Date: Tue, 16 Jan 1996 13:56:32 -0500 (EST)
From: Patrick Bernard <pber@loc.gov>
Subject: Report, Survey on CF Form Subdivisions

              FORM SUBDIVISIONS FOR COMPUTER FILES
                     A Report on the Survey

        This report is being distributed widely within LC and to several
Listservs (Emedia, Intercat, Conserlst, and Serialst).  I am also sending
the report to the ALA ALCTS SAC Subcommittee on Subject Analysis to
Computer Files with a request for recommendations for changes, if any, in
current LC practice based on the results of this survey.

     The survey on "Form Subdivisions for Computer Files" was
sent to four LISTSERVs (Emedia, Intercat, Conserlst, and
Serialst) around November 20, 1995.  By December 22, 1995, I had
a total of 35 replies.  This report compiles and tabulates those
replies.

     The survey was done rather informally and I claim no special
validity for my sample.  I sent the survey to the four LISTSERVs
through which I felt I could reach those librarians now creating
and using these records.  Practically all the replies were from
catalogers, and I did get more replies from Intercat subscribers
than from any of the other LISTSERVs.  My one and a half
reference librarian respondents is explained by the fact that one
respondent divided his/her time between reference and cataloging.
If nothing more, I believe these tallies and comments included in
the present report provide continuing fodder for the ongoing
discussion of these important issues.

     Below, I repeat each question as it appeared in the survey.
I follow it by the number of yes and no votes and the percentage
that each represents.  The percentage is always the percentage of
35, not the percentage of the yes and no votes to that particular
question.  Then follows a selection of the comments offered on
that particular question.  I included those comments that I felt
would further the discussion of these issues.  I edited the
comments freely for style, clarity, and ease of reading.  The
substance of the comment, however, is always that of the
respondent.  Where comments repeated each other, I generally
included only one of these.  Preceding each comment is a "Y" or
"N" which indicates whether the comment comes from a respondent
who marked Yes or No on that question.  The yes and no comments
are each grouped together, the yes comments appearing first.

-----------------------------------------------------------------

(1) To a monographic bibliography, catalog, directory, index,
etc. in electronic form, LC catalogers now assign two form
subdivisions, one for the bibliography, catalog, directory,
index, etc., and a second subdivision "--Databases" to indicate
that the bibliography, etc., is in electronic form.  The practice
is well established and seems to be accepted.  Is this practice
one that we want to continue?
          Yes  8 (23%)        No  26 (74%)
Comments:
Y  An alternative phrase form subdivisions (e.g., "--
Bibliography, Electronic)  would also be OK.
Y  Continue until the 655 option is implemented.
N  The subdivision "--Databases" should be restricted to records
for true databases.
N  The routine use of this subdivision for materials which are
not databases makes it useless for the purposes for which it was
designed.
N  It is time to expand to other terms.
N  This seems to confuse the carrier with the intellectual form.
In addition, this practice diverges from the way in which LC
treats most other non-print materials, such as microforms.
N  We would really prefer to see OPACs develop so that they can
use information which is already in the record to provide
searching and limiting functions.
N  Why not use a more explicit phrase such as "--Electronic file"
or "--Electronic format"?

(1a) Do your library's patrons understand these subject heading
strings?
          Yes  4 (11%)        No  19 (54%)
Comments:
N  Neither reference librarians nor the library's patrons are apt
to be able to distinguish fine distinction such as format vs.
content, nor do they always understand the differences in meaning
which occur as the result of changes in the order of the elements
(e.g. "--Databases--Periodicals" vs. "--Periodicals--Databases.")
N  We doubt they do, as some catalogers do not.
N  We hope they understand "--Catalogs," "--Directories," "--
Indexes," etc.  We are less sure about "--Databases," especially
as it is used both as a topical and as a form subdivision.
N  We doubt it.  Double interpretations of a subject heading
string make it even less likely that patrons will understand it.
N  I cannot really say whether the patrons understand them, but
the reference librarians clearly do not.

(1b) Are these subject heading strings confusing to the patron
seeking actual databases or work about databases?
          Yes  23 (66%)       No  5 (14%)
Comments:
Y  How can they not be, since they provide no way to distinguish?
Y  This is part of the general confusion between topic and form
that is still an integral part of LCSH.
Y  Well, it's confusing for me, so I assume it's also confusing
for a patron.
Y  If you limit "--Databases" to works about databases and use
subdivisions such as "--Computer file" or "--Electronic format"
when that is what is meant, then subject searching will be much
more accurate.

(1c) Once we implement subfield "v" for form subdivisions and can
clearly code these to distinguish between form and topic, does
that solve the problem?
          Yes  14 (40%)       No  18 (51%)
Comments:
Y  The displays will probably still be confusing to patrons
because they won't see the coding.  But in terms of designing
systems which will understand form vs. topical, subfield v will
be a great addition.
Y  All form subdivisions should be limited (divorced!) from the
subject headings and replaced with a term in a separate field.
When a form subdivision is used to mean the aboutness of a work,
it should be kept as a subfield x.
Y  It should help, although I think that form should never be
part of the subject heading strings.  Form belongs in a
separately coded field.
Y  But only IF many other things are also done: update and
standardize older records; educate our users regarding the new
subfield, its use, what it means; have subfield v available as a
separate index in keyword searching; have it available as a
searching qualifier in our online public access products; and,
make sure it displays in our OPACs.
N  Not if you use the same term for both, and your OPAC software
cannot differentiate the two.
N  We doubt patrons would distinguish on the basis of subfield
indicators, even were the indicators visible to then, or the
local system manipulated subfield x differently from v.

(2) Since other areas of the record indicate the electronic form,
is it important for monographic computer files to indicate in the
subject string that the item is in electronic form?
          Yes  12 (34%)       No  19 (54%)
Comments:
N  The format in which a resource is issued doesn't have a lot to
do with its subject matter.  (in a sense, we see this whole
discussion as parallel to the 008/006 implementation in terms of
choosing a primary attribute on which to focus.
N  We want to mainstream these items and we certainly don't add
form subdivisions for videorecordings.
N  The subject heading string is overworked as it is.

(3) Is it important for computer file serials to indicate in the
subject heading string that the serial is in electronic form?
          Yes  13 (37%)       No  20 (57%)
Comments:
Y  At this point in time, I think that it is.  We aren't yet to
the point where print and electronic journals are equivalent
either in accessibility or utility to a user.  Someday the answer
to this one may be different.
Y  How about a new subdivision: "--Periodicals, Electronic"?
Y  One idea for a possible subdivision would be "--Periodicals
(Electronic)."

(4) The present practice is that the form subdivision "--
Periodicals" is used for electronic serials without any
indication that the serial is in electronic format.  When other
form subdivisions are assigned to monographic computer files,
they are further subdivided to indicate that the item is in
electronic format.  Is this divergent practice for serials and
monographs acceptable, provided the practice is clearly defined
in the appropriate cataloging manuals and instructions?
          Yes  5 (14%)        No  28 (80%)
Comments:
Y  Uniform practice is desirable, but we can accept the different
treatment.
Y  I doubt that anyone other than catalogers is seriously
disturbed by the discrepancy.  It's not like there are no other
such contradictions in our catalogs.
N  We prefer the monographic practice with regard to
subdivisions.
N  There should be consistency in practice between monographs and
serials whichever practice is followed.  That would be a concern
for us if electronic serials are treated differently from
monographs.
N  The practice now in place for serials should be adopted for
monographs as well.
N  Practice should be consistent and standard.
N  Monographic and serial records all live together in our OPAC.
Divergent practices might be well-documented for catalogers, but
patrons do not see that type of documentation, and are ill-served
by inconsistent practices.
N  Should the form subdivision perhaps be "--Electronic
periodicals"?
N  Monographic cataloging practice should be changed to coincide
with serials cataloging practice.
N  Why have two sets of rules?  Having consistent application
reduces the chances of  misunderstanding and error.  Consistency
makes a great deal of difference to the user.
N  I catalog both monographs and serials and it's very confusing
to me|

(4a) Should we continue present practice until provisions now
under development for 6XX "v" subfield and 655 field for
form/genre headings are fully implemented?
          Yes  11 (31%)       No  20 (57%)
Comments:
Y  Yes, don't start something new, and temporary, which will
result in a multifaceted cleanup project later.  Don't cease the
monograph practice - it is useful, if not perfect, and people
have come to rely on it for software.
N  Subfield v is not a panacea.
N  If changes are to be made, they should be made as soon as
possible..
N  Better sooner.  There needs to be action as to what these
terms will be and how they will be coded--the issue is here
today.

(5) Should we implement the 655 field for form/genre/physical
characteristics and start applying it in records for electronic
serials?
     Yes  31 (89%)       No  3 (9%)
Comments:
Y  I do have concerns about users having to distinguish between a
650 and a 655 when the same heading is used as a topical heading
and as a form heading.
Y  We really prefer getting systems to use the information that
is already in the record.  However, we know that patrons want
access to materials in special formats, so this would be a way
for them to get the access given present system configuration.
Y  We should implement, but for the 655 headings, we also need
authority records with appropriate cross-reference structure.
N  We should be working on a solution to the multiple versions
problem rather than making separate records more and more
complex.

(6) Should we implement the 655 field for form/genre/physical
characteristics and start applying it in records for monographic
computer files?
          Yes  31 (89%)       No  3 (9%)
Comments:
(There were no comments on this question, although several
indicated that their comments on question no. 5 also applied
here.)

(7) Given implementation of the 655 field for form/genre/physical
characteristics and its use in records for electronic serials,
should we also indicate the electronic form in the 600/651
subject strings?
          Yes  12 (34%)       No  22 (63%)
Comments:
Y  The issue here seems to be the old pre-/post-coordination
argument.  Given that LCSH is intended to be a pre-coordinated
system, I would prefer that the form information be included in
the 600/651 subject headings, and that we not be forced to rely
on a Boolean search involving the 600/651 combined with the 655
fields.
Y  I'd give much weight to Thomas Mann's views on this.  I think
he favors keeping the form information in the 600/651 headings.
N  While these publications are "unusual" now, they will be very
common in the future and the exceptional practices we adopt now
will not be desirable.
N  This seems a bit redundant.  Boolean searching should remove
the need to code the same information in two different-but-
related fields.  Some OPACs either may not accommodate the 655 or
may require re-indexing the online catalog--but allowing current
technology to drive the practice perpetuates past errors and
seems to be the tail wagging the dog.

(8) Given implementation of the 655 field for form/genre/physical
characteristics and its use in records for monographic computer
files, should we also indicate the electronic form in the 600/651
subject strings?
          Yes  13 (37%)       No  22 (63%)
Comments:
Y  Until systems generally provide easy searching of data in
multiple fields in a combined search, we'll continue to need form
subdivisions in the subject heading string.
Y  The terminology itself should indicate distinction between
individual databases and works about databases.

-----------------------------------------------------------------
One respondent offered these general comments:

We find this is a very complex area to discuss.  There is some
feeling that indication of physical form such as CD-ROM or large
type books belong primarily in the description.  Some also
expressed strong support for this type of information in the
subject headings.  Subfield v should be used to indicate the
"internal" or "organizational" form of the document
(bibliography, index, dictionary, register, etc.).  We felt that
some quite broad subdivisions of this nature might be allowed for
electronic publications and computer files, such as the
subdivision "--Interactive multimedia."  Such subdivision should
use appropriate terminology and be clear as to their meaning.
For instance, we have qualms about the appropriateness of "--
Databases"  for the types of material to which it is currently
being assigned.

-----------------------------------------------------------------
In the survey, I also asked each respondent to supply certain
information about themselves.  The following are the tallies to
those questions.  For each item, I give the number of respondents
that checked off that item and the percentage (of 35) that the
number represents.

(9) My job is mostly as (Check one):
     Cataloger  26.5 (76%)         Reference Librarian  1.5 (4%)
     Library administrator  1 (3%)      Other  2 (6%)

(9a) If you checked "Cataloger" above, is your work mostly with
(Check one):
     Serials  7 (20%)         Monographs  3 (9%)       Both  16
(46%)

(10) My library has this many FTE professional library cataloger
positions (Check one):
     1 FTE  0       2-5 FTE  13 (37%)        6-10 FTE  10 (29%)
     11-25 FTE  5 (14%)       Over 25 FTE  4 (11%)

(11) I subscribe to the following ListServs (Check as many as
apply):
     Emedia  17 (49%)         Intercat  22 (63%)
     Conserlst 12 (34%)
     Serialst  18 (51%)       None of these  3 (9%)

-----------------------------------------------------------------
     What policy changes might come of this survey are for the
moment undetermined.  These issues are currently under study by
both the Library of Congress Cataloging Policy and Support
Office, and the ALA ALCTS Subject Analysis Committee,
Subcommittee on Subject Access to Computer Files.  We at LC are
committed to working closely with the ALCTS SAC Subcommittee.
Any changes to LC policy will be announced through the usual
channels of the Cataloging Service Bulletin, and the Subject Cataloging Manual.

     I would like to thank all of you who took the time to respond to the
 survey.  My warmest wishes to each of you for a prosperous and happy New Year.

Patrick Bernard, Cataloging Policy Specialist for Computer Files
Cataloging Policy and Support Office
Library of Congress    Internet: pber@loc.gov