Posting to NGC4LIB
I think we are actually on the same side, more-or-less, but there are some basic points of difference in our world-views. This may, or may not, have to do with the world-views of our respective roles as cataloger or systems person.
Library experience has shown that “better” is far too subjective to be a practical goal. What is better for me may very well be worse for you. As an illustration, I have always considered the process of cataloging to be similar to that of a mechanic (although I know this may irk some out there, from my working-class background, I consider comparison to a master mechanic to be a great compliment). The mechanic working on your car seeks to make your machine go the way it is supposed to go. If he tries to make your car go “better,” it must give you pause because then it will be going in the way it is not supposed to go. While it may end up genuinely going “better,” it may only seem that way to you and it may actually be worse in ways you don’t even know and will appear at unfortunate moments: polluting the atmosphere, being more prone to going out of control on curves, or in other ways.
This problem with coming up and implementing new ideas was something I pointed out in a thread I began about “lack of imagination” among catalogers. Boy, have I been excoriated for that one! But I still maintain it: the job of catalogers is to adhere to standards–just as mechanics must–and therefore, flights of imagination cannot be included within the system where those standards function. There can be brilliance within the system, and I have witnessed it many, many times, but not “flights of imagination.” This may be good or bad, but if catalogers want to let their imaginations free, which I think is vital today, they must work outside the system where those standards function, otherwise, they destroy it.
Library catalogs aim for reliable search results and not the “best” results. If I am interested in “love” and search in a library catalog with full cross-references, and I do it correctly, that is, following the methods laid down in the 19th-century for using card and printed catalogs (please keep reading!), I look under “L” browse to “Love” and find:
Narrower Term: Attachment behavior.
Narrower Term: Communism and love.
Narrower Term: Courtly love
Narrower Term: Courtship.
Narrower Term: God (Christianity)–Love.
Narrower Term: God (Christianity)–Worship and love.
Narrower Term: God (Hinduism)–Worship and love
Narrower Term: God (Islam)–Love [proposed]
Narrower Term: God (Islam)–Worship and love.
Narrower Term: God (Islam)–Worship and love [proposed update]
Narrower Term: God (Judaism)–Love.
Narrower Term: God (Judaism)–Worship and love.
Narrower Term: God–Love.
Narrower Term: God–Worship and love.
Narrower Term: Love, Maternal.
Narrower Term: Love, Paternal.
Narrower Term: Marriage.
Narrower Term: Platonic love.
Narrower Term: Summer romance
Narrower Term: Unrequited love.
Narrower Term: Yoga, Bhakti.
See Also: First loves
See Also: Friendship.
See Also: Intimacy (Psychology)
I think this is a provocative display that actually opens my mind to new possibilities that otherwise, I most probably would never have thought of. (Love to Summer romance, or to Bhakti yoga!) When I look at the individual records under the heading/concept “love”, I can be assured that I am looking at all items in my local collection (but not journal articles) where no matter in what language the item is in or when it was published, the topic of “love” equals 20% or more of the content (that is, if the catalogers behind the scenes have been trained to do this job and continue to do their jobs correctly, which is becoming increasingly problematic). Also, since there are so many books where 20% is on love, this topic has had to be subdivided into myriad groupings. And it is really handy to be able to separate out other things such as Love Act (Musical group) and Love actually (Motion picture). Please be aware that I also cannot use my imagination and decide that “Wedlock” is “better” than “Marriage” and enter it instead, because then the whole system breaks down.
Please understand, there can be no dispute about what I just wrote here. These are facts about how the machine works. These facts are being forgotten, and disputing them would be similar to arguing with your mechanic that the pistons shouldn’t work this way, but some other way. While you can have whatever opinion you like, it doesn’t change the fact that this is the way the pistons are supposed to work. Creating new records that fit into this system correctly is also highly complex.
Therefore, the library catalog does not aim for “better” results, but “reliable” results and “reliability” is based on a whole number of factors. I maintain that if Google is aiming for “better” results, that’s fine but “better” still remains a terribly subjective term that in practice equals “the result that makes me happy, but I may be happy only because I don’t know what I may be missing. If I knew what I was missing, I might not be happy, but if I don’t know, then ignorance is bliss.”
Now to change focus: can and should this traditional library catalog model, which relies on adherence to strict standards, be maintained in the new environment? This I do not know and I don’t think anyone will know without some kind of genuine research and “flights of imagination”. I will say that it is important to search by differentiated “personal, corporate, and geographic names” and “subjects” (I don’t know so much about titles), and that it would be very nice to be able to distinguish one Michael Gorman from another Michael Gorman.
The current method of browsing headings is definitely 100% obsolete, and has been since about 5 minutes after the implementation of keyword searching in the catalog. I would certainly like to see some new attempts to make the power of our headings more obvious in the new environment instead of just abandoning them. At last we are getting some tools such as Aquabrowser, but I personally find them unconvincing. Vivisimo and Grokster are on the right track and have also provided some interesting attempts but I haven’t seen anything with library-type headings.
If there are no imaginative attempts to make these traditional controls more useful today however, these standards probably will be abandoned as useless. These imaginative attempts presuppose open data however, and this has not been forthcoming from the library community. For example, while I applaud the id.loc.gov project, it was much too late, and we are still stuck with textual strings that must be browsed to be understandable, and therefore, we are still stuck in the 19th-century. If these incredible files had been put out 10 years ago for open development, they could very well have served as the backbone for the Semantic Web, but now, this will probably shift to something such as dbpedia, where I think we must get on board.
This is enough for now. I think that “better” will probably be the goal of the newer systems, although it must obviously be qualified by many, many caveats.