=========================================================================
Date:         Mon, 21 Jul 1997 13:52:00 -0400
Reply-To:     "International Conference on the Principles and Future Dev. of
              AACR" <AACRCONF@INFOSERV.NLC-BNC.CA>
Sender:       "International Conference on the Principles and Future Dev. of
              AACR" <AACRCONF@INFOSERV.NLC-BNC.CA>
From:         Stewart Marg <MAS@ABS.NLC-BNC.CA>
Subject:      Preconference Discussion List
Content-transfer-encoding: 7BIT

The Joint Steering Committee for Revision of AACR (JSC) has established an
open electronic discussion list to facilitate an exchange of ideas on the
topics under consideration at the International Conference on the Principles
and Future Development of AACR to be held in Toronto, Canada on October
23-25, 1997.  The conference papers will be posted as they are received.
 They are available on the JSC Web site at:

http://www.nlc-bnc.ca/jsc/confpap.htm

Final papers received from Tom Delsey, Ronald Hagler, Jean Hirons/Crystal
Graham and Martha Yee have been posted.

Discussion list participants are encouraged to consider and respond to
issues raised in the conference papers.  It is anticipated that feedback
arising from this forum will be taken into account by the speakers and
conference participants as part of their preparation for the conference.

For your convenience, we have taken the liberty of subscribing you, as an
invited conference participant, to the discussion list.

Information on the discussion list follows:

AACRCONF is an unmoderated list.
AACRCONF submissions are being archived.

 -------------------------------------
Your listowner can be contacted at:
     AACRCONF-REQUEST@INFOSERV.NLC-BNC.CA
as well as at:
     AACRCONF-COORD@NLC-BNC.CA
 -------------------------------------
To post items to the list,
send messages to:
    AACRCONF@INFOSERV.NLC-BNC.CA
 --------------------------------------

Common LISTSERV Commands
*******************************************
AACRCONF uses a UNIX-based LISTSERV software. Information
about advanced commands for LISTSERV can be retrieved
by sending the following e-mail:

     To: LISTSERV@INFOSERV.NLC-BNC.CA
     Message: INFO REFCARD

All commands to LISTSERV should be in the body of the
message (starting on the first line), and NOT in the subject
field (which should always be left blank).  Since the interface
used is English, all commands to LISTSERV must be in English.

SIGNOFF AACRCONF
* To remove your name from the list.

INFO AACRCONF
* To obtain information about the list.

GET AACRCONF WELCOME
* To obtain the latest copy of this Welcome message.

SET AACRCONF NOMail
* This will stop postings from being sent to you until you
reset your status -- Please use this when you go on holidays
or will not be reading your e-mail regularly.

SET AACRCONF Mail
* This will reset your NOMail status and allow you to begin
receiving messages from the list again.

SET AACRCONF DIGest
* To receive one message per day that contains all the daily
postings to the list.

SET AACRCONF ACK
* To receive a system-generated acknowledgment each time you post
a message to the list (this is the default).

SET AACRCONF NOACK
* To NOT receive acknowledgments each time you post to the list.

SET AACRCONF REPro
* To receive "carbon copies" of messages that you post to the list.

SET AACRCONF NOREPro
* To NOT receive copies of what you post to the list (this is the
default).

INDex AACRCONF
* To obtain an index of monthly archival notebook logs containing
what has been posted to the list.

GET AACRCONF LOGYYMM
* To retrieve a monthly log file from the list archive.  Specify
which year and month by replacing the YY and MM (for example,
GET AACRCONF LOG9707 will retrieve the log file for July 1997).

REVIEW AACRCONF
* To get a list of non-concealed subscribers to the list,
preceded by the list header parameters.

SET AACRCONF CONCEAL
* To conceal your name and e-mail address from anyone using the
REVIEW command for the list.

HELP AACRCONF
* To obtain a short list of common LISTSERV commands.

*******************************************************
National Library of Canada
July 1997
*******************************************************
*******************************************************
=========================================================================
Date:         Tue, 22 Jul 1997 11:43:14 -0400
Reply-To:     mac@slc.bc.ca
Sender:       "International Conference on the Principles and Future Dev. of
              AACR" <AACRCONF@INFOSERV.NLC-BNC.CA>
From:         "J. McRee Elrod" <mac@SLC.BC.CA>
Organization: Special Libraries Cataloguing, Inc.
Subject:      "Access points for works"
Comments: cc: autocat@ubvm.cc.buffalo.edu

It is well worth while to download and read the four papers so far posted
to the future of AACR conference site (http://www.nlc-
bnc.ca/jsc/confpap.htm).  Last night I found myself alternately
exclaiming "what?" and "right on!" as I read them.

Ronald Hagler in "Access Points for Works" (13 p.) used some terminology
in ways which would be foreign to us in dealing with our customers.
"'Entry' can only mean a single bibliographic record ..." he writes.
This meaning might work in our communications with our few remaining full
sets of catalogue card customers, but would not work in dealing with our
printed catalogue or OPAC customers.  For them "entry" means what Hagler
calls an "access point".  The entry (or access point) appears once, and
all titles associated with it appear following.  The entry (or access
point) is repeated only if there is a column break.

The titles which follow (regardless of their relation to the "entry"
above) are in alphabetical order in the author portion on the catalogue,
and in inverse chronological order in the subject portion.  Except for
the first title in each sequence, an "entry" as defined by Hagler is not
seen.  The same applies to the first screen in an OPAC search under
author, which looks not too different from the printed catalogue page.
The "author's name ... along with a title" as discussed by Hagler is only
seen in the classed portion of the printed catalogue (rarely used by
patrons), or when a particular title is chosen from a screen display.

Hagler asserts that "in the vast majority of all cases, a work has one
author .. and its only documentary manifestation bears one title proper".
This may well be true in Podunkville Public Library, but among the fifty
customers for whom we catalogue, I would say this characterizes less than
half of acquisitions.  Almost all Canadian government publications, for
example, have an added title page in French.  For the bibliographic world
in general, I would suspect the "vast" is an overstatement.

Hagler's point that deconstruction of the MARC record in OPACs makes us
less concerned with redundancy than earlier is very well taken, although
the practice he cites of dropping "Shakespeare's" from "Shakespeare's
complete works" in title proper died with AACR[1] (the example given in
1.1B2 was "Marlowe's plays").

Hagler's recommendations that all works have authority records, and that
all bibliographic records give access to all works contained in an item,
are noble goals.  We are already giving access to all papers contained in
Canadian law symposia (via SLC's Canadian Law Symposia Index on the QL
online law reference service).  To establish an authority record for each
of these works (20 or so per symposia), and to extend this practice of
all materials catalogued, would price us out of our market niche.  I
suspect this will have to wait for scanning and automatic generation of
author/title work authority records.  Based on the Blackwell North
America records which have been enriched by scanning, I would suspect
that this would also require the abandoning of standardization in
capitalization and punctuation.
=========================================================================
Date:         Tue, 22 Jul 1997 13:14:51 -0400
Reply-To:     mac@slc.bc.ca
Sender:       "International Conference on the Principles and Future Dev. of
              AACR" <AACRCONF@INFOSERV.NLC-BNC.CA>
From:         "J. McRee Elrod" <mac@SLC.BC.CA>
Organization: Special Libraries Cataloguing, Inc.
Subject:      Re: "Access points for works"
Comments: To: 0004652928@MCIMAIL.COM
In-Reply-To:  <97072219332428/0004652928PK1EM@mcimail.com>

>It would be helpful if everyone "signed" their name at the end of the
>message. Thanks.

Sorry.  Here it is signed, and with corrections.  Perhaps the list owner
can drop the first one from the archive.


It is well worth while to download and read the four papers so far posted
to the future of AACR conference site (http://www.nlc-
bnc.ca/jsc/confpap.htm).  Last night I found myself alternately
exclaiming "what?" and "right on!" as I read them.

Ronald Hagler in "Access Points for Works" (13 p.) used some terminology
in ways which would be foreign to us in dealing with our customers.
"'Entry' can only mean a single bibliographic record ..." he writes.
This meaning might work in our communications with our few remaining full
sets of catalogue card customers, but would not work in dealing with our
printed catalogue or OPAC customers.  For them "entry" means what Hagler
calls an "access point".  The entry (or access point) appears once, and
all titles associated with it appear following.  The entry (or access
point) is repeated only if there is a column break.

The titles which follow (regardless of their relation to the "entry"
above) are in alphabetical order in the author portion on the catalogue,
and in inverse chronological order in the subject portion.  Except for
the first title in each sequence, an "entry" as defined by Hagler is not
seen.  The same applies to the first screen in an OPAC search under
author, which looks not too different from the printed catalogue page.
The "author's name ... along with a title" as discussed by Hagler is only
seen in the classed portion of the printed catalogue (rarely used by
patrons), or when a particular title is chosen from a screen display.

Basically, I think in rule construction we need to be as concerned with
catalogue construction as with individual record construction.

Hagler asserts that "in the vast majority of all cases, a work has one
author .. and its only documentary manifestation bears one title proper".
This may well be true in Podunkville Public Library, but among the fifty
customers for whom we catalogue, I would say this characterizes less than
half of acquisitions.  Almost all Canadian government publications, for
example, have an added title page in French.  For the bibliographic world
in general, I would suspect the "vast" is an overstatement.

Hagler's point that deconstruction of the MARC record in OPACs makes us
less concerned with redundancy than earlier is very well taken, although
the practice he cites of dropping "Shakespeare's" from "Shakespeare's
complete works" in title proper died with AACR2 in 1978 (the example
given in 1.1B2 was "Marlowe's plays").  AACR 1967 continued the 1949
"green book" provision.

Hagler's recommendations that all works have authority records, and that
all bibliographic records give access to all works contained in an item,
are noble goals.  We are already giving access to all papers contained in
Canadian law symposia (via SLC's Canadian Law Symposia Index on the QL
online law reference service).  To establish an authority record for each
of these works (20 or so per symposia), and to extend this practice of
all materials catalogued, would price us out of our market niche.  I
suspect this will have to wait for scanning and automatic generation of
author/title work authority records.  Based on the Blackwell North
America records which have been enriched by scanning, I would suspect
that this would also require the abandoning of standardization in
capitalization and punctuation.



   __       __   J. McRee (Mac) Elrod (mac@slc.bc.ca)
  {__  |   /     Special Libraries Cataloguing   HTTP://www.slc.bc.ca/
  ___} |__ \__________________________________________________________
=========================================================================
Date:         Tue, 22 Jul 1997 22:23:22 -0400
Reply-To:     mac@slc.bc.ca
Sender:       "International Conference on the Principles and Future Dev. of
              AACR" <AACRCONF@INFOSERV.NLC-BNC.CA>
From:         "J. McRee Elrod" <mac@SLC.BC.CA>
Organization: Special Libraries Cataloguing, Inc.
Subject:      What is a work?
Comments: To: autocat@ubvm.cc.buffalo.edu

Martha M. Yee's paper "What is a work?" (34 p.) available on the future
of AACR conference web site (http://www.nlc-bnc.ca/jsc/confpap.htm) is a
delight.  Not only does she write well, but she reflects a thorough
familiarity with the great variety of materials currently passing through
catalogue departments.

She lays a firm foundation for the definition of a work.

Her discussion of relationships among works is exhaustive.  Her reasoning
concerning guidebooks and manuals which accompany other material,
including electronic resources, could also be applied, I think, to art
exhibition catalogues which have a relationship to the exhibit they
reflect.

If her concept of a work could end the recent change to entering later
editions of standard legal texts under their editors as opposed to the
original author (whose surname is often part of the title proper), my law
firm customers would be most pleased.

If her sound reasoning leads to more helpful treatment of serials which
change titles, it will be a great help to us all.  She points out that
the difficulty in finding the records for all the title variations of a
single serial, or editions with changed authorship, particularly if the
library does not have each link in the chain of changes.

The definition of one serial through all its changes of title as a single
"work", and relating the successive entry titles by a common access
point, would be wonderful.

Mac

   __       __   J. McRee (Mac) Elrod (mac@slc.bc.ca)
  {__  |   /     Special Libraries Cataloguing   HTTP://www.slc.bc.ca/
  ___} |__ \__________________________________________________________
=========================================================================
Date:         Wed, 23 Jul 1997 16:53:11 +1000
Reply-To:     "International Conference on the Principles and Future Dev. of
              AACR" <AACRCONF@INFOSERV.NLC-BNC.CA>
Sender:       "International Conference on the Principles and Future Dev. of
              AACR" <AACRCONF@INFOSERV.NLC-BNC.CA>
From:         Giles S Martin <ulgsm@DEWEY.NEWCASTLE.EDU.AU>
Subject:      Re: What is a work?
Comments: cc: "J. McRee Elrod" <mac@SLC.BC.CA>
In-Reply-To:  <asW1zEJ3Bg+X092yn@slc.bc.ca>
Mime-Version: 1.0
Content-Type: TEXT/PLAIN; charset=US-ASCII

I too have enjoyed Martha Yee's paper -- and I want to re-read it to
make sure that I have understood evrything she is saying.

However, the part that I want to discuss is the "Assumptions" on page
4.

I particularly like the last two:

   Users usually don't know about editions they don't find.
   Users rarely have a way to protest or complain ... .

A big part of the difficulty in improving cataloguing standards is that
users often:
   * are unsure of what they are looking for;
   * expect that if they can't find it, then it probably doesn't exist;
and therefore
   * don't let cataloguers know about problems in the catalogue.

However, that shouldn't let us off the hook.  We should still try to
find out what our users' need; and that may need some empirical evidence
as well as logical analysis of the bibliographic universe.

However, I think that the first assumption in Martha Yee's paper is an
over-simplification:

   Most users seek particular works not particular editions.

There are a number of cases when users want particular editions:
   * Students who have been told that they have to read a particular
edition of a work.
   * People who want the latest edition of a reference work.
   * Scholars who want the first edition, or another particular edition.

In addition, there are users who are looking for particular
manifestations, particlarly for non-book material, where specific
equipment is needed to view or listen to the work.

So what I think is needed is an analysis of what levels of grouping
library users want -- whether they are called "editions",
"manifestations", "versions" or "works" doesn't matter so much as how
useful the gathering is.  It doesn't matter whether Shakespeare's
Macbeth as a work includes all the different editions, or all the
different film versions, or all the diferent performances.  It is likely
that:
  * some users are interested in all the versions of Macbeth, including
books, films and sound recordings;
  * some are just interested in the text, but don't care which text;
  * some want a sound recording, but don't care which sound recording;
  * some want a particular edition of the text;
  * some want a particular film of Macbeth; and
  * some want a particular film on a particular video format.

The needs of all of these users need to be catered for: our catalogues
should group library items at all these levels, and display the grouping in
an intelligible way.

The next decision would be to name the levels: two kinds of names,
which would be:
  * the abstract name of the level like "version" or "edition"; and
  * the heading for the level, as it will be displayed to the user, like
"Macbeth (1971)" (not that I would advocate a heading like that, which
doesn't offer much help to the user).

Then you could move towards setting up a sort of hierarchy in
your catalogue display, similar to that on pp. 25-26 of Martha Yee's
paper.

Regards
Giles

          ####    ##       Giles Martin
       #######   ####      Quality Control Section
     #################     University of Newcastle Libraries
   ####################    New South Wales, Australia
   ###################*    E-mail: ulgsm@dewey.newcastle.edu.au
    #####      ## ###      Phone:   +61 49 215 828 (International)
                           Fax:     +61 49 215 833 (International)
                  ##
 The web of our life is of a mingled yarn, good and ill together
                        -- All's Well That Ends Well, IV.iii.98-99
=========================================================================
Date:         Wed, 23 Jul 1997 13:05:03 -0400
Reply-To:     "International Conference on the Principles and Future Dev. of
              AACR" <AACRCONF@INFOSERV.NLC-BNC.CA>
Sender:       "International Conference on the Principles and Future Dev. of
              AACR" <AACRCONF@INFOSERV.NLC-BNC.CA>
From:         "Jean L. Hirons" <jhir@LOC.GOV>
Subject:      The work and authority records
MIME-Version: 1.0
Content-Type: TEXT/PLAIN; charset=US-ASCII

Colleagues:

Following are my thoughts on the concept of the work and
the idea of an authority record for the work.

I've read the papers by Hagler, Yee, and Delsey and find them all
quite intriguing.  I'm very happy to see that they bring out some
of the points that Crystal and I would like to have covered but
couldn't (44 pages was long enough!).  What really interests me
is the fact that all of the papers seem to have similar threads
running among them.  Perhaps this is not so astonishing since we
all had to comment on the FRBR model, but, I still am pleased to
see that there seems to be some agreement on basic issues.

The aspects of most interest to those of us in serials, that
couldn't be covered in our paper, are 1) the concept of work and
what that means for a serial, 2) the potential for an authority
record for the work, 3) the aspect of corporate body authorship
for serials, and 4) the future of uniform titles (of both the
distinguishing and uniquing types).  I will address the first two
in this message and the second two in a separate message.

1. Concept of the work.  The FRBR model did not make it clear
what a serial work is: whether each separate title within a
succession constituted a new work or whether all of the titles
together constituted a work.  In her definition, Yee notes that a
name can change without there being a new work.  Thus, from her
definition, we would assume that serial title changes would not
constitute a new work.  But what about absorptions, mergers, and
splits?  I would assume that at least the latter two would be new
works.

What about changes of corporate body when the serial is entered
under the body?  Yee also says that the author can change without
causing a new work.  Would we want to follow this in all cases?

In our discussions on title changes, the CONSER AACR Review Task
Force tried to consider what would constitute a title change if
we were to create a separate record only when there was a new
work.  Thus, the change would have to be significant and signal a
change in scope or content.  While this is ideal, however, we
recognize that it isn't very practical. And we still need to
record the minor changes as well for purposes of access.
Nevertheless, the idea of "significant changes" is one that we
should pursue.

>From what I hear from other serialists, users want the holdings
in one place, regardless of title changes.  Thus, treating the
entire serial as a work, with all of its successive changes, is
no doubt a good idea.

But the use of ISSN, which is assigned to each title and the need
for small, clear records has led us to believe that successive
entry is still a good thing.  Can we have our cake and eat it to?
Perhaps the authority record is the solution???


2. Authority records for works

All three papers suggest this idea, Hagler specifically in
recommendation 2 and Delsey by suggesting the repackaging of work
and manifestation level data into separate records.  Yee notes
the problem for serials when one of the titles isn't held by a
library and the earlier and later titles cannot be linked,
suggesting that this is not an appropriate way to handle serial
works.

The idea of an authority record has much appeal. For serials, we
might see it as an umbrella record to which the successive
records for title changes would be attached.  But would we need a
multi-level record for the holdings (this is beginning to sound
like a familiar mulver discussion!) or could holdings for each
physical version be tied together through the use of the
authority record rather than being separated by title?  The first
question is what would be the title for this authority record???
The earliest would probably make most sense since this is an
authority record and not the primary public access record.  All
later titles would be included of course with corresponding
designations.

In discussing conference publications last year, the potential
for such a record became quite clear.  If we had this type of a
structure we wouldn't have to make the decision of monograph or
serial?  We could have both.  We could use the authority record
to house the collective call number and give the serial aspects,
while appending small monographic records with all of the detail
that everyone wants for access.

The ISSN record is intended as an authority record of a sort for
serials.  However, the ISSN record is tied to key title and there
must be a separate record for each ISSN.  Thus, at present, it
does not serve the function of bringing together the complete run
of a serial with all of its changes.  More thought should be
given to this, however, particularly as we deal with the
possibility of assigning ISSN to continuously-updated databases
for which we don't want to create successive entry records.

Another possible result of using such an authority record is that
it might free us up to make fewer title changes.  If the
authority record listed the earliest title and only those titles
significant enough to warrant a new record and the individual
records listed all of the title "variations" we would end up with
a hierarchy of title changes which, though potentially confusing,
might be the only practical way to approach the problem.  It
would give the user the access they need while providing the
librarian and user with a complete picture of the serial.

The idea of repackaging the work-related and manifestation- or
item-related details into separate records, as Delsey suggests,
is very appealing.  But Regina Reynolds and I tried out such an
exercise when we were reviewing the FRBR document last summer and
found that it wasn't quite as simple as it sounded.  One fear was
that the resulting records might be more complex and less useful
to the user, but that remains to be seen.  Our principle concern
was the difficulty of the cataloger in separating the data.

An important question is whether such an approach, which would
probably mean more work, is warranted for everything?  Hagler's
distinctions between the importance of "work" for scientific and
humanities publications is a valid one.  But even with scientific
publications we have serial titles changes and varying physical
manifestations.  So perhaps the answer is yes?

Jean Hirons
Acting CONSER Coordinator
Library of Congress
=========================================================================
Date:         Wed, 23 Jul 1997 23:09:49 -0400
Reply-To:     mac@slc.bc.ca
Sender:       "International Conference on the Principles and Future Dev. of
              AACR" <AACRCONF@INFOSERV.NLC-BNC.CA>
From:         "J. McRee Elrod" <mac@SLC.BC.CA>
Organization: Special Libraries Cataloguing, Inc.
Subject:      Issues related to seriality
Comments: cc: autocat@ubvm.cc.buffalo.edu

"Issues related to seriality" (44 p.) by Jean Hirons and Crystal Graham
is one of the four papers already posted to the future of AACR conference
site (http://www.nlc-bnc.ca/jsc/confpap.htm).

Much attention is given by Hirons and Graham to shifting the definition
of serials to include some continuing works now considered monographic,
such as loose-leaf services.  As a user of Catss, this discussion (like
the earlier format integration of MARC) is of little relevance to us.
The change in definition proposed would affect the coding of one fixed
field on Catss, and that is about it.

On the other hand, if the inclusion of more material as serials leads to
abandoning of AACR provisions in more records as CONSER is prone to do
(such as not always providing a date in imprint), my customers will not
be pleased.  Almost no law library retains the first instance of a loose-
leaf service, for example; but they do want the beginning date of the
service in 260$c.

I wonder if MARC powers that be in general, and other bibliographic
utilities in particular, should not take a look at Catss' single array of
fixed fields, and absence of templates (work forms).  It makes some
distinctions now made by rules, and discussed by Hirons and Graham, much
less important.

The distinction made between numbered and unnumbered series is also of
little import to our customers. If the individual issues of a series have
distinctive titles, our customers want them individually catalogued as
monographs, with series title added entries.  Whether they wish them
classed as a collection or a monographs also seems to have no relation to
whether the series is numbered, and varies from library to library more
than most such choices.  (For unnumbered series classed as a collection
we add a second Cutter based on main entry.)

Hirons and Graham make some welcome suggestions, including: basing serial
description on the file as opposed to first issue alone, consider
identification as opposed to transcription the main goal, base
description for services which supercede on latest iteration, use latest
entry for updating publications, and modify rules for title changes to
reduce the number of successive entries (would that mean we get MARC
field 247 back?).

These are all things we have been forced to adopt by client demand.  It
will be nice if rules changes make honest cataloguers of us.

In the discussion of the future of AACR, I hope the conference will not
neglect the impact of proposed changes on MARC and subject analysis.  Too
often these three proceed without reference to each other.


   __       __   J. McRee (Mac) Elrod (mac@slc.bc.ca)
  {__  |   /     Special Libraries Cataloguing   HTTP://www.slc.bc.ca/
  ___} |__ \__________________________________________________________
=========================================================================
Date:         Thu, 24 Jul 1997 14:05:12 -0400
Reply-To:     mac@slc.bc.ca
Sender:       "International Conference on the Principles and Future Dev. of
              AACR" <AACRCONF@INFOSERV.NLC-BNC.CA>
From:         "J. McRee Elrod" <mac@SLC.BC.CA>
Organization: Special Libraries Cataloguing, Inc.
Subject:      Choice and form of main entry
Comments: cc: autocat@ubvm.cc.buffalo.edu

With our customer mix of card catalogues, printed catalogues, single
entry new title lists, and/or OPACs, the choice of main entry *does*
matter, conventional wisdom not withstanding.

Choice of main entry has so far not received the amount of discussion I
would have expected on the future of AACR list.

Perhaps the discussion of work vs. item will lead to a reconsideration of
some present rules.

A title such as Smith on torts 10th ed. by Tom Jones, now gets main entry
under Jones.  Examining the text of successive editions of such titles
leads me to see Jones' role as the editor of this particular edition of
an ongoing work, as opposed to being its author.  Smith is the author of
the "work" of which this is the 10th ed.  If it is felt that authorship
of the work has become too diverse for entry under Smith's personal name,
it would seem better to me to move to main entry under title rather than
main entry under an editor who may be transitory in the ongoing history
of the work.  At least that would not hide the title from law firm
members in a single entry new title list.

Further discussion of corporate main entry seems needed to me.  While
main entry under corporate body seems appropriate for something like an
annual report which is both by and about the body, the mix of entry for
such things as law reform commission reports confuses patrons.  Patrons
in front of a catalogue, or cataloguers working from fore matter, have no
way of knowing whether a law reform commission report contains "official"
recommendations or is just informational.  Such reports are not the work
of whole body as are reports of royal commissions, or proceedings of a
conference, which spawn their work and die, but rather are issued by an
ongoing body.  They are not prepared by the body as a whole, but rather
by individuals or a committee.  If there were more than three personal
authors, title main entry seems appropriate to me.

Speaking of conferences, the overwhelming majority of "conferences" we
catalogue are not conferences at all.  They are continuing legal
education seminars.

"Made up" main entries have hardly been mentioned.  The ones most
disliked by my customers are those for treaties.  Catalogue users have no
idea which country is first in the English alphabet without knowing which
countries signed the treaty, usually don't know that is the basis of
choice for main entry, and certainly don't know the date of signing (the
two elements of the constructed main entry).  A uniform title such as
"North American Free Trade Agreement" seems much more patron friendly to
me.  (I will leave the subject of sacred works constructed main entries
to Hal from downunder.)  Perhaps it is time for constructed entries to
become added entries at most.

Simplification of main entry choice and form would speed cataloguing
greatly.  Whether we remain with the rule of three, or move to main entry
under a single author, title entry for all else, does not matter greatly
to me.  What does matter is being able to quickly determine (whether by
patron or cataloguer) that main entry.

   __       __   J. McRee (Mac) Elrod (mac@slc.bc.ca)
  {__  |   /     Special Libraries Cataloguing   HTTP://www.slc.bc.ca/
  ___} |__ \__________________________________________________________
=========================================================================
Date:         Thu, 24 Jul 1997 14:05:50 -0400
Reply-To:     mac@slc.bc.ca
Sender:       "International Conference on the Principles and Future Dev. of
              AACR" <AACRCONF@INFOSERV.NLC-BNC.CA>
From:         "J. McRee Elrod" <mac@SLC.BC.CA>
Organization: Special Libraries Cataloguing, Inc.
Subject:      Series
Comments: cc: autocat@ubvm.cc.buffalo.edu

Apart from their mention (numbered or unnumbered) as serials, series have
hardly been mentioned in the future of AACR list.

If the concept of "work" is applied to series, I can see my customers'
objection to series added entries under personal names the same as the
main entry for the title as a lost cause.

I've no objection to a series catalogued as a serial having personal main
entry if all issues were written by a single person.  But my customers
want series *added* entry as a *title* added entry.  It know it will
never happen, but I would like to see series as added entry divorced from
the work authority or catalogue record for the series as a serial.
=========================================================================
Date:         Fri, 25 Jul 1997 09:12:07 +1100
Reply-To:     Hal.Cain@ormond.unimelb.edu.au
Sender:       "International Conference on the Principles and Future Dev. of
              AACR" <AACRCONF@INFOSERV.NLC-BNC.CA>
From:         Hal Cain <hecain@ORMOND.UNIMELB.EDU.AU>
Organization: Joint Theological Library, Parkville, Victoria, Australia.
Subject:      Re: Series
Comments: To: mac@slc.bc.ca
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit

J. McRee Elrod wrote:
>
> Apart from their mention (numbered or unnumbered) as serials, series have
> hardly been mentioned in the future of AACR list.
>
> If the concept of "work" is applied to series, I can see my customers'
> objection to series added entries under personal names the same as the
> main entry for the title as a lost cause.
>
> I've no objection to a series catalogued as a serial having personal main
> entry if all issues were written by a single person.  But my customers
> want series *added* entry as a *title* added entry.  It know it will
> never happen, but I would like to see series as added entry divorced from
> the work authority or catalogue record for the series as a serial.

I'm still reading the conference papers (as available so far) and can't
claim to have them under my belt yet.  But IMO series is one form of the
work/manifestation dichotomy which seems to be one of the key issues to
be addressed.

When a work is catalogued in its own right, access is provided under a
full range of appropriate access points.  When a heading for that work
is provided as a secondary access point in a record either for another
work, or for an individual manifestation, conventional wisdom has
decreed that the main entry form (principal access point, or whatever)
must be the form used.

Some of what I have been reading (on the train -- sorry I can't cite the
reference at this point) has pointed to the fact that the way the data
is entered in the record *need not* correspond absolutely to what the
searcher sees in the display (or, presumably, in the printed-catalogue
page some of Mac's customers have -- Mac has explained very clearly to
readers of the Autocat list just why AACR2 headings for series under
personal author headings are rejected by them and effectively deny them
access under the series title).  So perhaps for cases like Mac's the
solution is to provide for more flexibility in output.

And this in turn points to the need to ensure that the USMARC
specifications enable us to construct the output we need for our very
different clients, on the basis of a standard source -- a topic related
to but strictly outside the scope of the conference and, presumably,
this list.

In respect of series I have another problem.  If a series produced by an
editor is catalogued as an entity there will be added entries for the
editor(s -- up to three, anyway, another practice which I think could be
challenged, but that's another matter).  When the series heading is
added to another record as a secondary access point, no access is
provided under the series editor's name.  Yet many scholars have made
important contributions by virtue of their editorship of series, and it
is clear that editing a series may constitute a form of intellectual
responsibility far more significant than compiling a collection of
existing articles.  Unless there is a record for the series as an entity
there is no link from the person to the work.

Hal Cain, Joint Theological Library, Parkville, Victoria, Australia
<hecain@ormond.unimelb.edu.au>
=========================================================================
Date:         Thu, 24 Jul 1997 20:10:16 +0000
Reply-To:     "International Conference on the Principles and Future Dev. of
              AACR" <AACRCONF@INFOSERV.NLC-BNC.CA>
Sender:       "International Conference on the Principles and Future Dev. of
              AACR" <AACRCONF@INFOSERV.NLC-BNC.CA>
From:         John Attig <jca@PSULIAS.PSU.EDU>
Subject:      Series, authority records and beyond
In-Reply-To:  <33D7D337.F74@ormond.unimelb.edu.au>
MIME-version: 1.0
Content-type: text/plain; charset="us-ascii"

At 09:12 AM 7/25/97 +1100, Hal Cain <hecain@ormond.unimelb.edu.au>  wrote:
>In respect of series I have another problem.  If a series produced by an
>editor is catalogued as an entity there will be added entries for the
>editor(s -- up to three, anyway, another practice which I think could be
>challenged, but that's another matter).  When the series heading is
>added to another record as a secondary access point, no access is
>provided under the series editor's name.  Yet many scholars have made
>important contributions by virtue of their editorship of series, and it
>is clear that editing a series may constitute a form of intellectual
>responsibility far more significant than compiling a collection of
>existing articles.  Unless there is a record for the series as an entity
>there is no link from the person to the work.
>
>Hal Cain, Joint Theological Library, Parkville, Victoria, Australia
><hecain@ormond.unimelb.edu.au>

In fact, there is a way of doing this -- which brings up another theme that
seems to be emerging -- the role of authority records in our information
systems.  In dealing with a series, it is quite possible to have an
authority record for the series title that includes the editors as see from
references, just as it is possible to have a serial bibliographic record
with the editors as added entries.  The tricky thing is for an information
system to deal in a meaningful fashion with these two very different ways
of giving the same information.  And, to anticipate Mac's response, we need
to recognize that different kinds of information systems (and I do include
various kinds of paper catalogues) will use different techniques to
accomplish this.

And that raises a perennial issue that I absolutely refuse to even try to
answer:  Just what is it that AACR2 is attempting to standardize?  Is it a
standard for recording information or a standard for displaying information
or something else altogether?  And what does each of these concepts mean in
practice?  If AACR2 is not to be taken as a standard for display, then we
have to look elsewhere for the solution to many of the problems we are
raising.  I hope that in the course of the discussions leading up to and
following the International Conference we can try to achieve a better
understanding of this thorny issue.  So much of what we are trying to
accomplish depends on coming to some sort of conclusion.
=========================================================================
Date:         Fri, 25 Jul 1997 10:28:17 +1000
Reply-To:     "International Conference on the Principles and Future Dev. of
              AACR" <AACRCONF@INFOSERV.NLC-BNC.CA>
Sender:       "International Conference on the Principles and Future Dev. of
              AACR" <AACRCONF@INFOSERV.NLC-BNC.CA>
From:         Giles S Martin <ulgsm@DEWEY.NEWCASTLE.EDU.AU>
Subject:      Re: Series
In-Reply-To:  <33D7D337.F74@ormond.unimelb.edu.au>
Mime-Version: 1.0
Content-Type: TEXT/PLAIN; charset=US-ASCII

Hal Cain's comment (quoted at the end of this rather long message) is the
heart of the problem for series, and for other entities which occur as
secondary entities.  However, I think that more precisely the problem
may be expressed this way:

AACR records (whether they are 3x5 cards, or MARC bibliographic records)
represent manifestations: either unique objects (such as manuscripts or
Web sites), or the nearly indistinguishable results of mass production
(such as printed books and compact discs).  They don't represent works,
but rather objects which may contain a single work, contain several
works, or contain part of a work (as in the case of an item in a series).

In AACR records works are represented by headings -- either uniform title
headings, or name-title headings.  This is obscured by the practice of
omitting the uniform title when it is identical to the title proper (in
MARC terms, you don't have a 240 field if it would be the same as the
title proper in the 245 field).  It is alspo obscured by the fact that in
many (or even most) records, you have the single physical manifestation
of a work, which will never be reissued again.

So the catalogue created using AACR has multiple access points for
manifestations, but strictly speaking only one access point for a work.
For example, in this MARC record:

 245  00 Anglo-American cataloguing rules /|cprepared under the direction of
         the Joint Steering Committee for Revision of AACR, a committee of
         the American Library Association, the Australian Committee on
         Cataloguing, the British Library, the Canadian Committee on
         Cataloguing, the Library Association, the Library of Congress ;
         edited by Michael Gorman and Paul W. Winkler.
 250     2nd ed., 1988 revision.
 260  0  Ottawa :|bCanadian Library Association ;|aChicago :|bAmerican
         Library Association,|c1988.
 300     xxv, 677 p. ;|c26 cm.
 500     First ed. published (1967) in two versions under the following
         titles: Anglo-American cataloging rules. North American text;
         Anglo-American cataloguing rules. British text.
 504     Includes bibliographical references and index.
 650   0 Descriptive cataloging|xRules
 700  10 Gorman, Michael,|d1941-
 700  10 Winkler, Paul W.|q(Paul Walter)
 710  20 Joint Steering Committee for Revision of AACR
 710  20 American Library Association

The access point for the work is in the 245 field: "Anglo-American
cataloguing rules".  On the other hand, the manifestation has access
points in the 245 field, the 650 field, the 700 fields and the 710 fields.
It just doesn't feel that way, because those access points are equally
applicable to the work: if you had a catalogue record for the work
(whether it was the work AACR2R, or the work AACR including all
editions and versions), then all those access points would be useful and
relevant.

As far as the user is concerned, the distinction between manifestation
and work doesn't matter here: the user looking for the work under any of
the access points will find the record for the manifestation which
includes the work.

However, it does matter in many cases.  One is for series entered under
personal author, where (for example) "The Oxford Shakespeare" gets a
heading like: "Shakespeare, William, 1564-1616. Works. 1982".  Even if
you have an OPAC with the right authority records present, it can be hard
for users to find their way from the title proper to the records for the
manifestations in the series.

Another example is translated works.  Recently, we had a user asking on
our OPAC's suggestion section about a work known as "Freud, S. A general
introduction to psychoanalysis".
  * The first problem is that it was published under the title "A
general introduction to psycho-analysis", so that keywords fail here.
  * The second problem is that it is a translation, so it files in our
OPAC under the uniform "Vorlesungen zur Einfuhrung in die Psychoanalyse.
English" when you do your author search under Freud.
  * The third problem is that we have other copies of this work, translated
by a different person into English, under three different titles: "The
complete introductory lectures on psychoanalysis", "Introductory lectures
on psycho-analysis" and "Introductory lectures on psychoanalysis" (note
the hyphen again in the second title!)  Searches under any of these
titles will only give you one manifestation, not the other manifestations
of the work.
  * The fourth problem is that we have yet one more copy of the work: as
volumes 15-16 of "The standard edition of the complete psychological
works of Sigmund Freud".  We do have a contents note for this, so you
could find it under a keyword search in our OPAC (as long as you remember
that "psycho-analysis" is hypenated); but we don't provide access points
for any of the works her under author or title.

This is not an unusually complex collection of manifestations of a work
-- but the present combination of AACR, USMARC, and our OPAC software
make it very hard for a normal user to find all of them.  In fact, I
suspect that most librarians here would have missed some of them in the
search.

What I think is needed is a new defiunition of a record for a work, which
would have some of the characteristics of a bibliographic record, and
some of the characteristics of an authority record.  Users of an OPAC
would be lead first to the record for the work, and then to all of its
manifestations.  I don't think this idea is original to me: it's implicit
in much of what Ronald Hagler and Martha Yee say in their papers, and
Rahmat Fattahi had a model like this in his "super records".

Giles

          ####    ##       Giles Martin
       #######   ####      Quality Control Section
     #################     University of Newcastle Libraries
   ####################    New South Wales, Australia
   ###################*    E-mail: ulgsm@dewey.newcastle.edu.au
    #####      ## ###      Phone:   +61 49 215 828 (International)
                           Fax:     +61 49 215 833 (International)
                  ##
 The web of our life is of a mingled yarn, good and ill together
                        -- All's Well That Ends Well, IV.iii.98-99

On Fri, 25 Jul 1997, Hal Cain wrote:

> When a work is catalogued in its own right, access is provided under a
> full range of appropriate access points.  When a heading for that work
> is provided as a secondary access point in a record either for another
> work, or for an individual manifestation, conventional wisdom has
> decreed that the main entry form (principal access point, or whatever)
> must be the form used.
=========================================================================
Date:         Thu, 24 Jul 1997 17:44:58 -0400
Reply-To:     mac@slc.bc.ca
Sender:       "International Conference on the Principles and Future Dev. of
              AACR" <AACRCONF@INFOSERV.NLC-BNC.CA>
From:         "J. McRee Elrod" <mac@SLC.BC.CA>
Organization: Special Libraries Cataloguing, Inc.
Subject:      Re: Series
Comments: To: hecain@ormond.unimelb.edu.au
In-Reply-To:  <33D7D337.F74@ormond.unimelb.edu.au>

Hal said:

>When a work is catalogued in its own right, access is provided under a
>full range of appropriate access points.  When a heading for that work
>is provided as a secondary access point in a record either for another
>work, or for an individual manifestation, conventional wisdom has
>decreed that the main entry form (principal access point, or whatever)
>must be the form used.

When dealing with an OPAC, it should be possible to map 4XX$t (in
older records) and 8XX$t to title added entry.  Perhaps we should look
at doing the same in our print catalogue programs, and just not use
4XX$a and 8XX$a for anything, since they duplicate the 1XX.  Then we
could live with what to us is a silly rule and its useless and even
sillier reflection in MARC coding.

Hal above expands the questions to all related works, not just series.
I hesitate to suggest to our programmer that he drop of 7XX$a when,
and *only* when, it duplicates the 1XX, and map the 7XX$t to title
added entry.  But that is what our book catalogue customers would like.
(The person in 7XX$a does not always duplicate 1XX as is the case with
4XX and 8XX.)

It has already been pointed out that MARC records and OPACs are more
accepting of redundancy in a record than were catalogue cards.  Printed
catalogues are much *less* accepting of redunancy in access points.

Mac

   __       __   J. McRee (Mac) Elrod (mac@slc.bc.ca)
  {__  |   /     Special Libraries Cataloguing   HTTP://www.slc.bc.ca/
  ___} |__ \__________________________________________________________
=========================================================================
Date:         Fri, 25 Jul 1997 00:30:54 -0500
Reply-To:     "International Conference on the Principles and Future Dev. of
              AACR" <AACRCONF@INFOSERV.NLC-BNC.CA>
Sender:       "International Conference on the Principles and Future Dev. of
              AACR" <AACRCONF@INFOSERV.NLC-BNC.CA>
From:         Ralph Papakhian <papakhi@INDIANA.EDU>
Organization: Indiana University W&G Cook Music Library
Subject:      Re: Series
In-Reply-To:  <Pine.3.89.9707250906.B4842-0100000@dewey.newcastle.edu.au> from
              "Giles S Martin" at Jul 25, 97 10:28:17 am
Content-Type: text

Hello.
I'm not going to address all of Giles' points. But in the case quoted below,
even the current authority record is adequate to meet the user's
question--the problem may be more with the opac interface (none
of which seem to correspond to AACR2 regarding filing (as in
ALA filing rules). <current authority record inserted below>

A somewhat related problem is the very antique idea that cataloging
rules and current authority control structures are really derived
from a Western European academic/scholarly citation tradition--namely
that books (and corollaries) are cited by author and then title.
That tradition seems now to be on the wane, which certainly presents
many new challenges (and one of the reasons for the AACRCONF).
In the meantime Giles should arrange to add the references that
are not already appearing. But, it looks like most of the required
references are already present in this authority record created
in 1988! I think Giles is really talking about an OPAC problem.
You see, Giles says "searches under these titles"... when of course
catalog users should really be searching for authors/titles.

That brings me to another point, namely that much of the concern
about cataloging rules now is not really about the rules, but
about their presentation in catalogs (aka OPACS). AACR never really
addressed that matter, and other national and international
standards bodies have not created standards for OPACS (except
in the psuedo-standard, Z39.50, designed to account for the lack
of standards).
I submit that much of the argument and brouhaha about AACR is
fundamentally an expression of displeasure with current OPAC
technology of even adequately presenting existing cataloging data not
to mention the proposed new and improved cataloging data.

--ralph papakhian

Ü   1  010     n  88610867  Ý
Ü   2  040     ICU ßc ICU Ý
Ü   3  005     19880216105218.9 Ý
Ü   4  100 10  Freud, Sigmund, ßd 1856-1939. ßt Vorlesungen zur
Einfèuhrung in
die Psychoanalyse. ßl English Ý
Ü   5  400 10  Freud, Sigmund, ßd 1856-1939. ßt General introduction
to
psychoanalysis Ý
Ü   6  400 10  Freud, Sigmund, ßd 1856-1939. ßt General introduction
to psycho-
analysis Ý
Ü   7  400 10  Freud, Sigmund, ßd 1856-1939. ßt Complete introductory
lectures
on psychoanalysis Ý
Ü   8  400 10  Freud, Sigmund, ßd 1856-1939. ßt Introductory lectures on
psychoanalysis Ý
Ü   9  400 10  Freud, Sigmund, ßd 1856-1939. ßt Introductory lectures on
psycho-analysis Ý
Ü  10  670     LCCN 20-12205: His A general introduction to
psychoanalysis,
1920. Ý
Ü  11  670     LC data base, 2-2-88 ßb His A general introduction to psycho-
analysis, 1935; His The complete introductory lectures on psychoanalysis,
1966; His Introductory lectures on psychoanalysis, 1977, c1966; His
Introductory
lectures on psycho-analysis, c1970. Ý

 Giles S Martin said ........

> Another example is translated works.  Recently, we had a user asking on
>
> our OPAC's suggestion section about a work known as "Freud, S. A general
> introduction to psychoanalysis".
>   * The first problem is that it was published under the title "A
> general introduction to psycho-analysis", so that keywords fail here.
>   * The second problem is that it is a translation, so it files in our
> OPAC under the uniform "Vorlesungen zur Einfuhrung in die Psychoanalyse.
> English" when you do your author search under Freud.
>   * The third problem is that we have other copies of this work, translated
> by a different person into English, under three different titles: "The
> complete introductory lectures on psychoanalysis", "Introductory lectures
> on psycho-analysis" and "Introductory lectures on psychoanalysis" (note
> the hyphen again in the second title!)  Searches under any of these
> titles will only give you one manifestation, not the other manifestations
> of the work.
>   * The fourth problem is that we have yet one more copy of the work: as
> volumes 15-16 of "The standard edition of the complete psychological
> works of Sigmund Freud".  We do have a contents note for this, so you
> could find it under a keyword search in our OPAC (as long as you remember
> that "psycho-analysis" is hypenated); but we don't provide access points
> for any of the works her under author or title.
>
> This is not an unusually complex collection of manifestations of a work
> -- but the present combination of AACR, USMARC, and our OPAC software
> make it very hard for a normal user to find all of them.  In fact, I
> suspect that most librarians here would have missed some of them in the
> search.
>
> What I think is needed is a new defiunition of a record for a work, which
> would have some of the characteristics of a bibliographic record, and
> some of the characteristics of an authority record.  Users of an OPAC
> would be lead first to the record for the work, and then to all of its
> manifestations.  I don't think this idea is original to me: it's implicit
> in much of what Ronald Hagler and Martha Yee say in their papers, and
> Rahmat Fattahi had a model like this in his "super records".
>
> Giles
>
>           ####    ##       Giles Martin
>        #######   ####      Quality Control Section
>      #################     University of Newcastle Libraries
>    ####################    New South Wales, Australia
>    ###################*    E-mail: ulgsm@dewey.newcastle.edu.au
>     #####      ## ###      Phone:   +61 49 215 828 (International)
>                            Fax:     +61 49 215 833 (International)
>                   ##
>  The web of our life is of a mingled yarn, good and ill together
>                         -- All's Well That Ends Well, IV.iii.98-99
>
> On Fri, 25 Jul 1997, Hal Cain wrote:
>
> > When a work is catalogued in its own right, access is provided under a
> > full range of appropriate access points.  When a heading for that work
> > is provided as a secondary access point in a record either for another
> > work, or for an individual manifestation, conventional wisdom has
> > decreed that the main entry form (principal access point, or whatever)
> > must be the form used.
>


--
A. Ralph Papakhian, Indiana University Music Library
Bloomington, IN 47405 812/855-2970 papakhi@indiana.edu
co-owner: MLA-L@listserv.indiana.edu
=========================================================================
Date:         Fri, 25 Jul 1997 16:32:56 +1000
Reply-To:     "International Conference on the Principles and Future Dev. of
              AACR" <AACRCONF@INFOSERV.NLC-BNC.CA>
Sender:       "International Conference on the Principles and Future Dev. of
              AACR" <AACRCONF@INFOSERV.NLC-BNC.CA>
From:         Giles S Martin <ulgsm@DEWEY.NEWCASTLE.EDU.AU>
Subject:      Re: Series
In-Reply-To:  <9707250530.AA14944@browndwarf.ucs.indiana.edu>
Mime-Version: 1.0
Content-Type: TEXT/PLAIN; charset=US-ASCII

Ralph Papakhian is right that a large part of the problem is the way that
our OPAC displays the data -- the information is already available within
the MARC record.  He also rightly says that our cataloging rules are
based on a tradition of citation of books under the name of the author.

However, our users seem to prefer title access.  About 21.5 % of searches
are for author, while about 43 % are for title (including about 8.5 % in
our "journal title" index, and about 34.5 % in our general title index).

Indeed, title access may often be far more efficient, especially for
prolific authors.  Users of our catalogue get more that 1300 hits under
"Mozart, Wolfgang Amadeus"; they get just 3 hits under "Jupiter
symphony", one of which is an authority record.  So given the time
required to process a large number of hits, a title search can be more
rational.

Giles

          ####    ##       Giles Martin
       #######   ####      Quality Control Section
     #################     University of Newcastle Libraries
   ####################    New South Wales, Australia
   ###################*    E-mail: ulgsm@dewey.newcastle.edu.au
    #####      ## ###      Phone:   +61 49 215 828 (International)
                           Fax:     +61 49 215 833 (International)
                  ##
 The web of our life is of a mingled yarn, good and ill together
                        -- All's Well That Ends Well, IV.iii.98-99
=========================================================================
Date:         Fri, 25 Jul 1997 11:32:03 METDST
Reply-To:     "International Conference on the Principles and Future Dev. of
              AACR" <AACRCONF@INFOSERV.NLC-BNC.CA>
Sender:       "International Conference on the Principles and Future Dev. of
              AACR" <AACRCONF@INFOSERV.NLC-BNC.CA>
From:         Bernhard Eversberg <EV@BUCH.BIBLIO.ETC.TU-BS.DE>
Organization: UB der TU-Braunschweig
Subject:      Works: AACR & Metadata

"Works" in AACR and Metadata : A few thoughts
---------------------------------------------

The AACR Future Development  papers now debated in this list contain
laudable attempts to come to terms with contemporary cataloging issues.
Especially, what merits everybody's attention are new approaches at the
concept of the "work" (definition of which had been deliberately (?) omitted
from AACR (M.M.Yee's paper, p.34)), and at the relationships between works and
whether or not, and how, these can or should be addressed by code and format.
In the wider bibliographic universe, beyond the MARC galaxy, some of these
concepts have been implemented in operative systems, but as everybody knows
intergalactic journeys are not technically feasible, so let's leave
this aside here.
Apart from that, there are a few more questions begging for an answer when
looking at present-day realities in information habits:

Who's going to turn to library databases when looking for electronic
documents? Doesn't everybody habitually fire up NetScape or MSIE and call
their favorite search engine? Has this not, for many users, even
become their first stop in the information seeking process?
And then, who - except hopefully librarians - can name the differences
between search engines and OPACs? The main difference is, of course, that
search engines exercise not an ounce of intellectual analysis on any single
document. What they do is not cataloging, it is just mechanical indexing,
albeit in maybe very ingenious ways. The fact that very often SOME useful
links show up instantaneously at the top of a "relevance ranked" list can
indeed deceive simple (and quite a few not so simple) minds into tacitly
assuming that it will take only a few improvements in indexing technology to
perfection the search engines into real "knowbots", which will serve up
ALL the relevant items for whatever the question is, AND place the real most
relevant items at the top. Search engines are even more overrated than library
catalogs are underrated, but both are not properly understood.
For users, the main difference between search engines and OPACs is probably
that the former bring up references to resources that are only mouseclicks
away. This creates the danger that resources further or much further away
than that (like, e.g., books) are losing market value and support - and
libraries with them?

It is not quite as bad. Serious users know a few things about the short-
comings of search engines, and that software improvement alone will never
be enough to overcome them. That's why the "metadata" concept was invented.
Embedded metadata elements, something like electronic CIP records, are
supposed to provide the missing search terms not found in a document "as is".
(But look at examples of metadata. You are in for surprises. They are not
being created by national bibliographic agencies, after all, but by
individuals primarily interested in drawing attention to their products.)

Now WE may all be aware that even Dublin Core compatible metadata can still
be far from the rigorous standards of data quality, controlled vocabulary,
authority controlled access points we have come to understand as necessary
for large databases to function well - but how do we get this across?
Information seekers nowadays have come or been made to believe that there
are quick and easy (and inexpensive) answers to all information needs
on the Web, or if not now, then soon. If you wonder what I'm talking about,
try to tell any non-librarian out there who is using or even running Web
services that what we need is more than the Dublin Core list of 15 elements
but show them the USMARC documentation and AACR2. Try to explain what
"collocation" is and why it is a good idea but not an easy one. Try to
make it understood that keyword searching is an inherently inadequate
concept for the document universe.

Most of us are aware that if we want to go on running libraries we have
to become players in new fields. For cataloging, this means we have to
re-evaluate and adapt or reformulate our concepts and standards for new
types of resources and new types of information needs. The Future Development
papers do point into that general direction. From what I read so far, however,
much more will have to be done to make it clear, and not just to catalogers
or even librarians but to the "information society" in general, what it is
we can contribute. The most urgent practical aim, IMHO, would be to make
it obvious where Dublin Core is falling short, and here I see two main
aspects:

-- there is no "work" concept (little wonder when not even AACR has one)

-- there is no clear concept of collocation in the sense of bringing together
   what belongs together; real relevance as opposed to the formal or
   statistical grouping the search engines do

Within the Dublin Core framework, one might elaborate on the elements of
"identifier" and "relation" to implement the "work" idea. In terms of MARC,
the uniform title concept as embodied in the 700s might serve as a
starting point. These fields carry identifiers of works, after all. Others
in this list have already expressed the need for a more workable definition
of "uniform title" as a means of identifying works and collocate versions.
Relationships between works, whatever these are, must always be based on some
kind of identifiers. Thus, it would be a huge step forward could one
define a "work identifier", including a set of edition/version and relation
codes or indicators. Presently, relations are mostly expressed simply by
URLs, which are inadequate for this purpose for more than one reason.
IdNumbers like ISBNs assigned by neutral agencies cannot be a solution either
but what's needed is an unambiguous identifier constructed from the most
stable and readily ascertainable elements that come with a work. The
uniform title as we know it, alas, is not among these elements.

Clearly, databases the size of those in the search engines could profit
from such concepts even more than library catalogs. Will the upcoming
conference, among other things, deal with this matter or regard it as
out of scope? Will it be an insider event, or will the opportunity be
used to generate public awareness or even appreciation of what libraries
attempt to do, and that and why and how they can make a difference?

John Attig asked, "Just what is it that AACR attempts to standardize?"
If the Conference is going to answer questions, this one should be
the first on their list. Does it go without saying that the answer should
be intelligible by more people than just librarians?

(BTW: "work identifiers" of a sort have been in use for a very long time
in the Science Citation Index: this indexing tool derives a very brief
but unique identifier from the bibliographic data of a periodical article
and uses this to collocate all references to this article...)

Just a few thoughts,   B.E.
=========================================================================
Date:         Fri, 25 Jul 1997 09:32:38 EDT
Reply-To:     "International Conference on the Principles and Future Dev. of
              AACR" <AACRCONF@INFOSERV.NLC-BNC.CA>
Sender:       "International Conference on the Principles and Future Dev. of
              AACR" <AACRCONF@INFOSERV.NLC-BNC.CA>
From:         Carroll Nelson Davis <cnd2@COLUMBIA.EDU>
Subject:      Series collocation

        As Mac Elrod and others have suggested, there is room for
improvement of catalog entries for series in a revised cataloging code
-- addressing accumulated concerns for access by users and ease of
management by library staff.  But the suggestion that collective
records for a series be located one place in the catalog and the
corresponding series added entries on records for analytics be located
in a different place is unacceptable.  That would abort the collocation
function of the catalog and create access and control failures for both
users and staff in catalogs like ours here at Columbia University.
        The Columbia catalog has a great many series with collective
bibliographic records and separate analytic records.  These work in
concert, both for internal processing and user access.  When the
collective series record and the analytic records fail to collocate,
processing errors and misdirections result; processing units demand
that the collocation be fixed.  Public service units relay user
complaints and demand that the collocation be fixed.  Our catalog has
plenty of collective series records whose entries do not match the
currently established series headings -- the legacy of fast-and-dirty
retrospective conversion projects; cleaning those up is an ongoing
task, for which there is steady demand.  So, my basis for reaffirming
the need for series collective/analytic collocation is not just
theoretical.
        Arguments for different forms of series (or serial, for that
matter) entry are worthwhile.  Arguments for title entry only for
these deserve serious evaluation.  Rules for choice of entry in
general deserve as thorough an overhaul as anyone cares to give them.
        But collocation, including series collocation, as a catalog
requirement needs to be preserved.

                                Carroll Davis, Serials cataloger
                                Columbia University Libraries
                                cnd2@columbia.edu
=========================================================================
Date:         Fri, 25 Jul 1997 11:13:54 -0400
Reply-To:     "International Conference on the Principles and Future Dev. of
              AACR" <AACRCONF@INFOSERV.NLC-BNC.CA>
Sender:       "International Conference on the Principles and Future Dev. of
              AACR" <AACRCONF@INFOSERV.NLC-BNC.CA>
From:         Cynthia Watters <WATTERS@MYRIAD.MIDDLEBURY.EDU>
Subject:      Re: Series

Giles, Mac, and Hal have all put their fingers on important challenges
for the problem of the future, and I agree with all.  My two minor
contributions are:

1) Author-title entries (those banes of Mac's existence for series)
*should* be construed (for cards and OPACs) as two entries: one under
author and one under title.  That's actually the way we (and many
others) have indexed our OPAC, so, indeed, I LIKE it when the single
entry for a related work (series or whatever) is in author-title form
because it is retrievable under EITHER author or title.  And I think
it should be.  My biggest concern about title main-entry for serials
(especially those eminating from corporate bodies like Journal / American
Philosophical Society (made up example)) and works produced under
editorial direction is that: when an added entry is made, there is
no access by body or editor.

I know authority records can provide this, but not only do not ally
(that's 'all') systems have interactive authority systems (a criticism
I reject because they should and, sooner or later, I think they will),
but also because that doesn't retrieve the records when key word
searches are done.

The editors, indeed, are rarely even present as 400s in series authority
records, which I find totally unacceptable!  So even authority records
don't help for many of these.

I'd love to see something like every series or related work entry to
have both an author (or editor or body) element and a title element
and to display regardless of which is searched.  Mac's clients can
have their cards produced and their OPACs programed so they are indexed
separately and, therefore, the author portion can be supressed as a genuine
duplicate.  Personally, I feel that I would lose access under author if
I didn't see the title I was looking for even if that book is represented
under a different title on the hit list.  (But if I had to choose BETWEEN
author or title, I'd go with Mac--it's just that it shouldn't be either/
or)

I think one of the primary shortcomings of catalogs at this point is that
a search on an added entry form does NOT retrieve records for related
works that contain only the authorized form.

Besides the series, etc. stuff above, one major one is subject entries
for works, especially translated works.  A subject search by title
(assuming your catalog indexes $t's) still doesn't work if you're using
the English title and the entry uses the original French title or whatever.
Nor indeed do authority references redirect you since they go only from
author-title not from title.  Besides all the authority records only
go from Author. Translated title. to Author. Original title. Language.
and subject entries do not include the language.

Sorry to babble on, and my second point somehow got into my first point
so there's no 2) in my message (perhaps you can be grateful for small
favors :=) )

Cynthia Watters
Catalog Librarian
Middlebury College
watters@myriad.middlebury.edu
=========================================================================
Date:         Fri, 25 Jul 1997 08:35:19 -0400
Reply-To:     mac@slc.bc.ca
Sender:       "International Conference on the Principles and Future Dev. of
              AACR" <AACRCONF@INFOSERV.NLC-BNC.CA>
From:         "J. McRee Elrod" <mac@SLC.BC.CA>
Organization: Special Libraries Cataloguing, Inc.
Subject:      Re: Series collocation
Comments: To: cnd2@COLUMBIA.EDU
In-Reply-To:  <CMM.0.90.4.869837558.cnd2@labdien.cc.columbia.edu>

>        Arguments for different forms of series (or serial, for that
>matter) entry are worthwhile.  Arguments for title entry only for
>these deserve serious evaluation.  Rules for choice of entry in
>general deserve as thorough an overhaul as anyone cares to give them.
>        But collocation, including series collocation, as a catalog
>requirement needs to be preserved.

Carroll, if seriality (including series) equals title main entry, I
could live with that.  The findings Giles reports seem to point in that
direction.  The need for collocation by editor pointed out by Hal needs
to be met, perhaps by a series "work" authority record.  Collocation by
author could be achieved the same way.  The practical problem with
series main entry under author is that it creates redundant entries.
A cross reference under the author would reduce multiple entries for
the same title under the same entry to one additional line for them all.
The theoretical problem is, how does one know that the series will in
the future have that author?  The arguments which led to title main
entry for serials apply also to series, it seems to me.

Mac

   __       __   J. McRee (Mac) Elrod (mac@slc.bc.ca)
  {__  |   /     Special Libraries Cataloguing   HTTP://www.slc.bc.ca/
  ___} |__ \__________________________________________________________






>
>                                Carroll Davis, Serials cataloger
>                                Columbia University Libraries
>                                cnd2@columbia.edu
>
=========================================================================
Date:         Fri, 25 Jul 1997 10:51:51 -0400
Reply-To:     mac@slc.bc.ca
Sender:       "International Conference on the Principles and Future Dev. of
              AACR" <AACRCONF@INFOSERV.NLC-BNC.CA>
From:         "J. McRee Elrod" <mac@SLC.BC.CA>
Organization: Special Libraries Cataloguing, Inc.
Subject:      Standards for catalogues

One of the lessons emerging from several posts is the importance of
standards for catalogue construction and display, whether printed or
electronic. We had that with the unit card, but have not had it since.

This may be more important that any tinkering with the rules for record
creation, or their reflection in MARC.

Mac

   __       __   J. McRee (Mac) Elrod (mac@slc.bc.ca)
  {__  |   /     Special Libraries Cataloguing   HTTP://www.slc.bc.ca/
  ___} |__ \__________________________________________________________
=========================================================================
Date:         Fri, 25 Jul 1997 14:45:38 -0400
Reply-To:     "International Conference on the Principles and Future Dev. of
              AACR" <AACRCONF@INFOSERV.NLC-BNC.CA>
Sender:       "International Conference on the Principles and Future Dev. of
              AACR" <AACRCONF@INFOSERV.NLC-BNC.CA>
From:         "Michelle (Collins) Flinchbaugh" <mflinchb@UMICH.EDU>
Subject:      When good rules are bad
Mime-Version: 1.0
Content-Type: text/plain; charset="us-ascii"

Hello everyone,

While the primary purpose of the Toronto Conference is to consider the
development of more effective rules and rules which will better encompass
new media, I'm hoping that the participants as well as JSC will also
consider this fact: A person's ability to comprehend and apply ACCRII is
primarily a function of their understanding and experience with cards.

I'd like to modestly submit that even good rules become bad when
they are termed in a nearly obsolete technology (cards) which fewer and
fewer people understand.

Having been a cataloger for only three years, my knowledge and
experience is substantially less than all of yours. To put this in
perspective, I've never cataloged on cards. The university I attended
implemented it's first OPAC my freshman year of college. To a large extent,
I really don't understand cards and how they work.

Because of this, some concepts, like main entry, for example, have always
been  problematic concepts. The glossary of AACRII says that main entry is
"the complete cataloging record for an item, presented in the form by which
the entity is to be uniformly identified and cited. The main entry may also
include the tracings." Now with an OPAC, it's quite clear that ALL entries
show the complete cataloging record for the item. After some cataloging
experience, I began to figure that main entry must really mean something
along the line of the how the item would be referred to in indexes displays
and in other records. My point is that regardless of whether or not we need
to change the rules about main entry, PLEASE change the definition!

I do completely understand that the primary purpose of AACRII is to serve
as an international standard, and I do understand that it may be of
relatively little import if those who aren't familiar with cards can
understand it or not. However, when the most basic and fundamental concepts
are based on cards, our rules also lose relevance to library administrators,
programmers who design our systems, and anyone else who needs to understand
our rules and their function.

My hope is that even good rules might be revised if their relevance and
meaning is obscured by card based concepts.

Michelle Flinchbaugh
Serials Cataloger
The University of Michigan
=========================================================================
Date:         Fri, 25 Jul 1997 14:31:09 -0700
Reply-To:     "International Conference on the Principles and Future Dev. of
              AACR" <AACRCONF@INFOSERV.NLC-BNC.CA>
Sender:       "International Conference on the Principles and Future Dev. of
              AACR" <AACRCONF@INFOSERV.NLC-BNC.CA>
From:         Daniel CannCasciato <dcc@MUMBLY.LIB.CWU.EDU>
Subject:      AACR and system implementation
In-Reply-To:  <Pine.3.89.9707250906.B4842-0100000@dewey.newcastle.edu.au>
MIME-Version: 1.0
Content-Type: TEXT/PLAIN; charset=US-ASCII

Hello All,

[something of a rambling message follows]

I think Giles example:

> Another example is translated works.  Recently, we had a user asking on
> our OPAC's suggestion section about a work known as "Freud, S. A general
> introduction to psychoanalysis".
> ...

>   * The second problem is that it is a translation, so it files in our
> OPAC under the uniform "Vorlesungen zur Einfuhrung in die Psychoanalyse.
> English" when you do your author search under Freud.
>   * The third problem is that we have other copies of this work, translated
> by a different person into English, under three different titles:  ...

exemplifies another characteristic of cataloging today:  we have not been
unified in our requirements for OPAC software.  In applying the code,
AACR, many times catalogers are confused by the way their local systems
react to the record structure.  Or, many times certain aspects of record
structure seem redundant in an online environment.  (An added title entry
that matches the title portion of a name-title entry, both on the same bib
record.)

In Giles' example, how many OPACs have an authority module that usefully
implements authority records and leads patrons to the
works/editions/manifestations that they might need?  Why would an
authority module be optional?

Anyway, I think that when we look at the future of our code, we need to
put the code and needs of our patrons first, and system implementation
second.  That is, the code should take the long view.  We can then attempt
to dictate minimum requirements for our OPACs to take full advantage of
the work we do.  A good example is the second indicator value of 1 for
related works, now abandoned though, as Stephen Hearn once posted to
AUTOCAT, a useful tool that was there for implementation and patron
benefit.

As others have mentioned, though, we do need to recognize that we are in a
automated environment, not a manual one, at least as regards producing
cataloging records.  Typing card sets is not much of an issue any longer.
Numerous added entries are not the time-sinks they once were.  The data
that we input today will only help us in the future.  I don't think I've
ever seen a bibliographic record that was too useful.


Daniel
-------------------------------------------
Daniel CannCasciato
Head of Cataloging and Interlibrary Loans
Central Washington University Library
400 East 8th Ave
Ellensburg  WA   98926-7548
509 963-2120     509 963-3684 (FAX)
dcc@cwu.edu
=========================================================================
Date:         Fri, 25 Jul 1997 15:16:49 -0400
Reply-To:     mac@slc.bc.ca
Sender:       "International Conference on the Principles and Future Dev. of
              AACR" <AACRCONF@INFOSERV.NLC-BNC.CA>
From:         "J. McRee Elrod" <mac@SLC.BC.CA>
Organization: Special Libraries Cataloguing, Inc.
Subject:      Re: When good rules are bad
Comments: To: mflinchb@UMICH.EDU
In-Reply-To:  <3.0.1.16.19970725143703.3e576aaa@m.imap.itd.umich.edu>

>Because of this, some concepts, like main entry, for example, have always
>been  problematic concepts. The glossary of AACRII says that main entry is
>"the complete cataloging record for an item, presented in the form by which
>the entity is to be uniformly identified and cited. The main entry may also
>include the tracings."

This AACR definition (and Hagler's paper) notwithstanding, every
cataloguing teacher I have ever had, every catalogue department in which
I have ever worked, and every one of my fifty customers, means by "main
entry" the person (personal or corporate) primarily responsible for a
work, or in the absence of such, the title of the item (title proper or
a constructed one).  It is the first field in the MARC record after the
zero fields.  It is in most instances the basis of the Cutter number.  I
agree that the AACR2 definition is less than useless.

Our customers will often give directions using this phrase, e.g., "Use
Smith as main entry" for something like Smith on Torts Tenth Edition by
Tom Jones.

When referring to what is described in the definition you quoted, all
cataloguers which whom I have ever worked said "main entry card" or
"unit card", as do my card customers now.

You are absolutely correct that the definitions need a complete
overhaul.

Mac

   __       __   J. McRee (Mac) Elrod (mac@slc.bc.ca)
  {__  |   /     Special Libraries Cataloguing   HTTP://www.slc.bc.ca/
  ___} |__ \__________________________________________________________
=========================================================================
Date:         Fri, 25 Jul 1997 16:29:55 -0700
Reply-To:     "International Conference on the Principles and Future Dev. of
              AACR" <AACRCONF@INFOSERV.NLC-BNC.CA>
Sender:       "International Conference on the Principles and Future Dev. of
              AACR" <AACRCONF@INFOSERV.NLC-BNC.CA>
From:         Linda Barnhart <lbarnhar@GORT.UCSD.EDU>
Subject:      Are bibliographic utilities in the driver's seat?
MIME-Version: 1.0
Content-Type: TEXT/PLAIN; charset=US-ASCII

As I read Ronald Hagler's very thoughtful paper last night, I was reminded
of an idea that was proposed to MARBI in 1993 as Discussion Paper 72,
wherein the bibliographic and authorities formats would be brought closer
to alignment regarding name/title fields. (It's too bad that this is too
old a paper to be on the USMARC website.)  In brief, the paper suggested
that in bibliographic records, the 240 field should be abolished and the
uniform title data entered instead in a "t" subfield which would parallel
the authority record (and the 700 field).  While many saw the logic and
consistency of such a structure, MARBI decided that the idea should not
move forward because of the "longstanding and extensive implementations of
the present configurations."

I was personally very disappointed in this decision, not only
because I thought it was the right thing to do, but also because it is
seemingly do-able and would be an incremental step toward more
fundamental change.  My concern for the group here, though, is:  if this
kind of change in coding practice is not possible because of the weight
of millions of existing MARC bibliographic records, what hope have we for the
kinds of changes we are charting here?

I do hope that, as a professional community, we can--once we have
consensus about where we are going--develop a incremental but well-paced
implementation scheme so that we can move from where we are now to that
glorious future.  (And we thought format integration was hard!) My concern
is with the bibliographic utilities, who may be key players in this
effort.  Are they positioned for change?  If they cannot do something,
what will happen?  Who is in the driver's seat?

Linda Barnhart
Head, Catalog Dept.
UC San Diego
=========================================================================
Date:         Fri, 25 Jul 1997 22:07:20 UT
Reply-To:     "International Conference on the Principles and Future Dev. of
              AACR" <AACRCONF@INFOSERV.NLC-BNC.CA>
Sender:       "International Conference on the Principles and Future Dev. of
              AACR" <AACRCONF@INFOSERV.NLC-BNC.CA>
From:         Thomas Brenndorfer <ThomasB@MSN.COM>
Subject:      Re: Series

It has already been pointed out that MARC records and OPACs are more
accepting of redundancy in a record than were catalogue cards.  Printed
catalogues are much *less* accepting of redunancy in access points.

Mac

   __       __   J. McRee (Mac) Elrod (mac@slc.bc.ca)
  {__  |   /     Special Libraries Cataloguing   HTTP://www.slc.bc.ca/
  ___} |__ \__________________________________________________________

[]

This issue of series and redundancy as it relates to cards vs. the OPAC is
something that should be carefully examined. The system I work with, Dynix,
cannot be described as being "tolerant" of redundancy -- it crumbles and
shatters whenever so-called redundancy saving features are incorporated from
AACR2 into the MARC coding.

For example, the system is incapable of dealing with 246 as both a note and an
access point. It is incapable of properly separating 490, the 8XXs, and the
redundancy-saving 440 -- all are present and labelled as "SERIES," even nearly
identical 490 1_'s and 830's. In other words, the system is not tolerant of
redundancy; it is largely ignorant of it.

The system has a great Related Works feature, but it only works with authority
controlled fields. I have looked at various ways of dealing with the titles,
but ultimately there is going to be a breakdown because 245 can never be
authority controlled, even if the exact same title proper appears as a 730 or
a 630. As a result, any attempt of building some effective gathering or
collocating function for titles in the Related Works screen is going to fail.
It would not fail if redundancy efforts were dropped. If every record
represented each work with discrete, sefl-contained headings, the system would
work spectacularly.

Why is this so? I picture any computer-savvy teenager as having a basic
explanation. Computers understand pure logic. An element is 1 and absolutely
is not 0. When field 245 is kind of controlled, but not presented as such, it
is obvious that a computerized system will quickly hit the wall in providing
the dual functions of a catalog, namely to find and to gather.

That's why a name-title for a series is not silly. A heading, a name of the
work, a principal access point must be one thing and one thing only. To
suggest a computer program is easily tolerant of this keystroke saving
features is to miss the mark entirely. A computer program will sputter along
and fail in many places because of the lack of adherence to a strict logical
framework.

A user looking for that "Thomas Covenant" series by some author named
"Donaldson" may find the series with a keyword search, if the user spells
correctly. A lot of trial and error may result in a hit if the user searches
the title index. But the user will be more successful by browsing under
Donaldson and noticing that the series title actually begins with "Chronicles
of ...." Plus the bonus! All of Donaldson's other works are gathered under his
name. If the user found an item in the series first, he or she would not have
far to look because **everything** is found under the author's name. The
name-title form of the series heading is the principal access point for good
reason -- it is the principal way in which a great many people would naturally
think of looking for a series title written by a single author.

What I've just described is the classical card catalog scenario. Dynix,
because MARC has all these horrible redundancy features in it, must use a lot
of creative programming to build a set of functional indexes. We mask the
title part of name-titles in the author index because they appear confusing to
the public. These subfields containing the title are copied over to the title
index, but there are no cross-references to the offical name of the work. In
fact, our title index is a virtual menagerie of titles, with subfield t
scraps, analytical titles, related titles, uniform titles, title propers, and
variant titles of all of these. It is a flat, bleak terrain, with no
hierarchical structure. Most users think a 246 filing close to a 245 (because
the 246 is a slightly different cover title) represents a different work, and
they check it out as such.

I think that, if anything, the structure of series authority records is a good
harbinger of things to come. Series authority records already look remarkably
like bibliographic records. All variants of the series title are added as
cross-references, not as added entries to the bibliographic records. There is
a clear separation of transcription and access in the bibliographic record,
with the exception of the non-mnemonic 440. There is the freedom to create
cutters out of either the series heading or the heading of each item --
whichever works best in the library. Whatever the case, a move to
computer-friendly discrete and logical structures would largely eliminate many
of the deficiencies that typical on-line systems face. It is ironic that some
say our weapon against the "intelligent agent" promoted by Gates et al. is the
authority-controlled collocating function of traditional catalog design, and
yet, such control is poorly implemented in a computerized environment, in
part, to placate those who want to reduce the pain of excessive keyboard
tapping.

Tom Brenndorfer
thomasb@msn.com
Guelph Public Library
=========================================================================
Date:         Mon, 28 Jul 1997 09:38:43 +1100
Reply-To:     Hal.Cain@ormond.unimelb.edu.au
Sender:       "International Conference on the Principles and Future Dev. of
              AACR" <AACRCONF@INFOSERV.NLC-BNC.CA>
From:         Hal Cain <hecain@ORMOND.UNIMELB.EDU.AU>
Organization: Joint Theological Library, Parkville, Victoria, Australia.
Subject:      What is AACR? (was: Re: Series)
Comments: cc: thomasb@msn.com, mac@slc.bc.ca
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit

Thomas Brenndorfer wrote:
>
> This issue of series and redundancy as it relates to cards vs. the OPAC is
> something that should be carefully examined. The system I work with, Dynix,
> cannot be described as being "tolerant" of redundancy -- it crumbles and
> shatters whenever so-called redundancy saving features are incorporated from
> AACR2 into the MARC coding.
>
I also work with Dynix.  What Tom describes was true for us under an
earlier release, but under the current release we can distinguish
between 440/8XX and 490 tags, if we wish.  I agree though that the
system does not deal well with tags which combine note and heading
data.   At the base of the problem is the architecture of the system,
where there is a one-to-one correspondence between MARC data elements
(at the tag or subfield level) and internal database fields.  I see no
intrinsic reason, though, why data elements cannot be replicated
internally.  IMO this is a case where customers need to apply pressure
to the vendor.

One the principle of replicating certain data elements within the system
is accepted, redundancy becomes no problem, surely?  Any particular
system can replicate to the programmer's heart's content, if that is the
approach required to obtain the display, searching, and authority
functions required.  The practical problem for the library is that
vendors are putting effort into devising GUI OPACs and Z39.50
functionality.  Other contributors to this list have already mentioned
the need for standards in OPAC displays.  Unfortunately the Z39.50
approach seems to have brought along a wave of lowest-common-denominator
implementation, where series is ignored as a search approach.  At that
level, the form of series heading is irrelevant.

I think this discussion will advance only if we stop and think about
what we are trying to do, in considering the revision of AACR.

IMO AACR is fundamentally a standard for describing and providing access
keys to bibliographic entities.

The specifications for organizing and displaying this data (including
standard punctuation and layout) reproduce those of ISBD and are *not*
fundamental to the code (in AACR1 they were, but the advent of ISBD
changed that).  (BTW if ISBD were respected by those who design OPACs
the problem of inconsistencies between systems would be much reduced).
ISBD's descriptive requirements are expressed, for the bibliographic
community which has adopted it, in AACR2.  It would be possible, for
that community, so turn our backs on ISBD and set about re-specifying
the descriptive requirements -- but that would put us out of step with
other parts of the international bibliographic community which have
adopted ISBD as the basis for their own codes.

AACR2 provided for three levels of completeness in description.  In
practice a new supplementary standard, that of the core record, has
emerged.

The most common specification for coding and transmitting the data
compiled under AACR is our old friend, USMARC.  USMARC makes demands not
based on AACR, sometimes in conflict (e.g. omission of initial article
in 246 $a) -- it also fails to distinguish data which ISBD/AACR2 does
distinguish (e.g. parallel title data) and specifies coded
representation of data which AACR requires in text form (so in practice
both are included).  I am unfamiliar with the specifications of UKMARC
and CANMARC (Ausmarc is obsolescent if not obsolete) but I assume they
exhibit similar differences.  USMARC was originally framed for AACR1 and
the requirements of one library (the largest) and conveys more than
AACR-compliant data.  It also is poorly-equipped to designate data to be
omitted from a less-complete display; so where brief or medium-level
detail is to be provided (a matter of specification in terms of the
descriptive code) system designers have taken over.

Of course, we can't discuss AACR in isolation from the other codes.  But
this process of preparing for revision of AACR will be more fruitful
(whatever the outcome) if we stop and think how AACR interacts with ISBD
and USMARC (and particular systems, too) and try to clarify its role.

It's too soon to talk about "AACR3".  But since "Revision of AACR" is a
mouthful (though it includes no redundancy) what about "AACR+" as a
convenient label?

Hal Cain, Joint Theological Library, Parkville, Victoria, Australia
<hecain@ormond.unimelb.edu.au>
=========================================================================
Date:         Sun, 27 Jul 1997 18:48:26 -0400
Reply-To:     mac@slc.bc.ca
Sender:       "International Conference on the Principles and Future Dev. of
              AACR" <AACRCONF@INFOSERV.NLC-BNC.CA>
From:         "J. McRee Elrod" <mac@SLC.BC.CA>
Organization: Special Libraries Cataloguing, Inc.
Subject:      AACR/MARC conflict
Comments: To: hecain@ormond.unimelb.edu.au
In-Reply-To:  <33DBCDF3.63FA@ormond.unimelb.edu.au>

Hal Cain <hecain@ormond.unimelb.edu.au> wrote:

>The specifications for organizing and displaying this data (including
>standard punctuation and layout) reproduce those of ISBD and are *not*
>fundamental to the code (in AACR1 they were, but the advent of ISBD
>changed that).

A respect for ISBD as a display standard would improve our bewildering
variety of OPAC displays tremendously.  It would obviate the need for
labels (which are often wrong; a Festschrift honoree is *not* an author,
nor is a defendant in a legal case, nor a pop music performer, nor a
composer for that matter.)  If code revision could add specific
references to ISBD as a display standard, it would be of great benefit.

In my opinion, one of the greatest conflicts between ISBD and MARC is
that area between edition and imprint for material specific information.
This is *one* area in ISBD.  In MARC is has been needlessly splintered
into a variety of fields.  We are going to soon run out of numbers
between 250 and 260.

Mac

   __       __   J. McRee (Mac) Elrod (mac@slc.bc.ca)
  {__  |   /     Special Libraries Cataloguing   HTTP://www.slc.bc.ca/
  ___} |__ \__________________________________________________________
=========================================================================
Date:         Tue, 29 Jul 1997 17:30:58 -0400
Reply-To:     mac@slc.bc.ca
Sender:       "International Conference on the Principles and Future Dev. of
              AACR" <AACRCONF@INFOSERV.NLC-BNC.CA>
From:         "J. McRee Elrod" <mac@SLC.BC.CA>
Organization: Special Libraries Cataloguing, Inc.
Subject:      Re: AACR/MARC conflict
Comments: To: cmboett@ucla.edu
In-Reply-To:  <2.2.32.19970729194858.006bdf3c@pop.ben2.ucla.edu>

Cheryl Boettcher Tarsala <cmboett@ucla.edu> wrote:

>I'm sure that you know that access points aren't part of ISBD, and that the
>important distinctions between writer, performer, defendant, editor, etc.
>aren't even well-addressed by AACR because they're all, in the end, just
>access points.

If the display retains the 245$c information, it is my contention, one
does not need a label for the 1XX or 7XX thus justified.  It *says*
"edited by", "in honour of", or whatever.  The "/"  replaces an
unexpressed "by".

Getting rid of redundancy by suppressing 245$c in OPAC displays saves
less room than is taken up by the labels thus made necessary.  For close
to 100 years we got by without labels on catalogue cards, and I don't
think we need them now with properly formatted displays.

Mac

   __       __   J. McRee (Mac) Elrod (mac@slc.bc.ca)
  {__  |   /     Special Libraries Cataloguing   HTTP://www.slc.bc.ca/
  ___} |__ \__________________________________________________________
=========================================================================
Date:         Wed, 30 Jul 1997 08:14:45 -0400
Reply-To:     mac@slc.bc.ca
Sender:       "International Conference on the Principles and Future Dev. of
              AACR" <AACRCONF@INFOSERV.NLC-BNC.CA>
From:         "J. McRee Elrod" <mac@SLC.BC.CA>
Organization: Special Libraries Cataloguing, Inc.
Subject:      Records lacking collation

A recent discussion on autocat calls to mind rules which call for the
omission of collation.

With the introduction of the smd to collation, the collation becomes
more than just a physical description.  An smd indicating the nature of
the item being catalogued seems appropriate to me, e.g, "www electronic
resource".
=========================================================================
Date:         Wed, 30 Jul 1997 08:17:12 -0400
Reply-To:     mac@slc.bc.ca
Sender:       "International Conference on the Principles and Future Dev. of
              AACR" <AACRCONF@INFOSERV.NLC-BNC.CA>
From:         "J. McRee Elrod" <mac@SLC.BC.CA>
Organization: Special Libraries Cataloguing, Inc.
Subject:      Records lacking imprint

I would like to see imprint expanded to include provenance, e.g, the
university at which a thesis was produced.  With electronic texts and on
demand print copies, theses are no longer an original typescript with a
few carbons.  One should not have to look at a note to see from whence
it came.

Mac

   __       __   J. McRee (Mac) Elrod (mac@slc.bc.ca)
  {__  |   /     Special Libraries Cataloguing   HTTP://www.slc.bc.ca/
  ___} |__ \__________________________________________________________
=========================================================================
Date:         Thu, 31 Jul 1997 13:47:00 -0400
Reply-To:     "International Conference on the Principles and Future Dev. of
              AACR" <AACRCONF@INFOSERV.NLC-BNC.CA>
Sender:       "International Conference on the Principles and Future Dev. of
              AACR" <AACRCONF@INFOSERV.NLC-BNC.CA>
From:         Stewart Marg <MAS@ABS.NLC-BNC.CA>
Subject:      Bibliographic Relationships
Content-transfer-encoding: 7BIT

Please note that the following conference paper is now available on the JSC
Web site:

Bibliographic Relationships by Sherry L. Vellucci