Message-Id: <199706050109.SAA290788@dns.ccit.arizona.edu> Date: Wed, 4 Jun 1997 21:10:06 -0400 From: "Robert A. Baron" <mailto:rabaron@PIPELINE.COM> Subject: Re: database construction To: mailto:IMAGELIB@LISTSERV.ARIZONA.EDU
At 04:38 PM 6/4/97 -0700, Richard Rinehart wrote:>I know you can take a raw database program and fit it to any institutional
>need by defining fields/relations, but when looking at off-the-shelf
>systems it's good to look at intended use factors in addition to the
>technical requirements/power ratios.
>
>my two c :)
>
Richard is undervaluing the worth of his advice. While the technical issues that must be faced to deliver large image databases are immense, I think there is a tendency among those who must consider these questions to undervalue the complexity of the data structures needed to provide a serious intellectual backbone to their work. Without mentioning the obviously sophisticated requirements associated with object management that museums need, any intellectual database of even modest pretensions must be able to handle 1) multiple dated attributions (maker, school, style, bibliography and commentary) and their history, 2) part/whole/associated relationships among objects, some of which inherit and some of which do not inherit attributes from their parents (and some of which may not even exist any longer or are simply thought to have existed), 3) relationships between objects and the images which document them, and the makers of those images, and 4) other sets of attributions to an ever changing/evolving list of stylistic and cultural criteria and provenance. (Not to mention meaning, iconography, literary, historical and social associations.) While the above-stated concepts may suit historians and curators, there will be entirely different sets of requirements for graphic artists and illustrators, for anthropologists, etc.The problem at hand (as I see it) is not simply one of how to deliver the images, but rather how to find those to deliver. When we look at the eight or ten standard fields that typically accompany some museum's contribution to web-based databases, we know that while they might satisfy eighty percent of the public's need to find information, they do not come anywhere near what might be demanded by scholarly inquiry.
It may be that the need to engage the public is the force driving the development of image database systems, and that scholarly requirements must accordingly take a back seat, but if all the work that goes into serving that hypothetical public is not at some time transformed into lasting scholarly systems of record, I imagine that once the public infatuation ebbs, much of the effort expended in their behalf will have been wasted.
And the reason why this is so is because it is the scholarly tradition that keeps the data alive and causes it to evolve to suit contemporary needs and issues. Without scholarly input (even though it is often hidden from the public) information quickly becomes stale and moribund.
=========================== Robert A. Baron Museum Computer Consultant mailto:mailto:rabaron@pipeline.com