On the Impossibility of Classification

Last week was an exciting week for me, cataloging-wise, as the library where I work just received our subscriptions to Cataloger’s Desktop and Classification Web. In addition, we wrapped up training on our integrated library system, Millennium. I also finished up the new book of Alberto Manguel’s compiling his writings, A Reader on Reading. I don’t think there is a more erudite and pleasant author who writes on books and libraries than Manguel. Of course, little did I know that the Classification Web subscription and finishing the Manguel book would end up in the same busy week, prompting me to write this short discourse.

One of my charges as a librarian with expertise (such as it is) in cataloging is the classification of the books into some overarching and coherent schema of knowledge. This concept of classification is fraught with problems which the “classifier” must accept to seriously attempt to place an item into a scheme of knowledge. The process is highly subjective - should an exhibition catalog, for example, be classified with works by the artist, or with works at that geographical location, and so on. On the shelf, the classification of these items seems so definitive when, in reality, classification is quite subjective. To be of any use, any scheme must be (or at least present itself as) a way to organize all of the knowledge that has, does, or will exist. This is, of course, an impossibility - there can be no schema that represents unknown knowledge. Instead, creators of these schema must look to accommodate what we have yet to learn or discover, so that this new knowledge will find an appropriate and logical place in the schema when it is classified.

Why, then, do librarians attempt to undertake this seemingly futile task? The answer is in two parts - to increase access and browse-ability. Without a “call number” (or a shelflist number, if you prefer), you can imagine how difficult it would be to find a title or item you desire to use on the shelves of a library of any size. Would they be sorted alphabetically by title, or author? This would preclude one knowing one or the other of those facts about a book, which is often unknown by the searcher. A simple alphanumeric tag greatly assists in finding the item in the library. Another benefit of classifying items in a library is that (ostensibly) it aids in discovery of different titles by browsing. In a perfect world all of the titles a person is searching for would be shelved next to one another so that when you found one book, you would find them all. As you read earlier, this is a bit of an impossibility, but we librarians do try our best.

Of course, it all looks so simple from the finished product of this Sisyphean task. Indeed, many people do not even realize that the call numbers one finds on the items in any sizable library are actually an alphanumeric representation of the item’s (and the knowledge it contains) place in a scheme for classification. The one we use where I work is the Library of Congress Classification, which is commonly used in academic libraries. Here’s an example:



It is far more likely you have interacted with the Dewey Decimal Classification, which looks more like this:



Rather than trying to make some sweeping final statement about the impossibility and necessity of classification in the library, let’s just let a far better writer than I (Alberto Manguel, from the book mentioned earlier) do that heavy lifting:

In the middle of the third century B.C., the Cyrene poet Callimachus undertook the task of cataloguing the half-million volumes housed in the famous Library of Alexandria. The task was prodigious, not only because of the number of books to be inspected, dusted, and shelved, but because it entailed the conception of a literary order that was supposed somehow to reflect the vaster order of the universe. In attributing a certain book to a certain shelf - Homer to “Poetry” or Herodotus to “History” for example - Callimachus had first to determine that all writing could be divided into a specific number of categories, or, as he called them, pinakes, “tables”; and then he had to decide to which category each of the thousands of unlabeled books belonged. Callimachus divided the colossal library into eight tables, which were to contain every possible fact, conjecture, thought, imagination ever scrawled on a sheet of papyrus; future librarians would multiply this modest number to infinity. Jorge Luis Borges recalled that in the numeric system of the Institut Bibliographique in Brussels, number 231 corresponded to God.

No reader who has ever derived pleasure from a book has much confidence in these cataloguing methods. Subject indexes, literary genres, schools of thought and style, literatures classified by nationality or race, chronological compendiums, and thematic anthologies suggest to the reader merely one of a multitude of points of view, none comprehensive, none even grazing the breadth and depth of a mysterious piece of writing. Books refuse to sit quietly on shelves: Gulliver’s Travels jumps from “Chronicles” to “Social Satire” to “Children’s Literature” and will not be faithful to any of these labels. (p. 26-27)