Classification and languages

Can I ask a dumb question about classification? It seems to me that all commonly used systems and the EC separate out books by language—chiefly or entirely fiction, it seems.

I can see obvious reasons for doing so. It might not be fun for everyone to browse a section with stuff in it they can't read. And, if you spoke Swedish, it might be useful to enter an American library and quickly check out what novels you could grab.

At the same time, it's not a perfect system as the fact that it only applies to fiction demonstrates. Browse Greek History in LCC and you'll find German, French and English books mixed up together. I guess this makes some sense.

I raise this not only because I'd like to hear someone on the reality and intellectual basis of the reality, but because, as many know, LibraryThing's "works" system is language-independent.

If any system—Cutter included—is going to deal with LibraryThing, works have to be the starting point. (Yes, if we get down far enough, editions and so forth will have to be added on, but works should share most of their call number otherwise.) Languages can get in the way of that.

Thoughts?

Tim,

Do you have a pointer to some place where you have discussed your concept of "works" in some detail? Your questions can be puzzling to me since most libraries don't begin there.

Another distinction that could be useful to the discussion--what is your model? Academic and public libraries, not to mention special or school, etc., are very different.

In the classification system that I am building/writing for my own books language and subject are two separate fields so that a library could conceivably separate their books by language or by subject.

Language is a three letter field and subject is a five digit numerical field.

Hi. The LibraryThing concept of works is explained here. It's pretty close to the FRBR concept of works, although it was developed separately and is user-driven (FRBR *could* be user-driven, but no implementation is).

I want to question the notion that academic and public libraries need a separate classification system. Of course all libraries are different, but that's true within each type as well, across time, for each user, etc. There is no such thing as a perfect system for anyone. But a system can be equally good (or bad) for the major types. Cutter, for example, was embraced by all types—the high-tide was 42% public, 26% academic, 14% government, 14% athenaea, and 4% theological. The libraries that abandoned it (all but four) went the predictable directions—public to DDC, academic to LC.

One feature of Cutter that's very attractive here are its seven levels (or "expansions") published one after another. Each went a level deeper. A public library might choose to go to one level, an academic library all the way to the complex end. This is similar to how you can "round" DDC*, if somewhat more elegant. As someone pointed out to me, rounding of DDCs mean that copy-cataloging DDC call numbers is pretty useless.

I'd like to see a system that suits all your types as well as personal libraries of persnickety people, and bookstores. Yes, bookstores. They too shelve books.

I'm a dreamer, perhaps.

Anyway, I had a long talk with Winke, who may be the top expert on Cutter. I'll summarize my findings and my thoughts in another post, when I get a chance.

*Technical term anyone?

I just deleted my above post about consideration of other classification systems, because i realize it veered off-topic from this particular thread. Thus i have started a new thread and reposted the deleted message there.

>1 timspalding:

The only reason I'm aware of for shelving books by language is for the convenience of the user, the patron of the library. Many public libraries include non-fiction as well. Foreign language collections, fiction and non-fiction, are often shelved together, separately from English in U.S. public libraries. Translations into English are shelved with the English language titles. Users ask for materials that way. They want novels written in Spanish, philosophy written in German. Some of it may have stemmed from the practice of shelving materials for children and young people separately, but that is only conjecture. It is a convenience based on physical location. Similar books are close together. It saves users time.

These collections are labeled for patrons on the bibliographic record screen of library catalogs. There each item, usually one copy, that is attached to the bibliographic record, is described by location/collection, call number and status (checked in, on shelf, due date, etc). As long as there is a location/collection code in the catalog for an area, a record can describe it.

It is also more likely in libraries because the different language versions have separate bibliographic records. And the increasing number of formats, the different shelving needs for those, also are breaking up the pattern of one shelving location.

"In the classification system that I am building/writing for my own books language and subject are two separate fields so that a library could conceivably separate their books by language or by subject."

Actually, that sounds like a pretty good idea. If we have a (say) 2-part classification number where the first is a topical classification and the second a number to identify the book (usually an author Cutter Number).

Now, if we replace that second number with one that identifies the language (we could use the Marc codes for that purpose, as they are already encoded in), we get topical books further organized by languages.

We could have a second number for authors, and by allowing the users to choose the ordering, they can get a classification by languages, then author, then topic. Or by topic, author, and language (keeping all the Agatha Christie alternate languages together.)

I think Cutter had a two-digit code for the language/country which could be put before or after the class mark depending on how people wanted to arrange their libraries.

Making up numbers:

YF|34 = literature|English
B|34 = philosophy|English
34|B is the same thing, but if you wanted to arrange all the English books together.

Cutter recognised early on the importance of flexibility (within an orderly framework to avoid chaos)

Actually the same thing can be achieved with other classifications (Dewey, or other) by using them with lanuage abbreviations which can be put ahead of or after the class code. The MARC language codes would be good for this as they can be generated automatically.

The language codes represent the language of the actual item at hand. Dewey and other classifications give class codes for the original language in which the work was created, or to put it another way, the source culture.

E.g., Dewey 820s is English literature, whether in the original language or in translation. Likewise Dewey 840s is French literature, whether in the original language or in translation.

EN would be items in the English language whether they were translations or originally in that language -- ditto FR for French.

So, as in the example of shmjay's post, you could put the 2 codes in order in whichever order you preferred depending on whether you wanted to group items first by language of item or by subject (or original lang.)

FR 629.2887 Motorcycle repair manuals in French language
FR 823 English fiction in French translation
FR 843 French fiction in the original language

Or the FR could go after the class no. and before the Author name or Cutter no. to group things by suject topic, and then within the topic by lang.

Libraries already use letter prefixes all the time, like J for juvenile/childrens and R for reference.