James Weinheimer wrote:
These sorts of practices always interest me and I try to come up with
ideas that bring them together. One way of looking at this would be
that a record for a serial is the manifestation, and that this single
manifestation has variant titles (not necessarily earlier ones, but
variants), similar to monographs that have spine titles, a variant
title on p. 4 of cover, and so on. That is how AACR2 and RDA consider
them. But the Germans (and I assume others--many?) would consider them
in the way you describe.
In RDA there is a difference between "variant title" (2.3.6) and
"earlier title proper" (2.3.7, only possible for integrating resources
at the moment) or "later title proper" (2.3.8) - they are seen as
separate elements. So I don't see a fundamental difference here between
the Anglo-American and the German view of things.
It occurs to me that we have the concept of *the* title of an item but
as we see here, there are problems with choosing a single title and
there always have been. Why do we have to pick one as being *the*
title? We always have but perhaps matters could be reconsidered. New
systems allow novel possibilities. Let's imagine something rather
blasphemous and almost impossible to conceive of in a card
environment: that 245a and b could be repeatable. As a result, all
246s (and 740s?) would be equal titles to what is in the 245 now. This
would mean that when there is more than one title, there is not *the*
title of an item but different titles of equal worth. And each 245
could have its own note explaining where it comes from, as they do
now, perhaps in a subfield i, as in the 246.
For retrieval, it certainly doesn't (or shouldn't) matter which title
you use. But I'm not so sure about display. Would we really do our
readers a favour if we presented all titles as having equal worth
(perhaps in an alphabetical or random order)? I think most users would
agree that a title on the title page is more authoritative than e.g. one
on the spine. If those responsible for the resource wouldn't have wanted
us to associate it primarily with this version of the title, they
wouldn't have put it in the most prominent position.
But things are indeed different with earlier and later titles. You could
say that B, C, and D from my example (minor variations in a journal
title, not leading to title splits) are on a level of equal importance,
with all of them stemming from the chief source of information of the
relevant issues. What makes them different is the chronological
dimension. Therefore I would readily accept a display as you have
suggested, giving the different titles closer together than they are now
(with one in the main part and the others in footnotes):
it could be something like:
Title proper: B (time period)
Title proper: C (time period)
Title proper: D (time period)
Users could be given a choice as to whether the list should start with
the oldest or the latest title. I'm fairly certain that most would
prefer the latest, so the default display could be D - C - B.
Theoretically (i.e. if we take the effort), the complete bibliographical
information for all chronological stages could be stored with the same
detail of description in repeatable and identically structured fields,
always with an indication about the time period for which this
information is releveant. We do something like that in the publication
area. Here's part of a record in the ZDB (in the internal format of the
database):
4030 Malden, Mass. $n Blackwell
4035 Beverly Hills, Calif. [u.a.] $n Sage $h -1982
4035 Guildford $n Butterworth $h 1983-1990,2
4035 Guildford $n Butterworth-Heinemann $h 1990,3-1991
4035 Cambridge [u.a.] $n Blackwell $h 1992-1996
Field 4030 gives the latest Information, i.e. the current publisher and
place of publication. The 4035s (which are repeatable) give the same
information for earlier stages, including a time specification in $h.
Unfortunately, it is not always done as precisely as you see it here.
Sometimes it just says "$h anfangs" (i.e. "in the beginning"), so that
we don't know when exactly the change took place.
This is not all that novel of an idea, since the VIAF brings together
different headings for a name, and does not choose any as *the* form,
and these can be displayed in different ways.
With respect to personal names, I quite agree: If identifiers are used
instead of text strings (authorized access points) for expressing
relationships, it really should no longer be necessary to decide on a
main form. The different forms of a name could be coded, e.g. as to
their language (e.g. "Horace $l eng" vs. "Horaz $l ger" vs. "Horatius
Flaccus, Quintus $l lat"). Then a user could set his or her preferences
for display accordingly, e.g. "if there is an English form, choose this
for display".
Heidrun
--
---------------------
Prof. Heidrun Wiesenmueller M.A.
Stuttgart Media University
Faculty of Information and Communication
Wolframstr. 32, 70191 Stuttgart, Germany
www.hdm-stuttgart.de/bi