I’ve been working with a lot of MARC records lately, from a handful of different libraries, and a handful of different library systems. Over the next few months, I’ll be working with MARC records from approximately 80 different libraries. And what is really striking to me, as someone who is fairly new to libraryland, is how different all these MARC records are. Isn’t this supposed to be a standard?
Now, I’m not talking about the content of tags, and whether things are abbreviated properly, or where the commas are (although these things are problematic, too). The conversation about RDA versus AARC2 isn’t the one I want to have right now. I’m talking purely about where in the record various pieces of information can expect to be found: the structural stuff.
MARC is a hugely complicated structural standard for bibliographic data. When I first started dealing with MARC I thought it was overly complicated, but I’ve since come to understand that bibliographic data requires some complexity. However, when you’re working with a huge and multifaceted standard, it’s kind of important that you get it right. There are tons and tons of fields potentially involved in a single record, and in order for libraries to work effectively with each other, the records have to be interoperable, and the same data should really be in the same fields. Sadly, this is not the case.
I don’t blame catalogers for this, at least not entirely. Library systems are the major culprit here, putting system control numbers and OCLC numbers and library holdings into any darn field they want on export. Library systems take huge advantage of those locally-defined fields and they stuff all kinds of system information into them, without offering any explanation of what is going where. Half the time, systems librarians and catalogers don’t even fully understand what their library systems are exporting and how. What is the purpose of having a standard if there is so much leeway that the systems in which it works can do whatever they want with it?
Catalogers are a little bit to blame, as well. When you get a set of records and find that only half of the numbers in the fields designated for Library of Congress Control Numbers or ISSNs are, in fact, those numbers, that data has quickly become less useful. And I don’t even want to get started on serials holdings, because that is one royal mess.
We’re in the process of creating a “standard” MARC record for our project, and will have to create individual converter programs for each library with which we work in order to convert their record to “our” standard. This seems like an enormously frustrating waste to a newbie like myself, who can’t help but think that if MARC was actually treated as the standard it is supposed to be, we wouldn’t be in this mess. Which is why I, for one, cheered when I read that the Library of Congress is going to start discussions about doing away with MARC.
It’s time for us to start thinking about doing this work better. Interoperability should be at the forefront of our minds as we come up with new standards for storage, encoding, and transmission of bibliographic data. And it would be nice if we could move a little faster than usual on this one. I’d prefer not to have to wait 10 years to see library data that actually works.