Chapel Hill Library School writes that "the classification system essentially supplies a browsing path from one documents to another." (“How to study classification systems and their appropriateness for individual institutions.” In Classification: Options and Opportunities, ed. Alan R. Thomas. Binghamton, NY: Haworth, 1995, pg. 45)
So if we acknowledge that the reason for a subject-based shelving system is to permit browsing, we then surely have to ask the next question:
What kind of subject-based classification system is going to encourage the users of my library to browse productively so that they are led to books which are of interest to them? We think that this is a question worth asking and trying to answer, rather than submitting to the monopolistic dominance of Dewey as a supposed one-size-fits-all system.
The question puts users at the center of the issue, and keeps them there. Ultimately, that is the essential thing.