We did not include all the values you mention from the Leader/07
(Bibliographic level), i.e. monographic component part, etc. in MODS
because this character position mixes different types of information and
we wanted to split these out. It includes issuance (i.e. serial,
monograph; whether it is finite or infinite), aggregation level
(collection, item) and mode of issuance (integrating updates or complete
as issued). So in MODS we defined issuance separately. "Collection" is an
attribute in type of resource, since each of those values could be a
single item or collection. Component part seemed to be such a difficult
concept that is relative to the particular situation and not particularly
useful that we didn't provide anything specifically for it. Components of
a larger work being described can be expressed in relatedItem (as
type="constituent") or if a component itself is described by linking to
the parent in relatedItem with type="host".

In all this discussion about being able to better handle journal articles
and the like this question does arise about whether we need something that
specifically says "I am a component part" or whether the other data will
make it obvious (particularly the relatedItem data). I think the latter is
the case, particularly if we enhance the ability to give more information
about the parent/host.

^^  Rebecca S. Guenther                                   ^^
^^  Senior Networking and Standards Specialist            ^^
^^  Network Development and MARC Standards Office         ^^
^^  1st and Independence Ave. SE                          ^^
^^  Library of Congress                                   ^^
^^  Washington, DC 20540-4402                             ^^
^^  (202) 707-5092 (voice)    (202) 707-0115 (FAX)        ^^
^^  [log in to unmask]                                          ^^
^^                                                        ^^

On Sun, 13 Apr 2003, Karen Coyle wrote:

> The "reference type" thing is a bit complex. I think there are two ways to
> look at "reference type" --
> 1) you can look at it structurally - this means that you divide your
> universe into items published singly (books) and those published serially
> (journals). (There is another category, but let's not go there for now.)
> Then you can have your "types" be structural parts of those items -- i.e.
> chapters of books, articles in journals. There is a code in MARC for
> "monographic component part" and "serial component part." A bit terse, but
> I find these very useful for data processing because they tell you what
> data elements to expect and how to interpret them.
> 2) Some systems name "types" that are more about the nature of the
> contents, but that often also include your "reference type": journal
> article, conference paper, literature review, patent, preprint, technical
> report, etc. These are tricky because they don't always unambiguously
> correspond to one of the structural elements in 1, above, and different
> communities interpret them differently. However, sometimes that's all
> you're given to work with. In that case, I usually find that we have to
> extrapolate using other data elements in the record (i.e. if it has an ISSN
> and gives a page range, then it's probably a journal article).
> I think we should strive to clarify 1 in MODS. Even that isn't easy, and
> I've been working on a scheme that I'll try to post tomorrow if I get more
> time to work on it.
> kc
> At 05:11 PM 4/12/2003 -0400, you wrote:
> >Karen wrote:
> >
> >>So if we can extend the MODS record to accommodate data from citation
> >>indexing then we are working toward that goal of integrated
> >>information resources.
> >
> >So what is missing, beyond elements for things like volume and issue
> >number?
> >
> >One thing I'm starting to understand is this:
> >
> >On one hand, MODS/MARC has a "resource type" that is broader and more
> >general than the "reference type" you see in applications focused on
> >bibliographic formatting (where type examples include book, journal
> >article, chapter, map, etc.).  On the other, it has the "genre" which
> >is more specific and related to content.  But in order to format
> >bibliographic records, you do need an intermediate categorization that
> >can describe how the record ought to be formatted.  I don't know
> >exactly what that ought to be in practice, though, or whether MODS
> >ought to try to include that.
> >
> >Bruce
> *********************************************
> Karen Coyle           [log in to unmask]
> **********************************************