3 messages, 186 lines:

(1)------------------------
Date:         Thu, 25 Apr 1996 17:49:08 -0400
From:         "Bremer,Robert" <bremerr@OCLC.ORG>
Organization: OCLC Inc, Dublin, Ohio USA  (614) 764-6000
Subject:      Manuscript serials?

Frieda Rosenberg wrote:

Here's a question of definition arising from my own recent query to
Serialist.  The question is: Can a serial, defined as "a *publication*
in any medium..." be a manuscript?  OCLC has said no--since serials are
"publications" they cannot also be manuscripts ...
OCLC in fact will not validate manuscript coding on the serials
workform.  But the definition of "manuscript" as "writings (including
musical scores, maps, etc.) made by hand, typescripts, and inscriptions
on clay tablets, stone, etc." does not exclude the idea of seriality,
nor does it imply that manuscript items must be unpublished.  If we are
cataloging theses as manuscripts, despite their wide distribution via
UMI reproduction, it seems illogical to say that a serial in longhand
with even more limited distribution must be a book!  One CONSER
librarian who replied seemed to agree. CONSER documentation
redefining the leader, byte 6 (type of record) has yet to appear to my
knowledge, but perhaps someone in the know can contribute a word about
what will happen.

I know it's really all one format.  But OCLC's exclusion of options is the
wrong way to handle a process (format integration) that was supposed to
make all codes available for all materials.   Besides, many systems
allow limitation by format or even (though I'm not advocating it)
segregate serials from monographs in two different files.

 ----------------------------------------------------------------------------
Response:

The question concerning the potential coding of manuscript serials had come
up here at OCLC while planning the implementation of format integration.  We
consulted LC staff, studied the rules and RIs, and came to the conclusion
that the AACR2 definitions and rules governing the description of
manuscripts and serials are mutually exclusive.  The definitions in the
glossary of AACR2 use the word "item" in defining monographs, supplements,
and numerous other concepts, but it specifically defines a serial as a
"publication" in any medium "issued in successive parts" ...  etc.  We don't
believe that this is simply an oversight on the part of the editors of
AACR2, and "publication" is defined in English-language dictionaries as
"something that is published, esp. a periodical."  The AACR2 definition
of serial also indicates "issued in successive parts."  English-language
dictionaries also define "issue" as "something that is printed or published
and distributed, esp. a given number of a periodical."

Likewise the dictionary definition of manuscript says it is "the original
text of an author's work, handwritten or now usually typed, that is
submitted to a publisher," i.e., the version of a work that precedes its
publication.  While the AACR2 definition of manuscript does not specifically
say manuscripts are unpublished, it is clearly the intent per rule 4.4.
 AACR2 chapter 4 rules govern the description of manuscripts.  The rules
instruct the cataloger to only record a date in area 4 with no place and no
publisher, further indicating that manuscripts are not published.

How can you then have a manuscript serial since that would mean that you
have a manuscript publication, i.e., an unpublished publication and a
contradiction in terms?

UMI dissertations are cataloged as reproductions of manuscripts.  Current
practice is to describe the original manuscript in the body of the
description, again omitting place and publisher in field 260 $a and $b, and
describe the details of the published reproduction in field 533 including
the place and publisher of the reproduction.  Fixed-field coding for
reproductions is based on the original item being reproduced so that type of
record (Leader/06) code t is used on such records reflecting the manuscript
nature of the original material, so it's not clear how this practice relates
to the discussion of manuscript serials.

Keep in mind that format integration did not change AACR2 cataloging rules,
but rather it ironed out the differences in coding practices among the
bibliographic formats.  With the implementation of PRISM validation in OCLC,
we have included relationship rules between elements to assist users in
avoiding illogical or invalid combinations of elements in records that
should not occur according to cataloging rules.  For example, PRISM
validation will not let you include field 255 for a map scale on a record
for a score (Leader/06=c).  Technically, field 255 is valid in all formats,
but cataloging rules that actually govern the content of MARC records would
not have you add a map scale to a description of a score.  In such a case,
the user probably made a typographical error and had intended to add field
254 instead.  Likewise, cataloging rules would exclude certain combinations
of type of record (Leader/06) and bibliographic level (Leader/07), such as
manuscripts and serials.  We have added relationships to PRISM validation
for these invalid combinations to assist users and avoid input of incorrect
records.

The CONSER Editing Guide update number 3 for spring 1996 does include a
revision page for Type of record (Leader/06) which indicates that codes d,
f, p, and t are not applicable to serials cataloging.

The kinds of manuscript materials sometimes cited as requiring serial
treatment include ships logs, meeting minutes, diaries, etc.  In applying
AACR2 rules to the description of such items, you would probably have to
supply the title and otherwise base the description on the first issue.  But
what is the first issue?  An entry in a ledger which would not meet the
serial definition requirement of "issued in successive parts."  Or is it the
first physical volume the first issue?  Does the first physical volume then
carry the required designation?  Would a person's letters qualify as a
serial with each letter being an "issue" carrying it's own date designation?
 Where would the line be drawn since our archives and manuscript colleagues
are working with these same materials and continue to call them collections.
 There would need to be some consistent treatment agreed upon by both groups
to facilitate the exchange of records and the reliability of the data they
contain.

The current definition of serial in AACR2 may need to be changed in the
coming years for a variety of reasons.  At that point, we would change
validation in support of the rules, until then we continue to support
current standards.

 --Robert Bremer, OCLC
   <bremerr@OCLC.ORG>

(2)-------------------------
Date:         Fri, 26 Apr 1996 11:21:23 EST
From:         Joe Orth <JOrth@CK6.USCOURTS.GOV>
Subject:      Re[2]: Manuscript serials and format integration

     Frieda,
        I never should have opened my mouth on this one.  I've only
     cataloged monographs and done copy cataloging on serials (that's one
     of the reasons I follow this list -- to learn more about serials!).
     But now that I've demonstrated how I can "open mouth before engaging
     brain", I have to ask what you mean by ignoring manuscript coding.
     Will you add the serial 006 to the books record and place the 362
     field info in a 500 type field?  Or just catalog it as a serial and
     note that it is in manuscript form.  Your situation does seem rather
     unique.

     I went to a class on Format Integration 6-7 weeks ago, and I remember
     them saying that all fields would be valid for all formats (so I
     thought you could add a 362 to a bks record, but clearly I'm
     mistaken).  Is there any chance that this statement is true for USMARC
     itself, but not for OCLC's implementation of USMARC?  In TB212 on p.
     36 the introduction to BLvl says "Values are valid for all formats;
     however, *not* every code is appropriate for every type of material.
     You must use an appropriate code for the material type for your record
     to pass validation."  So does the first statement apply to USMARC
     itself and the second part OCLC's implementation?

______________________________ Reply Separator _________________________________
Subject: Re: Manuscript serials and format integration
Author:  Frieda Rosenberg <friedat@EMAIL.UNC.EDU>
Date:    4/25/96 4:04 PM

...[snip]
Our solution is to ignore manuscript coding until we get further
information.

Regards,
Frieda Rosenberg, UNC-CH
<friedat@EMAIL.UNC.EDU>

(3)---------------------------
Date:         Thu, 25 Apr 1996 19:21:36 -0500
From:         Joel Hahn <jhahn@CCS.NSLSILUS.ORG>
Subject:      Re: Manuscript serials and format integration

On Thu, 25 Apr 1996, Frieda Rosenberg wrote:
> On Thu, 25 Apr 1996, Joe Orth wrote:
>
> >      I thought this was part of the reason for format integration. The
> > main format would be mixed materials, possibly even allowing archival
> > control.  Then one could add an 006 for the "seriality" of the item.
> > Having cut my teeth on OCLC, I tend to think of serials only as
> > published items.

>     On OCLC, manuscript serials would be cataloged not as mixed
> materials, but as books. (Mixed materials can only be collections
> organized around a person or body.) With the Books format, you are not
> allowed to input a 362--it will not validate.  With the serials
> format, you are not allowed a Type t--it will not validate.

Call me silly, but wouldn't it work to use the serials format and add an
006 line for the bks format?  Type t will validate just fine when it's
in an 006 line.

Joel Hahn
Niles Public Library District
<jhahn@CCS.NSLSILUS.ORG>