|
On
the differences between
Date: Sat, 21
Jun 1997 17:16:12 -0400 ---------------------- Information from the mail header
----------------------- At 02:57 PM 6/20/97 EDT, E.O'Donnell wrote: For some time now this oclc-marc-fmp discussion has been taking place and no one has thought that there might be a distinction between the purposes of library cataloging and image cataloging that mitigate against following the library model too closely. I know that many slide-room directors have been trained in the library sciences, and from that perspective the chaos of cataloging images uniquely for each collection seems like a terrible waste of time; and, of course, if anything, it is time-consuming and may seem inefficient. But before everyone climbs aboard the bandwagon, allow me to pull out an old war-horse (or, to put it another way, a pet hobby horse of mine) and draw a distinction between what library cataloging and image cataloging aims to accomplish, and a distinction between their respective sources of information and roles in the galaxy of information science.. Libraries record publications primarily: books, tapes, and various media that are issued in multiple copies. There are unique archives and rare books, of course; but, they form a separate story of their own. Most of these "library" publications are identified on their title page or box by giving author, title, place of pub, publisher, date, etc. etc. The function of library cataloging and of uploading the data into OCLC and of using a standard format like MARC is to take benefit from the fact that one library's book will be essentially the same as that found in another library. So, theoretically, if one library catalogs the item, the resulting record should be suitable for all. That, of course, is not strictly true, standard cataloging procedures and practices do tend to bend in order to accommodate the differing demands of special libraries and different patronage. But, in general, one library can usually depend upon the "facts" addressed and the cataloging decisions made by another. The facts of cataloging are there -- part and property of the item cataloged. Fine arts cataloging is quite different, although, from our common point of view as a 20th century art audience, it may seem to have much in common with book cataloging. But, when looked at from the perspective of "evolving history" the identification criteria of objects of fine art are not a collection of facts; indeed, hardly an object, save a few signed works, contain any facts at all. Our factual basis of art identification is a facade being held up by a set of common beliefs about how our past has evolved. And this facade is quite wobbly; throughout the history of art it has changed considerably -- with the discovery of new personalities, new objects, with the evolution of connoisseurship, the development of scientific identification techniques and so on and so on. And it is changing still, in slow motion perhaps, but to be sure, nothing in it is really firm. The only thing that is sure is change itself. One of the jobs of visual resources curators is to press on through this protean knowledge base and constantly to adapt their collections to changing conditions and the changing base of knowledge. Curiously, this process tends to occur rather automatically as new images are pressed into the collection from new books with new identification tags. Old images, with old identifications tend to drop out. [An argument for using books as a source for images.] Imagine the teaching visual resources collection as a long canoe pointed into the flowing stream of time. As time sweeps by, without even moving, its bow breaks new headwaters and its stern passes through old headwaters. The information in such collections contains the present state of the discipline and part of the past state of the discipline. And all of it, together, is an uncertain and unsteady body of knowledge built upon the ever rippling surface of opinions. (Check out Heraclitus' concept of change, and Plato's idea of nature in flux.) Who then has the better record? From the point of view of some future researcher is it better to have today's set of assumptions catalogued, or yesterday's. The Rembrandt scholar is as much interested in the corpus of works we ascribe to the master today as in those works said to be by Rembrandt by scholars in the 19th century. I'll admit it; any use of a device like OCLC to share information would be certainly convenient. For despite my protestations, having a body of authoritatively cataloged works (Okay, so "authoritatively" is not an absolute.) can, indeed, save lots of time for lots of people. In the long run, and makes the life of the typical user so much easier. However, the "authority" claimed by such records is suspect and, unlike records made for books, will eventually be undermined by the vicissitudes of evolving opinion. The life of a Visual Resources curator is a constant challenge to make the information in his system useful to the collection's clientele. Untouched, the collection quickly becomes stale. Now for cataloging schemes: MARC was devised as a method of transmitting library cataloging data on linear tape for distribution to multiple libraries. The very structure of MARC was developed under the assumption that MARC data is the kind of factual book-data one finds (where else), but on books. The kinds of complex associations and opinions, citations, inscriptions, attributions to person, place, style, culture that we encounter in fine arts cataloging is not easy to accomplish in MARC. Though it might be easy to record, it is difficult to track and to attribute. Further, MARC uses an architecture that better supports lists than relationships. Historical associations and the kinds of connection that we have become used to when using "relational" databases are not native to MARC. No, I know (without quite knowing all the facts) that with increased power and new sophisticated tools, MARC has been recast somewhat by allowing the use of various lexicons and so on. But in general, it is a library tool -- a finding tool -- and not a scholarly one. Susan Williams' desire to establish a "communication" tool for use among collections is important, and once a set of essential fields has been defined (hasn't it already?), one should be able to proceed. Transmitting information, as MARC has shown us, is best handled by non-relational systems, and perhaps a cheap tool like FMP does hold some promises there, but there are file transmission protocols, too. But I suspect that it is far better to take a VERY powerful tool and under-use it than to take an easy and ultimately weak took and attempt to overuse it to the point at which it breaks. Based upon SGML (Standard Generalized Markup Language), a subset of which is the basis for HTML, and under the auspices of the Getty Trust the Museum Computer Network (MCN) and other interested parties, John Perkins was commissioned some years ago to develop an interchange format for museum information. I am not familiar with all the details of this project: CIMI (Consortium for the Interchange of Museum Information -- http://www.cimi.org/cimi ), but I can tell you that it was developed to incorporate complex museum data as well as data of use to various cultural organizations. Right now it is in its test phases which you can read about at the URL cited above. In conclusion, I want to note that the questions posed on
VRA-L this last week bring up an old problem: How much effort
should be expended in serving the needs of current users, and how
much should be dedicated to preserving the present and past for
future users. The answer that always pops up first is that the
present must be served first. At the same time, we often hear
laments to the effect that it is too bad that the past wasn't so
well protected. But what is the present? It is just this
razor-edge of the moment keeping its place in a flowing river of
time in which the future and past spread out infinitely before
and behind us. I think it is possible to dedicate ourselves to
the demands of the future by paying due homage to our obligations
to the past. Indeed, preparing for the future IS serving the
present. How we see ourselves in the fullness of time will
influence and determine what our databases eventually look like. [ Go to part two ] [ Home Page | E-mail Menu ] |