All,
This thread has taken an interesting turn. The purpose of my message was not to say that traditional library methods are better or worse than the relevance ranking, but simply to give some very basic informatio--to those who do not know--how a catalog is supposed to work. It is obvious on this list that many people do not understand how library catalogs are supposed to work. Whether catalogs fulfill the task well or badly is a separate point, but this is how they supposed to work.
What I tried to show is that library catalogs are designed to give access to the items in a collection by their authors, titles and subjects and they do this not by using "text" but they do it by "arrangement," which tries its best to allow concept searching. There are known parameters beyond which this does not work, such as the rule of three, 20% of an item for subject analysis and so on. This part is not up for discussion: these are simply facts about how the "library machine" works. Disputing this is similar to someone who is learning to drive a car that it would be better to use levers to steer instead of a steering wheel, or that how fast you go and how you stop should be through voice commands instead of foot controls. Perhaps the person's beliefs are right or wrong, but all that is irrelevant to the task at hand: when someone is learning to drive a car, there is a method to do it and you must use a steering wheel and foot pedals. There is also a way to use a catalog,!
whether
we like it or not. Getting angry about it make no sense. If you don't "agree" with driving, then don't drive a car or build your own and hope you can get a street license for it; if you don't like the way catalogs work, use Google or whatever else you can find.
Still, to maintain that traditional methods do not work is to be standing on rather thin ice. This is because this is how catalogs have worked for a long, long time and until not that long ago (only 25 or 30 years?), this was the only method that people could use since keyword searching was unavailable. Therefore, we would have to conclude that research before the late 20th century is highly suspect, and I for one, would not want to try to maintain such an extreme position. Do we really want to maintain that the work of tens of thousands for the last 150 years is worthless? Do we want to say that the work of Darwin, Kant, Einstein and countless others was inferior and that they couldn't do any research because there was no keyword? This is how I read some of these messages, and I cannot agree, since they did some excellent work. If this is not the upshot of the messages, please correct me. If we conclude that they did some good work, then we must ask why it worked then, but !
does not
work today.
I still have seen nothing in these replies that convinces me that Google can do concept searches. I have seen references that say, for example, there is a link to Wikipedia which is number 1 or number 2, and from there you can find concepts (this in itself is highly disputable, but I'll avoid it for now), but what this has to do with Google ranking is beyond me. If we want to say that Wikipedia can emulate the authority files, that is a point for discussion, but just because the Google spider includes Wikipedia, and that many people link to wikipedia articles so that they come up higher in the results does not equal concept searching.
And yes, the library catalog absolutely can do concept searches, and I believe I have already shown it. It's not perfect, but it can do things quite different from relevance searching and is very, very useful, as our forbears have shown. For those who wish to disbelieve, there is little I can do at this point except say: don't use the authorized forms and stick with keyword.
As Karen points out, users want to search things only once and get exactly what they want. Librarians know that to find materials, you need to search in many ways using any tools you may have. Where I disagree is that I think creating new tools will be very much a cooperative affair and will nee
d people from various fields. These people must share a common knowledge of the strengths and weaknesses of the various tools that are currently available. That does not mean common agreement or common opinions, but common knowledge.
That was the primary reason why I described very simply how a catalog is designed to work--not to say that it is the best or the only way that it should be done. And yes, it has been a success for many people for a long time. I would like to try to build on this success, but others obviously take a different opinion.
By the way, I'll be at ALA midwinter, myself.
James Weinheimer
Received on Sun Jan 06 2008 - 15:55:00 EST