Tim Spalding wrote:
The rule of three came about during the catalog card era, and was maintained after, but your argument does hold up as far as it goes. In doing so, however, I think we can see that classification is done for a reason, and generally in the context of a specific--and rapidly obsolete--use case. It is so very easy to slip into the notion that classification is something more than that-something simply true and appropriate to every context and use.
The purpose of assigning a specific classification number to an item has not really ever been because it is more "true" or "appropriate" than any other number. Catalogers have known since Day 1 that most resources that are even halfway complex in their subjects can easily be placed in 3 or 4 places within the classification system they are using, plus catalogers are aware that there are many different classifications in the world, some of them quite different in how they arrange their subjects. The most widely used in the library world are the Dewey Decimal, the Library of Congress, the Universal Decimal, but there are many others, some of them used only in a single institution.
So, the purpose of assigning a classification number is to provide a level of predictability to an item, or in other words, a bit of standardization. The purpose of classification (as well as assignment of all headings) is to *bring the same things together*, so much as possible. Why?
Because otherwise, people will have to look around in different places for resources on the same topic. I went into more detail on this in an article I wrote on my former website several years ago and it is still available in the internet archive:
http://web.archive.org/web/20020406052541/http://www.princeton.edu/~jamesw/mdata/MetadataCreation.html (the images don't seem to have survived, as happens so often in the Archive, but maybe they aren't so important to understand the ideas there)
Therefore I do not assign a call number to an item because I "like" it or I feel it is "best"; people who do this often make huge mistakes (and I have seen this happening more and more as training of catalogers goes by the wayside), rather I must find similar items and assign the same numbers, whether I happen to like them or not. This same thing occurs for assigning controlled vocabulary (Rome vs. Roma, Tolstoy vs. Tolstoi, etc.). There are many headings I have assigned that I think are absolutely stupid, but since I am a cataloger, it is my job to assign those headings and not those that conform more to my own taste.
In terms of classification, the task is the same: to assign a controlled "relative location" within a specific collection, thereby making the resource more predictable for others to find--often this "other person" can only be another expert in the classification. Still, it is predictable.
Problems with assigning classification numbers have been known since the days of the Library of Alexandria. Many of these problems may not be relevant today with virtual materials and more powerful systems.
We should also realize and use the so-called "syndetic structure" buried in the arrangement of the subject headings. For example, "Love" in the LC subject headings is in the arrangement:
This final heading then has related terms for
Man-woman relationships, and
We see a "classification system" at work here. There are lots and lots and lots of other references and notes within all of this structure, and it seems to me as if this could be incredibly useful and highly popular if more advantage were made from it.
[BTW, the rule of three deals with assignment of name headings and not with subjects]