how do name authorities work anyway?

Question for the catalogers out there.

I thought to go look and see if I had an LC authority record. I don’t think I’ve written anything that would end up cataloged by LC, so I doubted it.

But, see, I’ve known for a few years now that there’s this OTHER guy out there with my name! I’m not sure which of us had it first, we’re about the same age. (It’s quite possible we were born in the same year, which would make things especially inconvenient, but I’m really not sure). And he does have an authority record.

So if I ever do end up publishing something that gets cataloged, are catalogers going to accidentally think I’m him and use that authority record?

(PS: My father happens to have an authority record already. None of the few others in there with my last name are related to me as far as I know. )

This entry was posted in General. Bookmark the permalink.

12 Responses to how do name authorities work anyway?

  1. brian says:

    Good question. We pay a firm (LTI) fairly big bucks to keep our authority records in line and I’m not convinced it worth it. Although I can see where being confident of all an author’s works is useful, I think using the money spent on authority work to buy more books and then relying on an author’s name to simply show up in a keyword results list sounds viable, at least for my tiny library. Someone tell me I’m wrong so I don’t have to pursue dropping our authority work contract….

  2. Jon Gorman says:

    I’m sure they will. After all, what’s to distinguish you?

    Brian – the issue will be how many people rely on things like clicking an authors name once within the record or if you have certain power users (read – librarians) that you want to support. Do you have a lot of people who follow authors? Series? Perhaps the answer is that some more focus work is needed rather than authority work

    I’m convinced that authority work shouldn’t need to be as hard as it it and that the current situation is a mix of vendors who don’t want the problem fixed (the ones doing things like fixing your records), those who know it’s a subtle enough feature that it’s not going to land them any sales (big ILS folks), and the dogma that we should do the bare minimum with these records to save time despite the fact in this day and age it saves us little time and ends up costing more in the long run.

    One of these days I’ll be able to create a nice system for authors…or just use systems that more modern ones have developed.

  3. Jonathan, Brian,

    At my library (ca. 13 million volumes) the investment in authority control is definitely worthwhile.

    As a member of the NACO program, I can say that if I received a book written by Jonathan, and I wanted to create an authority record for him, I would try to find something to distinguish his identity in the name heading. Preferably his year of birth. Otherwise, maybe a middle name or something like that. The general principle is that the first time a name is established, if there are no conflicts in the authority file, there is no need for qualification (e.g., birthdate, occupation, birth place, other names). Then, when the same name appears again, and we know it’s for a different person, we try to insert a qualifier. In rare cases that we can’t find any way to distinguish them, we create something called an “undifferentiated name heading”. But this is an absolute last resort (http://www.library.yale.edu/cataloging/authorities/tips/tips.html#undiff).

    In theory, the LC/NACO record number could serve as a unique identifier, and distinguish two persons with identical names. In which case the qualifier might not matter as much. On the other hand, it helps users navigate the catalog if they can visually distiguish, say, “Smith, John, 1580-1631” the explorer, from “Smith, John, 1749-1831” the artist. The qualifiers provide cues that don’t appear in the numeric identifier (but would still appear elsewhere in the record, i.e., as non-public notes in the 670 field)

  4. Jon Gorman says:

    Daniel –

    The problem that happens is that there are cases where birthdate is NOT enough to distinguish for very common names. (Smith, John 1880-1930 vs Smith, John 1878-1950). It requires you to know information from an external source that you may not have OR you investigate each option. I’ve never done any statistical evidence to figure out how frequently this happens though. Would love to do it.

    Part of the problem of course as you say is the tools. I tend to think that we should treat the LC/NACO record number as the unique identifier as you suggest. Then we need to “pre-compile” some data from the catalog and essentially create a record of the author that has biographical details, bibliography, etc. An advanced interface browsing authors might even be able to tweak what details are displayed.

  5. Daniel says:

    What do you think of WorldCat Identities? This service seems to mine data and bring it to the surface to help make those distinctions, e.g.: http://orlabs.oclc.org/identities/find?fullName=john+smith

  6. jrochkind says:

    I wonder if fields such as the 670 should in fact NOT be considered “non-public”, but instead be available for users, albeit only fairly sophisticated users would want to see it. But isn’t the name of one book by that author a useful thing for a user to see, to make sure they have the right person and learn more about them?

    It would certainly help if the 670 source fields were written in such a way that software could actually create a reliable _link_ to the bib record representing the work cited. Ie, using some kind of reliable, unambiguous, machine-actionable identifier.

    I also notice that my attempt to bookmark links to the authority records failed, sorry. Blame LC. Or SirsiDynix Horizon, really.

  7. Jon Gorman says:

    I love worldcat identities. It’s still not perfect, but it’s a great project and headed in the right step. I’m actually quite disappointed it’s not more tightly integrated with WorldCat and also doesn’t have many feedback mechanisms (it’s been a while since I looked though).

  8. Daniel says:

    To Jrochkind: Good point about the 670s. I don’t like the idea of hiding useful information from the public. There is a MARC field defined for public notes, i.e., the 680 field, http://www.loc.gov/marc/authority/concise/ad680.html, but it seems mostly used for scope notes in subject records.

    And LC has those permalinks for bibliographic records, but I guess not (yet?) for authority records. At least the LCSH in SKOS, once it’s back up, should have persistent links.

    To JGorman: I agree WC Identities are very promising. We should definitely be encouraging OCLC to do more with it.

  9. jrochkind says:

    In general, having to enter the same information multiple times: Once in a ‘non public’ form, once in a ‘public form’, perhaps once more in an actual machine-actionable form, etc.–is the kind inefficiency that makes our processes more expensive than they ought to be.

    Oh, and you guys may have noticed that I’ve included links to Worldcat Identities in my catalog and link resolver. There’s a lot more that could be done with this, to make it more powerful and reliable, but the simple steps I’ve taken are ( to me) interesting, and easily reproducible.

  10. I do have an authority record, and although as pointed out in a earlier comment the general rule is to add middle names and dates only in a conflict situation, that’s sometimes not how things work. In my case, my home institution set it up, and they had the info, so used it. Ergo, you can figure out how old I am using the LCNAF. Oh well.

    Rather than worry too much about how the current authority system works, it might be a good thing to take a look at FRAD, which is almost cooked, to see where we’re headed.

  11. jrochkind says:

    I am VERY dissapointed in FRAD. I think it’s gets it’s conceptual model completely wrong. It frustrated me so much that I am loathe to look at it ever again.

  12. Pingback: Unique authors » CommonPlace.Net

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s