From: Peter Graham, Rutgers University Libraries
Here are some comments on the revised, 2/21, Qualifiers document of Jon K and
Martin H, which looks better and better.
1. Contemporary ID information: for author, publisher, contributor, etc. I
continue to think that these are ephemeral and should be reserved for a
different element, and would be interested in hearing how a later version of
DC would allow them to be moved. If the DC is for "intrinsic" information,
this information seems extrinsic. (Author name is intrinsic; this week's
address is not. As we used to say in the library biz, "the future is
longer than the past.")
Even if we don't change it now, I think the terminology of "HomePhone" or
"HomePostal" could do with a change. "Personal" perhaps? In particular I'm
not sure what a "home" email address is, since one can use it from anywhere.
The whole terminology of Phone, Fax, HomePhone, has a smell of informal
definitions about it to me. "Primary" and "alternate" or "secondary" could
make more sense and fewer judgements.
2. Dates--and I'm not talking format here, but content.
a. The DATE element under Type speaks of "the resource", but this is
there ambiguously defined. What is the difference between "the
resource" and "current form of the resource"? What is meant by "first
created"? What is the difference betrween "current form" date and the
"date on which the resource was last modified"?
b. In my mind in all this is the use of DC to describe artifactual
materials, what we used to call books. In this context, the SOURCE
element in an earlier version of January had several additional date
specifying fields, as I recall. Consider the case: At Rutgers we
have an 1856 broadside we intend to digitize. We want it to be
accessible in searching terms in terms of its originary date, not the
date of the networked image resource. At the moment the SOURCE and
DATE fields don't seem to allow useful specification of these matters.
3. IDENTIFIER: The ISBN is notoriously not dependable as a unique
identifier, as it is publisher assigned with all sorts of errors and
multiple assignments. Does this matter to our thinking? It would be
no worse to search on it as a DC element than in MARC records,
however.
(The ISSN is in somewhat better shape as it is agency assigned.)
4. IDENTIFIER--PRIMARY VERSION: Who decides what is primary? Would
this be used to refer to an artifactual version of the resource? (Only
if this is a separate DC metarecord, I assume.)
But in what terms is "primary" meant--or is this to be left to
localoption? ("The primary version that we got from the vendor is on our
main server; the copy is on our mirror site.") Do changes in format
affect the canonicity? "This is the WP6.0 version of the WP4.0
document, and is now the canonical version.") Does this kind of
definition matter at this point?
5. AUTHENTICITY / INTEGRITY: I'm not sure this is intrinsic or
extrinsic, but needs to be dealt with when people want to be sure they
are getting what they want (which is why I've put this here after the
canonicity mention.) I would think the RELATION element might have
an entry that allows reference to a validation file, or a reference
location for an authentication link. Needs to be somewhere, and it's
more intrinsic than rights, though less intrinsic than subject,
perhaps.
6. Inconsistent Relation Type Specs: In evaluating the Type specs I
mentally create a sentence, which has "this resource" as the subject
and the "type" given as the predicate, with the resource being pointed
to as the object of the sentence. Doing this uncovers an
inconsistency of use. The first 5 (IsParentOf, etc.) describe the
resource being metadescribed. The next 5 only make sense if the
related resource is being described as the subject of the sentence
(IsRevisionHistoryFor, etc.). The final one (IsDataFor) is at the
moment uninterpretable by me. I would urge a consistency of syntax
here.
In fact this document is an enormous help. Onward!
--pg
Peter Graham [log in to unmask] Rutgers University Libraries
169 College Ave., New Brunswick, NJ 08903 (908)445-5908; fax(908)445-5888
<URL:http://aultnis.rutgers.edu/pghome.html>
