Thursday, May 04, 2006

LC Series Decision

American Libraries is taking a poll on the decision by the Library of Congress to stop creating series authority records.

Will the Library of Congress' decision to cease creating series authority records affect your library's technical or public service capability?

It does not seem that there is a good explanation, on the Web, written for our users that explains what this means to them. There are some very good posts in e-mail lists, but those require registration and login. There are some good explanations for librarians of what this means for our patrons. Something needs to be written for the general public that clearly and without jargon says what it means to them. No mention of MARC, 440 vs. 490, OPAC, or other terms like that. I may take it on but would not object if someone else found the time.


Anonymous said...

I'd certainly appreciate any commentaries or explanations you care to give or point to--I'm still a bit muddled, even as a librarian.

Anonymous said...

I had started a discussion, in plainer language, on Library Thing, via the comments (it starts here: I did mention the MARC 240 field, but only in the context of using MARC at all). It does get discussed in later posts on the Library Thing blog (, but I have to spend time looking for it. If I can find the discussion, and if you think this is valuable at all, I can send you the links.

Anonymous said...

Well, it's kind of confusing. Becuase the world of cataloging is kind of confusing in general these days, with, in my opinion, even some catalogers not neccesarily understanding the implications of what's going on. (Not to mention that cataloging is kind of broken in general these days in my opinion; but that's another topic; this LC decision will just break it worse.)

Some books exist in a series. [Someone else will have to fill in on what sorts of books this effects; I'm not sure if fiction series like 'Dragonriders of Pern' or whatever are actually currently cataloged in the 440; that's what most users will think of when you talk about series, but I'm not sure how this decision effects those. This is the biggest hole in my ability to explain what this means to users.]

So, catalogers have traditionally noted the name of the series in the catalog record, as it appears on the item. So far so good. But this is a 'human readable' thing, not really a machine readable thing.

What if you want to retrieve all the items from the catalog that belong to that series? Well, you can do a keyword search on the series name. That's exactly what LC is suggesting is all you need to do. But. There might be other series with very similar (or even the same!) names. A given series might also change it's name slightly (or just accidentally, as a typo, appear slightly differently than usual on a given book). All of these things will impede your ability to find everything in a given series by a keyword search. (Either becuase you want to know everything in the series, or becuase you're looking for a particular book, and you know it's in that series, but you can't remember exactly what it was called, etc.)

Also, we really want our catalog systems to be able to automatically assemble all the members of a series. We want the computer program to be able to tell what's in what series, for all sorts of keen purposes. [Sadly, most of these keen purposes are not actually implemented by today's systems, which makes this even more confusing to explain.]

The solution to all these things is making sure that there's a unique series name for every series, and that that unique series name appears on the records for all the items in the series. [If you are a computer geek like me, you can almost think of this as a 'foreign key' to the series record].

Now, we don't need to worry about the ambiguity or confusion in series names as written on the books themselves. Catalogers have said, dont' worry, we've provided this unique key on every record to unambiguously and clearly state which series a given item belongs to. Not only is that of use to a user searching for a series (dont' just have to use unreliable keyword search), but it let's the computer system easily identify what series exist and what their members are in an unambiguous way too. (Maybe you could subscribe to an RSS feed for newly published items in a given series; the system has got to know for sure whether a newly published item is a member of the series you care about!).

That's what LC is getting rid of---the catalogers method for making sure that each series can be unambiguously identified, and all the items belonging to that series can be unambiguously identified. That method is "series authority work"---making sure there is a unique string for every series.

Whew. Not sure if that's actually comprehensible to a non-cataloger (or cataloging geek like myself), but I gave it a stab.


Anonymous said...

PS: I believe that for certain scientific and engineering kinds of publication, series are frequently used, and users often come knowing hte name of the series, and need to use that to find the books they want (or might want). These users are going to be most impacted. But like I said, I'm not entirely clear on what classes of stuff this is most crucial for in acutal experience.

PPS: I _think_ another part of series authority work is saying "This series used to be known as X, but is now known as Y." [I think this is done for series, right? Someone correct me if I'm wrong]. You can see why this would also be important information for a user who knows about a series and is trying to use that to find an item of interest. That won't be done anymoe either.

PPPS: Yoko, that link you posted to blogger is broken, or I woudl post all this stuff there but I can't find where there is.