Alexander Johannesen wrote:
> But, I believe that LCSH can potentially provide users much more useful browsing,
> using the great syndetic structure, so long as people do *not* have to navigate it as
> they did in the card catalog, where it worked a lot better. I think the see alsos are
> great, so that when I think I want "Authority" I find the exceedingly helpful:
> Narrower Term: Divine right of kings.
> Narrower Term: Example.
> Narrower Term: General will.
> Narrower Term: Power (Philosophy)
> See Also: Authoritarianism
> See Also: Consensus (Social sciences)
<snip>
Well, that in itself is good stuff, but the problem is; what - if anything - is at those endpoints? Will a book that potentially *could*
have been classified as "Power (Philosophy)" be classified as such? If I rely on this rigid system, will I miss out on books if I trust it
too much? I also know in modern days people are more liberal with the number of categories per book, but there's a 100 year backlog which is
a bit more scant. And then, just because we got "Power (Philosophy)" attached to a book, what part of it deals with it? Big part, small
part? Why is it categorized as such?
These are all human questions that cannot be solved by computers on *our* side, apart from wide speculation. But a realize that this isn't
LCSH criticism, this is just me asking for some infra-structure or methods that could be shared with our users in order to improve, and
get away from the rigid nature of the beast. There is some balance between useless junk and rigid hierarchy where I suspect you *should*
be with LCSH that will halt its popularity.
</snip>
I can tell you what it is *supposed* to be: when you navigate through these subjects, you should be looking at the major topics covered by 20% or more of a resource. Of course, this depends on the catalogers: their level of training, amount of time for each record, ease of use of the cataloging system, level of morale, and other variables. Many problems are inherent in this system. Primary among them is the fact that research pretty much consistently shows that when assigning subjects, two different well-trained catalogers will assign two different subjects. So, for example, one will assign to a certain resource "Authority" while another will assign "Power (Philosophy) or another will assign "Authoritarianism." Many people therefore conclude that subject assignment is more or less useless. Q.E.D.
I don't agree with this conclusion, *so long as* the syndetic structure is easily available. While I have never done any research in this, my experience leads me to believe that while two well-trained catalogers may not agree 100% with one another on all the subjects of every book, they will come close and that is one of the purposes of the syndetic structure, to help the users who can therefore, also come close. It seems to me rather like putting in golf. If you try to sink the putt in one shot, lots of times you miss and you have to go running after the ball. But, if your goal is to two-putt, all you have to do is get close and then you sink the shot. That is much more possible. Without the syndetic structure, the user would never on his or her own, make the connection between "Authority" and "Power (Philosophy)" and as a result, is pretty much doomed to sinking the putt in one shot.
Now, what about the problem of those subjects that are less than 20%? This is where a symbiotic relationship with automated means and Web2.0 tools could prove highly useful. Again, I have done no research in these matters and don't know if any has been done, but in my experience, automated methods don't do very well at determining higher-level subjects of a resource, e.g. this book as a whole is about baroque architecture, or this book as a whole is about aquaculture methods in India. But, automated methods do pretty well for ferreting out the bits and pieces of information. That is, automated methods are strong where traditional methods are weak, and vice versa. Add to this the Web2.0 possibilities, and we have something genuinely new and useful (at least, I would think so).
But yes, LCSH must become much more flexible in its use, incorporate about 10000 times more cross-references, hopefully open it up to cross-references added by reference librarians and the general public, and managed by us. I think LCSH can demonstrate this flexibility, but apparently the RDF format is not sufficiently robust and other solutions must be sought.
James L. Weinheimer j.weinheimer_at_aur.edu
Director of Library and Information Services
The American University of Rome
Rome, Italy
Received on Tue Nov 17 2009 - 03:30:48 EST