[UPS] <journal> journal

[herbert van de sompel]

As I said my suggestion is in no sense in contradiction to Stevan's.  We
are talking about different things, basically.  Stevan wants a field to
describe "Published in".  I suggest a field for "Author is interested to
publish in".  My suggestion to put the latter in the Subject field is
based on my primitive understanding that a journal title often gives a
good indication on the subject area.  But this info could as well be
held in another field, and Stevan has good points to illustrate why it
should indeed be in another field.  And I gave some illustrations of
potential other use for the field.

A problem I see with Stevan's suggestion is that the Journal field will
sparsely be filled out, because it requires the author to go back to the
preprint metadata, once he knows in which journal, issue etc... his
paper will be published.  In order to avoid this, we might come to a
convergence of Stevan's and my suggestion and use Journal tag
- initially to contain "author is interested to publish in"
- whenever a real publication occurs and the author thinks of editing
the metadata, use it to contain "published in"
The XML DTD could give both a seperate attribute.

Apart from that, allow me to contradict that I would be against Stevan's
suggestion to include this Journal tag, being inspired by an underlying
idea that journals will vanish at some point in time.  Although I have
my personal opinion on the matter, I would never go as far as to
restrict a metadata scheme to reflect my very personal opinion.  Quite
to the contrary: all of you have seen that I used SFX to link eprints
with the established journal system, meaning that I do have a sense of
reality.  

All of this is left for the group to decide upon.

greetings

herbert







Stevan Harnad wrote:
> 
> On Wed, 27 Oct 1999, herbert van de sompel wrote:
> 
> > I suggested to (ab)use the Subject field as
> > a carrier for the information on the journal(s) in which the author
> > would like his eprint to be published. Whether this information would
> > indeed belong in the Subject field or in a seperate Journal field is
> > open for discussion.
> 
> I don't understand (perhaps it is because I am not a metadata expert)
> why (ab)using the Subject field is preferable to declaring a Journal
> field explicitly and unambiguously in our minimal tagging specs.
> 
> And "would like his eprint published" is NOT the right descriptor for
> this field. I am not talking about wishful thinking, before acceptance.
> I am talking about the journal in which an eprint HAS been published.
> 
> > - There is a universal understanding regarding the subject areas of
> > journals, meaning that there exists an agreed upon global
> > subject-oriented namespace being the journal namespace (that actually
> > even has unique identiefiers - ISSNs).  This can hardly be said about
> > other subject classifications.
> 
> That is an argument against a Subject field, not against a Journal
> field. I am in favour of a Subject field too, even though it (and
> Keywords) are not used very useably/usefully by many authors. One could still
> mine something useful out of Subject tags. But the name of the Journal
> does NOT belong there. When I look for subject "psychology" I don't
> search on the term "Psychonomic Science" (a journal name) and vice
> versa.
> 
> Sometimes journal names reflect subject matter, sometimes not, but
> either way, they do not belong in a common grave with Subject terms.
> 
> If I want to retrieve, say, all papers in (the journal called)
> Psychonomic Science in (Date) 1998, I want to be able to do that with two
> definite fields, the Date (2) field, for which Herb argued successfully
> for inclusion (and I opposed, but TOGETHER with a Journal field is
> makes terrific sense), plus the Journal field.
> 
> And if I want the journal (not subject) called "Biochemistry" for all of
> 1998, I don't want to retrieve everything in subject Biochemistry dated
> 1998!
> 
> Etc. etc. I don't think it requires meta-data expertise to see that this
> field deserves an unambiguous category of its own.
> 
> I continue to detect animus against journals motivating this reluctance
> to declare a specific, dedicated Journal field. (The "would like to be
> published" phrase above is a sign of Herb's preferred outcome, which
> will be the day when there are no more journals. That is not yet the
> case; it may never be the case; and in terms of the way I, for one,
> happen to see it, it SHOULD never be the case. So our tagging scheme
> scheme should not be predicated on the assumption, or hope, that it
> will, or should, please.)
> 
> > - It is a type of Subject classification that can indeed be added by the
> > author, who IS a connoisseur of the journals in his domain, but hardly
> > is a subject classification expert.
> 
> True, but irrelevant. Why not, then, put the Title or Authorname into
> the Subject field too...?
> 
> > It is a
> > subject classification that could be used by whomever and that could
> > lead to very interesrting features in end-user services.
> 
> True, and irrelevant.
> 
> > - It could be used as a gateway mechanism between eprint archives and
> > publishers: publishers could harvest records that have their journal
> > names in the Subject/Journal field and decide whether they are or are
> > not interested in publishing the paper.
> 
> And how do they distinguish, say, Subject-Biochemistry from
> Journalname-Biochemistry? Are there going to be embedded sub-meta-tags
> within the Subject field? But why? Doesn't Journal-name, especially for
> this sort of "harvesting" (which I see Reader-Users of Author-tagged
> Self-Archives as using, by the way, not particularly publishers), call
> for an explicit, unambiguous field of its own, in any "minimal" tagging
> system?
> 
> > - I see this as a very nice way to gently parasite on the existing
> > communication system.
> 
> As soon as I understand why, say, Author and Title should not likewise
> parasite on Subject, I will understand why Journal should.
> 
> Cheers, Stevan