ALCTS CCS Subject Analysis Committee
ALA Midwinter Conference 2011, San Diego, California
Sunday, January 9th, 2011 8:00 a.m.–11:00 a.m./GASLAMP-Santa Rosa
Members present January 9: Daniel N. Joudrey, Linda Ballinger, Jeffrey Beall, Christopher
J. Cronin, Jimmie H. Lundgren, Scott A. Opasik, Tony Olson, Deborah A. Ryszka, Adam
Liaisons present January 9: Julianne Beall, Sherman Clarke, Stephen S. Hearn, Yael
Mandelstam, Joseph Miller, Joan Mitchell, Ed O‘Neill, Deborah Rose-Lefman, Hermine
Members absent January 9: Mary Catherine Little, Molly D. Poremski
Liaisons absent January 9: David Miller
1.1 Welcome and introduction of members and guests
1.2 Adoption of agenda
Joudrey announced changes to the order of the agenda—move agenda item 2.12 to
after agenda item 1.3, followed by agenda item 1.8 in order to accommodate some
scheduling conflicts. Changes were unanimously accepted. Adoption of the revised
agenda was moved, and approved unanimously.
1.3 Adoption of 2010 Annual minutes
Schiff suggested correcting ―the Gettysburg‖ to ―the Getty Institute‖ under agenda
item 1.10—Report of the Liaison from the Art Libraries Society. Adoption of the
revised minutes was moved, seconded, and approved unanimously.
2.12 New Business
2.12.1 RDA and Group 3 entities
John Attig, ALA Representative to the Joint Steering Committee for Development
of RDA (JSC), and Lori Robare, Chair of the Committee on Cataloging:
Description and Access (CC:DA) met with SAC to discuss an upcoming change in
the development of RDA. Attig explained that as the ALA Representative he is
responsible for submitting proposals from the American Library Association
(ALA) for revisions to RDA (Resource Description & Access), submitting
responses from ALA to proposals from elsewhere, and representing ALA at JSC
meetings. He gets advice from CC:DA on RDA content and passes it on to the JSC.
Page | 1
He explained that unlike in the past, RDA will now contain some content on
subject entities, attributes, and relationships. The RDA chapters addressing the
Group 3 entities from the various Functional Requirements for
Bibliographic Records (FRBR) models may only explain what subject
attributes and relationships are, but most likely will not address how to apply them
in a particular standard. As the JSC moves into the realm of subject analysis, the
current advising structure with ALA needs to change. CC:DA is not charged with
addressing subject access issues--that falls under the purview of the SAC. The ALA
Representative needs support on subject-related issues. Attig, Robare, and Joudrey
need to recommend to the Cataloging and Classification Section (CCS) Executive
Committee a process for determining and reporting ALA‘s positions on subject-
related content in RDA.
Joudrey suggested a possible approach: the CCS Executive Committee will appoint
a SAC liaison to CC:DA (a new position), who will chair a new standing SAC
subcommittee on RDA. The liaison would report to the SAC to vote on the
recommendations; the liaison would work with Attig and report progress to
CC:DA. Lori Robare elaborated: this Subcommittee would include some people
who know CC:DA procedures. The liaison would be a voting member of SAC.
SAC would make the decisions, which will be reported at CC:DA, but not voted on
by them. Attig added that he would work with the subcommittee and be at some
SAC meetings so he could give and get direct feedback.
Attig reported that this is somewhat urgent as the JSC may produce a discussion
paper by the time of ALA Annual 2011. Joudrey and Robare are setting up a joint
CC:DA/SAC meeting to discuss the issues and get to know one another. Schiff
asked if they will meet at ALA and how this is different from just coming to
general SAC meetings. Attig thinks that CC:DA and SAC see RDA differently,
especially since RDA is a major priority for CC:DA. The SAC can add an agenda
item to discuss RDA, but it does not have to be the only thing discussed at SAC
meetings. A permanent structure will be helpful in the future as proposals continue
to be made, rather than Schiff‘s proposal of a task force as needed, which could be
useful for specific tasks.
Robare says that if decisions come through CC:DA, the organization can continue
to use the documentation process that has already been established during CC:DA‘s
evaluation of RDA, and SAC won‘t have to develop a process of its own. Stephen
Hearn wanted to know if this is now part of the job description for the ALA liaison
and if so, would anyone else ever take this job? Attig will discuss this point with
the CCS Executive Committee to figure out if the description is practical since it is
not clear if anyone else would need to do this job after Attig. Joudrey took an
informal vote, which was unanimously approved, with a caveat that if it doesn‘t
work, we can change it in the future.
Joudrey, Attig, and Robare will present the idea to the CCS Executive Committee.
Robare said there was a discussion of having a joint meeting at the 2011 Annual
Page | 2
ALA Conference on Friday, but this may conflict with the preconference. Cronin
proposes shortening the first SAC meeting and having a joint meeting during the
Monday session of SAC (in the afternoon) or CC:DA (in the morning). The
consensus was that this could work, although it is unknown how long it would be.
A proposal is needed to change the meetings.
1.8 Dewey Decimal Classification Reports
1.8.1 Report on Dewey Classification and OCLC Dewey Services
See written report on ALA Connect (link provided). [SAC11-MID/5]
1.8.2 Report of the liaison from Dewey Section of the Library of Congress
See written report on ALA Connect (link provided). [SAC11-MID/5]
Miller asked if there‘s a way for users of the Abridged WebDewey to be
told about changes that will impact them. Mitchell said that would probably
1.8.3 Report of the liaison from the Dewey Classification Editorial Policy
Committee (Deborah Rose-Lefman)
See written report on ALA Connect (link provided). [SAC11-MID/6]
Mandelstam asked about the classification of e-book readers as
publications, and commented that there is a difference between the content
and the device. Rose-Lefman said there was a debate about that, but there
isn‘t a place for e-book reader as a device. Mandelstam pointed out that
many devices, including iPads and iPods and mp3 players, can read e-
books, and their purpose is constantly changing. Mitchell said there is a
general statement that a device should be placed according to its primary
use. She compared this to rock music, which is also hard to define and
classify. Hearn asked if the electronic meeting is a viable alternative. Rose-
Lefman said typically an electronic meeting is mostly the follow-up from a
face-to-face meeting, which works just fine.
1.4 Old Business
There is no old business.
1.5 Report on the Sears List of Subject Headings (Joseph Miller)
See written report on ALA Connect (link provided). [SAC11-MID/3]
Page | 3
WilsonWeb is hoping to eventually have PDF versions of every previous edition,
back to 1923, available. The list in Spanish will be available in about a month. It is
in beta right now and on display at the Wilson booth. All the headings that were
created since the print Spanish edition was published have been translated and will
be available once the system goes live.
Janis Young asked how the bilingual link works. Miller said records will have a
7XX linking field corresponding the record numbers to other record numbers in the
Thesaurus. Young followed up with a question about the variant languages issue
that the terms don‘t exactly match. Miller said if the two languages don‘t match,
they won‘t link. Clarke asked about changes in the headings ―Indians of North
America‖ and ―Native American,‖ if anyone had done a mashup of Sears List of
Subject Headings and LCSH (Library of Congress Subject Headings). As far as
Miller knew, the answer is ―no.‖
1.6 Report of the liaison from the Policy and Standards Division of LC
See written report on ALA Connect (link provided). [SAC11-MID/4]
Janis Young added that the 32nd edition of LCSH does not contain validation
Schiff asked about footnotes in the 072 Pilot Project. Young doesn‘t know. She
gave a brief preview in her ACIG presentation.
Young mentioned that some of the approved cartographic headings are being
reconsidered after catalogers actually tried to use them. Vermeij asked for examples
of types of cartography headings that were problems. For example, plat map and
cadastral map have similar definitions. There was a proposal to cancel the headings
for Cadastral globes and Lunar globes and make clear that ―Globes‖ isn‘t used for
the Earth, but rather for post coordination with the heading for the planet.
Joudrey asked about the status of LCGFT (Library of Congress Genre/Form Terms)
for literature. Young said the literature project is scheduled to begin this year. This
project will require the rethinking of how we catalog literature at Library of
Congress. Schiff asked if the Library of Congress will work with other groups for
this project. Young said ―no,‖ because there isn‘t any organization that she knew of
devoted to literature librarianship, unlike religion, law, and music. Schiff suggested
the Modern Language Association (MLA) or the ACRL Western European Studies
Section (WESS). Mandelstam suggested that it doesn‘t have to be a group that
comes from specific associations, but it could be people from outside groups that
catalog literature collections. Young said it is something that she would be
interested in exploring. If there‘s interest, she would go that way. Schiff asked if
the SAC wants to appoint a subcommittee. Joudrey asked if there was interest from
Page | 4
SAC members but there was not much response. Hearn asked if this project is
underway. Young said the project has not yet begun, but it is on the timeline for
this year. She wants to finish with cartography, law and religion first, but is aiming
to start this year. Hearn suggested using the experience from working with the
American Association of Law Libraries (its Genre/Form Project) or other groups
(Genre/Form Projects) to help set up the process before having SAC involved.
Joudrey said this would be discussed again at the Annual Meeting.
Miller asked about subdivisions (e.g., Thirty Years' War, 1618-1648—Aerial
operations)--if there is an attempt to force validation of combinations of subject and
subdivision. Young said that it is a way to validate the headings that the catalogers
have assigned without having authority records for each combination, and also a
way to suggest an appropriate subdivision to the cataloger. For example, if you type
―Thirty Years‘ War, 1618-1648,‖ then click a button, you could get a pop-up list of
applicable subdivisions that would help to speed up the cataloging process and
improve accuracy. However it won‘t work for geographic headings.
Miller pointed out that about 10 years ago, Wilson started using a new editorial
system and this system forced validation on all the subjects; as a result, they ended
up with much cleaner data. They took all of the subjects that were on bibliographic
records, established authority records for each of them, and from that point, every
subject and subdivision combination had to be established before it could go
through the system. It seemed like a lot of work but it was actually less work than
figuring out how to make an automated validation system work for all the
possibilities of all the different kinds of subdivisions. Young agreed. O‘ Neill
added that FAST had a similar experience.
Young said LC is keeping validation records and continuing to create them if
heading strings are clean. They also get comments about validation records that
don‘t make sense, e.g., Dogs—Politics and government. Young admitted that the
LC validation records aren‘t perfect.
072 fields can be algorithmically applied to some headings, but other headings need
to be examined individually. Miller commented that free-floating subdivisions are
problematic for the system. Young agreed and added that the subdivision order is
difficult for a computer to understand. For example, History—Historiography and
Historiography—History mean different things. How can we make the computer
formulate these correctly? Schiff would like to see validation records for regional
headings, if LC would consider creating a common heading such as ―Metropolitan
area.‖ Young will take Schiff‘s request back to LC.
Olson wanted to know where the data is recorded when the links between LCSH
and additional languages are created. Young doesn‘t know the technical details.
She will find out and will e-mail the SAC list. [Response from Young to SAC list:
The SKOS interface (authorities.loc.gov) has the ability to link to other thesauri,
including those in other languages. Currently, the only thesaurus linked to LCSH
Page | 5
in this way is RAMEAU, and the link was accomplished by hard-coding the URI of
the LCSH heading into the RAMEAU record. This is still experimental, and has
the drawback of being only a one-way linkage (i.e., you can link from LCSH to
RAMEAU, but not from RAMEAU to LCSH, since the URIs for RAMEAU
headings are not included in the LCSH authorities). LC continues to investigate
other methods for linking data.]
Clarke asked if the ATLA (American Theological Library Association) project
would include liturgical objects, and if so, could Young contact ARLIS/NA,
because the ARLIS/NA may not want ATLA establishing genre/form terms for
liturgical art. Young agreed to keep Clarke informed.
Hearn asked if LC is indexing and displaying the BISAC (Book Industry Standards
and Communication) and American Mathematical Society (AMS) subject headings.
Young believed they are probably keyword indexed; Schiff thought they‘re in a
browse list. Schiff commented there are no BISAC authority records loaded into
the LC database. Young confirmed this and said that the BISAC headings are in
the database as a snapshot, much like NLM's headings.
Elaine Winske requested that, when deleting a form heading in favor of a genre
heading, LC give notice in the authority record of what it will be replaced with so
we don‘t delete the old headings and lose data. Young will check on it. By policy,
LC is trying to keep the two thesauri (LCSH and LCGFT) separate.
1.9. Report of the liaison from the Music Library Association (Hermine Vermeij)
See written report on ALA Connect (link provided). [SAC11-MID/7]
The timeline for the music genre/form project is unclear. A number of things still
need to be decided: for instance, will topical headings with form subdivisions be
changed to genre headings? For example, how do we denote music intended for
children as an audience versus music intended for children as a performer? The
Music Genre/Form Task Force will be meeting with the LC group that is working
on the music genre/form project in February at the Music Library Association
Annual Meeting in Philadelphia.
O‘Neill asked about discussions of topical headings that connote the medium of
performance, e.g., Guitar music. Vermeij said they talked about it, decided to
separate medium of performance, e.g., instruments like guitar or voice, from the
genre/form, e.g., symphonies or sonatas. It is possible for a piece of music that has
no genre to have no genre headings at all, but to have a broad subject like Music
qualified with a medium of performance heading. Schiff asked if that would be a
new field. Vermeij doesn‘t know yet. There is a 3xx field but there is difficulty
with the definition of the field in RDA. No decisions have been made yet, but there
Page | 6
could be a MARBI proposal eventually. All LCSH that are clearly genres have
O‘Neill asked if those terms will be dropped as 150s. Vermeij answered no,
because they are still valid for topical heading use. There are some dual terms, like
Opera/Operas, which will likely be cancelled. The subdivision term ―History and
criticism‖ may no longer be used.
Hearn asked about proposals to put headings into authority records; Vermeij said
currently the information is in bibliographic records, but they‘d eventually like to
have it in authority, work, or expression records. Vermeij also said that they‘re
looking at linking fields, i.e., medium of performance with the corresponding genre
and also anthologies with multiple media of performance and genre.
John Mitchell asked if they have come up with a definition of ―Intended audience,‖
e.g., ―Music for children or juveniles‖—is it performed by [children] or listened to
by [children]? Vermeij said there has been a great deal of discussion about
audience and performer and which terms refer to both. Currently ―Children‘s
music‖ refers to both. It means music performed by or for children. However, the
work ―Peter and the Wolf‖ is performed for children, but not performed by
children. The ―Music for children‘s audience‖ term specifically denotes the
audience and not necessarily the performer, whereas the term ―Children‘s music‖ is
a broader term meaning music for children. Mitchell asked if coding in the record
for intended audience, rather than in headings, is being considered. Vermeij said
this is something about which the group needs to have further discussion.
Mandelstam mentioned associating headings with other information elsewhere in
the record. Mitchell said perhaps some of the issues can be avoided by taking this
out of the heading area. Vermeij and Mandelstam agreed that‘s probably where
they will end up. Hearn again addressed the issue that the audience code applies to
the whole record, but the subject string applies to elements in the record.
Mandelstam refers to subfield 8 as a linking between specific genre term and
specific piece of information in the record. Schiff thought the fixed field may not
be adequate. Vermeij thought the audience fixed field doesn‘t even currently exist
for the score. O‘ Neill pointed out there is a fixed field that only addresses age of
audience. Vermeij mentioned that currently ―Music by child composers‖ is only
valid for collections in LCSH. In addition there are still discussions about genre
headings in which characteristics of the creator are addressed.
1.10 Report of the liaison from the American Association of the Law Libraries (Yael
See written report on ALA Connect (link provided). [SAC11-MID/8]
Schiff said that they could look at old subject headings that should be genres, where
it was not clear if it was meant to be a topical or form heading. Mandelstam said
that this is a natural place to look, but there aren‘t that many of them, and
Page | 7
identifying a work as being about a subject or being something is difficult to do
automatically. Some K schedules have form tables, but the schedules may not be
totally consistent; they may need to start with one schedule and see if it works for
the others. They could also use subdivisions (e.g., Legislation). They will test
before applying globally. She hopes the project will be done next year. They are
working with LC on the project.
Mandelstam mentioned that the headings ―International criminal law‖ and
―International law‖ have been proposed, and wanted to know about their status.
Young said they are on the tentative list for this week, but she wasn‘t sure when
they will be approved. Mandelstam pointed out that all books dealing with
international crimes will primarily deal with the legal aspects, not the social,
economic, etc., aspects that general crime books deal with. Mandelstam and Young
agreed that they still need to work on the reference structure, not the terms
1.11 Report of the liaison form the Art Libraries Society (Sherman Clarke)
See written report on ALA Connect (link provided). [SAC11-MID/9]
Clarke noted that art will probably need genre terms as well, and he may ask
ARLIS/NA members what their thoughts are. He asked for advice from other
groups about websites such as Blurb and Lulu, which publish books for artists and
photographers (not books on demand per se) and about how to distinguish them
from other items.
1.12 SAC Research and Presentations Working Group (Linda Ballinger)
This working group was set up at the 2010 Annual Conference. At the time they
got two volunteers, and they now have a mailing list. On that list, they discussed
what to explore, especially in terms of a presentation for the Midwinter Meeting.
They quickly agreed on linked data. They had hoped to have an introduction from
someone at LC about id.loc.gov, but they couldn‘t schedule it. There will be a
presentation tomorrow about a project (The Online Books Page) that John Mark
Ockerbloom is doing with id.loc.gov, and they have advertised it widely. They are
looking for more volunteers and ideas on subjects for presentations, and the group
is figuring out how to share what‘s going on elsewhere. They are considering a
wiki that everyone can post to as they find places where people have written on
subjects or related topics (blogs, papers, etc.). If you want to volunteer, let Joudrey
or Ballinger know.
1.13 SAC Faceting Task Force (Scott A. Opasik)
Opasik provided the written report below:
Page | 8
After Annual, Daniel Joudrey, Stephen Hearn, and I discussed the direction of the
group’s work. It was decided that the direction of the Task Force would be
exploratory, with no specific outcome/product in mind as we started out.
Task Force agreed that we would start by building a bibliography of sources about
faceted subject access.
Kathryn La Barre [of the Graduate School of Library and Information Science,
University of Illinois at Urbana-Champaign] was added to the Task Force after
ALA Annual and offered to set up a wiki as the means to build the bibliography.
The Task Force agreed.
Presently there are only two publications on the bibliography.
Submitted by Scott Opasik
Ballinger asked if the wiki is open to anyone. Opasik said it is currently by
invitation only, but they want to change that, or periodically update it on
ALAConnect. Ballinger says it would be nice if everyone can see it as it is updated,
and they should ask for more participation. Schiff suggested making it public but
The meeting was adjourned at 10:10 am.
Monday, January 10, 2011 1:30 p.m.–5:30 p.m. / SDCC-Room 27B
Members present January 10: Daniel N. Joudrey, Linda Ballinger, Jeffrey Beall,
Christopher J. Cronin, Jimmie H. Lundgren, Scott A. Opasik, Molly D. Poremski, Tony
Olson, Deborah A. Ryszka, Adam Schiff, Phillip Young
Liaisons present January 10th: Julianne Beall, Sherman Clarke, Stephen S. Hearn, Yael
Mandelstam, Joseph Miller, Ed O‘Neill, Hermine Vermeij, Janis Young
Members absent January 10: Mary Catherine Little
Liaisons absent January 10: David Miller
2.1. Welcome and introduction of members and guests
2.2 Update on MARBI (Stephen S. Hearn)
See written report on ALA Connect (link provided). [SAC11-MID/10]
Page | 9
Young asked about the coding type for geographic feature—numeric identifier,
etc.? Hearn said that he believes a code for type of geographic entity should be
entered into the record, corresponding to the parenthetical qualifier. That is, they
are defining an RDA element, therefore a MARC element, identical to the qualifier.
Young was in favor of that type of jurisdiction, but still had questions about the
Schiff noted that as with other RDA elements, this value can be encoded even if not
used as an access point, to facilitate searching, etc., in the authority file. Hearn
pointed to a tension between access using the fullest form of name vs. the form that
gets used in the parenthetical qualifier in the natural name heading, and Schiff
responded that we‘re moving away from the complete formulation of the access
point, so there is no need for the qualifiers to be added to the access point if merely
used to distinguish entities in the authority record. Hearn agreed, noting that
emphasis should be placed on the explanatory information conveyed rather than on
the form of the qualifier.
2.3 Report of the SACO-at-Large meeting (Jeffrey Beall)
See written report on ALA Connect (link provided). [SAC11-MID/11]
2.4 FAST reports
2.4.1. Report of the Subcommittee on FAST (Jimmie Lundgren)
See written report on ALA Connect (link provided). [SAC11-MID/12]
Joudrey asked for the name of the CCS interest group and its charge, and Lundgren
said the name is ―Faceting Subject Access Interest Group,‖ and that the charge is to
discuss theory and application relating to subject terminology intended for faceting
application, including FAST (Faceting Application of Subject Terminology), AAT
(Art and Architecture Thesaurus), LCGFT (Library of Congress Genre/Form
Terms), and others, to ―provide a forum for users and others interested in the
faceting approach to subject access and to compare notes to discuss further
developments and implementations of subject faceting for digital projects and
Schiff, Mandelstam, and O‘Neill discussed whether or not genre/form should be
made explicit in the name of the group, with O‘Neill pointing out that FRSAD‘s
position is that a genre/form is not a subject.
Lundgren mentioned that there is interest in exploring how faceting works for the
discovery systems with elements other than genre/form. She also point out that the
―tag cloud‖ displays can also show similar relationships.
Page | 10
Schiff suggested that the name of the interest group be broadened so as not to be
Joudrey said that there are some overlaps with the SAC Faceting Task Force, and
that a connection should be made with Scott to clarify the issues. Hearn called for a
written charge, and Olson agreed that a charge was needed.
2.4.2 Update of the FAST Project (Ed O‘Neill)
See written report on ALA Connect (link provided). [SAC11-MID/13]
Hearn asked if small record sets (weekly, etc.) could be uploaded for conversion,
and O‘Neill said this should be the case, but that he is hesitant since the system is
new, and stability is an issue. The situation should become clearer as they get more
experience with it.
Schiff asked if there will be something in Connexion that allows users to do the
same thing to a bibliographic record that‘s sitting there before they export the
record needed to add to the master record, or get those before they export the copy
of the record to the local file. O‘Neill said that there is no such plan at the moment,
but he‘d like to see that capability. As to whether the service would be run across
WorldCat, he answered that they‘ve thought of it, but it must be demonstrated that
it is practical to do so.
To Ballinger‘s question concerning experimenting with mapFAST in the non-
MARC database, O‘Neill answered that some libraries (including the National
Library of New Zealand) are using some aspect of FAST with non-MARC, but he
doesn‘t have a list of them. He offered to send more information via e-mail.
Schiff asked if the 1,813 authority records include only LCGFT terms or other
terms, and O‘Neill said that there is not a perfect match, but more will be included
Mandelstam asked how the conflict between a former 185 that is now a 155 in
FAST is resolved. O‘Neill replied that they are mapped as well as possible—all
185s and subfield v‘s—and Mandelstam asked what happens when the existing 185
defines a thing differently than a new 155: is the 185 deleted? O‘Neill explained
that the record is deprecated, and a 700 field [i.e. 7xx field; the exact tag would
depend on the facet] is included in the deprecated record that links to the
replacement .The previous term will map to the new term, like a cross-reference,
and if it doesn‘t normalize in the same form, the heading will be changed, and a
new record will be created with the corrected heading. Mandelstam asked if the
original record number is retained in the authority record, and O‘Neill said no, but
it is logically linked, so the information is available.
Page | 11
Janis Young asked which record, in a case in which the 185 and 155 terms are used
differently, is deprecated, and O‘Neill said that the 155 will be favored.
Mandelstam pointed out that in the LC authority record for 155 there is no link or
reference to 185, 181, or 150, so the issue would pass unnoticed. O‘Neill agreed
that this is difficult, but that duplicates are deleted whenever possible.
Olson mentioned that MeSH should not be forgotten in the list of faceted thesauri,
but that this does not need to be put in the charge.
2.5. IFLA liaison report (David Miller & Ed O‘Neill)
See written report on ALA Connect (link provided). [SAC11-MID/14]
2.6. Report of the Subcommittee on Genre/Form Implementation (Adam Schiff)
See written report on ALA Connect (link provided). [SAC11-MID/15]
Schiff stated that the subcommittee met on Saturday, January 8. There were reports
from the Library of Congress (LC), the American Association of Law Libraries
(AALL), the Online Audiovisual Catalogers/Cataloging Policy Committee
(OLAC/CAPC), and Music Library Association (MLA). Agenda and accompanied
documents posted on ALAConnect.
He stated, ―The ultimate goal is to send this list with recommendations to Janis and
recommend that the Library of Congress add to its project. One additional project
would be the recommendation to convert, create certain 155 genre terms, and
possibly to consider revisions in some cases to the existing form subdivisions. In
some cases, we were suggesting that 185s be turned from singular to plural form. In
some cases, we recommended expanding the scope of the term.‖
One example is the religious form subdivision limited to Christian denominations.
It was felt that this is an unnecessary limitation and should apply to all religious
faiths and denominations. In very few cases the subcommittee recommended some
cancelation or replacement with something else. Schiff said he assumed that SAC
would have to approve this before it goes to LC, and asked whether this could be
done before Annual or would have to be done at Annual. Joudrey said that this
should be possible before Annual, but the vote would need to be reconfirmed at
Annual. Schiff said that the project should be completed in the next few months.
He mentioned the next project, agenda item 7, which is facets that are related to
genre/form. Discussion revealed that there are things that they want to put in as
genre terms but that aren‘t genre/form—things like audience, setting, nationality,
ethnicity—for example, Poetry by African Americans, or Music for children. Many
are in LCSH already. For Russian literature, there is a language and a nationality
component. About sixteen were identified. Though they have not been discussed,
the structure for dealing with them is in place. Mandelstam suggested adding
Page | 12
another term, ―denomination,‖ instead of creating a genre for separate
denomination associated with them.
Schiff initiated a discussion of several issues under consideration by the committee.
―Things like geographic origin never work. People want to search for all the
Argentine films in your collection, but that is country of production information.
People wanted LC to allow subdivision, but LC said no. How do we encode that?
Geographic setting… You have a novel set in San Francisco—ethnic, nationality,
audience, often class of persons information. Example in LCSH: Cancer patients‘
writings. Is the term itself a genre/form or is the writing part the form but ‗Cancer
patients‘ is the class of person. Also chronology, type of literary form…?‖
He said that the committee would discuss these issues once the 185/155 project was
completed in March. At that time, discussion will begin and decisions will be made
concerning which issues will be undertaken, including what to do about overlap
(e.g., nationality vs. geographic origin). The work will be divided up among
committee members, and a structure put in place for each category. Existing
MARC formats, both bibliographic and authority, will be examined and evaluated.
Considerations will include the possibility of new fields or subfields or new
indicator values, followed by discussions concerning recommendations and how
information can be linked. If a thing is faceted, how can it be linked back? For
example, a collection of poetry by children and short stories by some other classes
of person: when these are faceted, how are they brought together for correct
linkage? In MARC, there are linking-fields available, but they are not currently
Because an hour and a half meeting will not be sufficient to deal with these issues,
the committee agreed to meet for up to an additional four hours on Tuesday in New
Orleans, and the result will be a recommendation to LC and to the community
about best practice, and possible MARBI proposals.
Schiff continued that the committee did not get to other agenda items, notably
talking about future projects to deal with genre/form. On LC‘s timeline, the current
genre/form projects are to be completed by the end of 2012.
The committee will make recommendations to LC about some other projects. In
doing 185/155 evaluation, one area that stood out was the need for an Art/Visual
Resources/Architecture, etc., project to do with genre/form. The religion project is
already underway, and the committee felt it would be presumptuous to appropriate
the genre/form discussion for the future art project, since the art visual resource
community knows best what might be needed for genre/form terms in LCGFT. A
similar point was made about an archival project. Therefore, recommendations will
be made to LC about the areas the committee thinks are high priority that LC might
want to do next.
Page | 13
O‘Neill asked about considering a separate field to break out national origin or
places, or country, and Schiff noted that there are already many places in MARC
for that information. O‘Neill said, ―I have always had a problem with headings, for
example, ‗Australian Science Fiction,‘ because all science fiction comes from one
country or another, so you end up with the only things cataloged with Science
fiction [are] collections from many collections or when you don‘t know what
country it came from. So if you could break out the country of origin into a subfield
or class, then Science fiction would apply to every item. The same would apply to
the creative or the intended audience of things like children‘s poetry.‖
Schiff noted that the group thinks ―we are heading to the faceted way rather than
pre-composed string of long complicated terms, e.g., Poetry for African American
Jewish lesbians. We have to know what part of that information belongs in the
authority record and what part belongs in work authority record versus the personal
name authority record.‖ Does the nationality belong in the name authority record or
the record for the individual work? In order to function, these linkages must be
worked out. In the short term, it might have to be in the bibliographic record. A
way to distinguish the intended audience and the creator‘s class of persons is
required. The music and literary communities are dealing with this, and there must
be something for children within the religious genre.
Young said that those working on the religious genre have age level, audience level
and denomination terminology that they are trying to deal with, and Clarke added
this is also the case in art.
Schiff will prepare a written report.
2.7 Report of the OLAC/CAPC Moving Image LC Genre/Form Heading Best Practices
Task Force liaison (Deborah Ryszka)
See written report on ALA Connect (link provided). [SAC11-MID/16]
Ryszka said that one of the unresolved issues being dealt with is defining and
finding examples of things like webisodes, podcasts, and other items that originally
appeared on the Web. Work is progressing, and there will be more to report at the
annual conference. Details of the work are in the submitted written report.
2.8. Report of the chair of SAC (Daniel Joudrey)
See written report on ALA Connect (link provided). [SAC11-MID/17]
2.9. SAC Genre/Form Pre-conference Update (Janis Young & Daniel Joudrey)
Joudrey reported that they are still working on content; speakers are in the process
of developing their presentations. The preconference will address both theoretical
Page | 14
and practical components. It will have exercises, historical background, and
information about the experiences of both MLA and AALL in identifying
Young said that they want people that attend the preconference to feel comfortable
and confident in assigning genre/form headings, to understand where the committee
is coming from and where it is going with the genre/form project, and to understand
the faceting aspect. They want people to have a good grounding in the genre/form
thesaurus—how it applies and how it has been developed. The preconference will
be held Friday, 8:30 a.m.–4:30 p.m.
O‘Neill expressed disappointment that the faceting group would be held at 4 p.m.
on Friday, overlapping with the genre/form meeting, but Young noted that the only
day available for a preconference is Friday. Clarke said that the
preconference emphasis should be on LCGFT rather than faceting, and Young
replied that it would be on assigning headings from the new thesaurus and answers
to the questions people have about it, using examples. Schiff asked if best practices
will be ready by then, since people are interested in assigning these terms but often
―don‘t get it right in the right way.‖
2.10 H1095 Project with LC Update (Janis Young & Daniel Joudrey)
Phase 3 of 1095 project with LC: There has been discussion on what needs to
happen with some subdivisions. In some of the areas, there is a consensus, and in
others there is no agreement on certain subdivisions. Some of the issues are
―Quotations, maxims, etc.‖ and ―Quotations‖—looking at combining those two;
another one is ―Rating of‖ and ―Evaluation.‖ The primary use of ―Rating of‖ is
with class of person in H1100, and ―Evaluation‖ is used for everything else. The
two terms could possibly be combined. There is agreement on using ―Evaluation‖
instead of ―Rating of‖ for classes of persons. Another is the area of ―Terminology,‖
―Glossaries, vocabularies, etc.,‖ and ―Vocabulary.‖ ―Terminology,‖ ―Glossaries,
vocabularies, etc.,‖ and ―Vocabulary‖ could be combined, but there is a question
about ―Conversation and phrase books,‖ which really belongs to genre/form, so it
could not be combined with the others. Another example is ―Acronyms‖ and
―Abbreviations‖--it is recommended that these should be combined as ―Acronyms
and abbreviations.‖ Another example is ―Education‖ and ―Knowledge and
learning‖; it is recommended that these be combined. ―Education‖ applies to
individual and ―Knowledge and learning‖ to classes of persons. This information is
available to the SAC members. Terms having no consensus or agreement include
―Ability testing‖ and ―Testing.‖ There are a lot of disagreements about whether
―Testing‖ could replace ―Ability testing,‖ since some hold that ―Ability testing‖
should be part of the subject heading instead of the subdivision.
Another issue discussed concerned identifying books of a general nature;
suggestions included ―Introduction,‖ ―Overview,‖ ―Basic information,‖
―Compendium,‖ ―Primer,‖ ―Introductory treatment,‖ and ―Core knowledge.‖
Page | 15
The last issue discussed was the use of ―Biography‖: some want to see ―Biography‖
authorized under individual names, and there is some concern about retrospective
work or splitting the catalog.
Young said that ―Biography‖ used to be authorized under individual names, so it
should be referred to the previous policy. How much biographical information is
required in order to get the subdivision applies in the personal name headings or
classes of persons headings.
According to Joudrey, this is a continuing effort; he would like more people to be
Mandelstam said that coordination is needed between this and the 185/155 project;
at some point there should be a discussion and recommendation. Joudrey asked for
suggestions on how to proceed with the project. Mandelstam suggested doing it
group by group through ALA Connect; the list can be recast to narrow the
discussion. Young said that she prefers an alphabetic approach and suggested both
the genre/form works on 185/155 and that this group work on H 1095 posting and
have a discussion [on ALA Connect], so whoever is involved can see it.
2.14 Presentation on SKOS and its application (John Mark Ockerbloom, University of
See PDF file on ALA Connect (link provided). [SAC11-
John Mark Ockerbloom, digital planner and researcher at the University of
Pennsylvania, gave a presentation on open and linked subject authority data from
the Library of Congress (http://id.loc.gov/) and how it can be used to improve
subject cataloging and browsing. Ockerbloom described how the LC data is
structured (including a brief overview of RDF (Resource Description Framework)
and SKOS (Simple Knowledge Organization System), and explained how he used
it to improve both the Online Books Page (http://onlinebooks.library.upenn.edu/)
and the catalog at the University of Pennsylvania. In addition, he also discussed the
augmentation and enhancement of authority data from LC, and discussed how
widespread use of linked bibliographic and authority data may change how we
build and use catalogs in the future.
Ockerbloom answered questions after his presentation:
Q1: Sherman Clarke asked how the Hamlet page was constructed in the Online
Page | 16
A1: It was semi-handcrafted. Ockerbloom created a work record for Hamlet and
linked bibliographic records using identifiers. The work record had text fields with
notes and descriptions, which were just supposed to give the user enough
information to be sure they identified the correct play and told them basic
information about it. He categorized (he seemed to start to say ―classified‖) the
editions and placed their identifiers in that category. Clarke also asked about his
almost-use of the word ―class.‖ Ockerbloom said he used his own classification,
not a standard one. Some of the page was automatically built, because since the
name-title headings have identifiers, he could query those and find out what records
used them as a subject heading and link to them. This could be extended to a
database as large as WorldCat as well. Schiff asked where the parenthetical
description came from (HTML, PDF, etc.). Ockerbloom said that in this case, he
had to code it, but if he is pulling from someplace like HathiTrust, he can use their
Dublin Core record to get the data.
Q2: Martha Yee asked if Ockerbloom has a position on whether subject is a class or
a property in RDF? For example, with a book by Shakespeare, person is clearly a
class, but if there‘s a book about him, he could become a property. If subject is a
property, there are places in RDF where it can‘t go, because they won‘t take
properties, and the data becomes too complex to fit into RDF.
A2: His position is ―whatever works,‖ and he pointed out that when faced with any
large set of cataloging data, especially if it comes from various linked sources,
different people will make different decisions about how to organize that data, so
some might have Shakespeare as a class and some might have him as a property.
Ideally, his data would do the basics and recognize that a particular record is about
Shakespeare. He realizes this is imprecise, but he feels we need a certain amount of
imprecision. However, he also feels we could spend forever discussing this and
never come to a conclusion.
Q3: Christopher Cronin asked about the possibilities for collaborative efforts with
departmental and faculty projects on campus, collaboration between library and
other organizations on campus that have and manage data such that everyone wants
to have the data together in some kind of way. Where does Ockerbloom see it
playing out with other people‘s data on, e.g., university campuses or research
institutes? Does he see id.loc.gov‘s data being used in this instance?
A3: This could definitely happen, but it depends on the field. He thinks scientists
would not be as interested, because they write articles more than books, but in the
humanities, a faculty member might want to link to our data to talk about the books
they have written that relate. He doesn‘t know how things will fit together, but he
thinks they will. The WEbook Project is trying to assemble data about faculty
projects across campuses, talking about what they‘re publishing and working with.
The focus is on not recreating data, so they‘re using concepts like open researcher
IDs, which link them to databases. He suggests you could link ORCID name
identifiers to LC name authorities. Once you do that, you can link scientists to their
Page | 17
articles and books that mention them. Linked data allows us to build bridges
between datasets that already exist.
Q4: Schiff mentioned that we aren‘t putting many URLs in subject authority
records, but there are a lot in name authority records. If those are in SKOS, can you
link directly to them? He gave the example of the record for Southern Sudan, which
has links to their constitution and the main government page.
A4: Probably, but he doesn‘t know off the top of his head, because he doesn‘t
remember all the fields present in SKOS. SKOS is simplified, and a work in
progress. There will be features added, but right now it doesn‘t have all of them.
One of its features is simplicity. RDF records can combine terms from various
vocabularies, including SKOS, and mix and match with anthologies that do have
Q5: Daniel Lovins (Yale University) asked whether the "Linked Data Cloud,"
which Ockerbloom showed on a diagram [slide #6], really exists.
A5: Ockerbloom said that it does really exist, and that there are ways of linking the
data more dynamically. He thinks we really are using linked data even if we‘re
downloading it. Even if we download a set, we link to or query individual items.
The Shakespeare page isn‘t all downloaded; much of it is just links--download
what‘s useful to process locally and link to the rest. Lovins followed up: If data is
downloaded onto a server, it‘s accessible to others. To what extent does the RDF
model lend itself to local data as opposed to linked data? Ockerbloom responded
that Google Books is doing something similar to linked data; ―Find in a Library‖ is
a link that is based on a search they run. If the link isn‘t relevant, it doesn‘t display:
it depends on the application and what is most efficient for it.
Q6: Eva Bolkovac (Yale University) asked about the identifiers in the bibliographic
records that linked back to the work record for Hamlet. What identifiers were used?
A6: Ockerbloom said he uses internal identifiers for now; he considered using the
name/title identifier, i.e., ―Shakespeare, William, 1564-1616. Hamlet,‖ but there
aren‘t enough to be useful. If a standard work identifier comes along, he will
happily switch. He feels there are a lot of local projects waiting for standards.
Q7: Jeffrey Beall asked about Ockerbloom‘s Toronto example, he said he used
lexical analysis to generate some of the pages. How does he deal with ambiguous
terms like ―Mass‖ (which could be the place, or related to physics)?
A7: Another example of problems are inverted terms (―Philosophy,‖ ―Philosophy,
Jewish‖). We can usually automatically infer between ―Term‖ and ―Term,
[something],‖ (―Term, [something]‖ is a narrower term to ―Term‖) but it doesn‘t
always work. (―Israel, Mark‖ is not a narrower term to ―Israel‖) In his database, it
occurs rarely enough that he uses an exception list, and if he has subfield types that
Page | 18
tell him what kind of term it is (geographic vs. name) it helps. Clarke says Flikr
already has some disambiguation, and wants to know if he can do something like
using context to help distinguish. Co-occurrence analyses can help: if ―African
American abolitionists—Biography‖ and ―Sojourner Truth‖ often both occur as
subject headings on the same books, we can reasonably conclude that Sojourner
Truth is an African American abolitionist. People who write about the same thing
many times need to be excluded. Schiff says we need to be careful about multiple
works on different subjects in one carrier. Ockerbloom says he is conservative
about this analysis; if there are other names in the record, he doesn‘t make the
inference. We need to be pragmatic and figure out if the inference gives us
Q8: Janis Young asked about the MADS (Metadata Authority Description Schema)
interface, which Ockerbloom mentioned people have said is too complicated. She
had not heard that before and wants to hear his impression.
A8: This is secondhand knowledge for him, so he can‘t be sure what people said.
Part of the objection was how MADS is being represented as XML, and he doesn‘t
want to give too many specifics as this was not his experience, but something he
heard. He thinks one criticism was that new RDF terms were created when similar
ones existed in other ontologies, and those could have been reused. He mentioned
that sometimes it‘s good to come up with your own ways of organizing, and
sometimes it‘s useful to use what‘s already out there. He knows that SKOS doesn‘t
perfectly capture everything in MARC authorities, but thinks it‘s close enough for
him for now. Once MADS is available, SKOS will still be available, so he may
transition to MADS eventually, but if he doesn‘t find that it works for him, he‘ll
keep using SKOS.
Q9: Schiff said that Ockerbloom‘s ―Scientific‖ example stated the differences
between subject and genre, because some of the terms Ockerbloom has will
become genre terms coded as such. Schiff asked if Ockerbloom will be working
with LCGFT terms.
A9: He‘s considered it, but LCGFT is still in flux, so he‘s waiting to decide, once it
settles down. Young pointed out that those terms come with the download of
LCSH. Schiff asked how they are identified. Ockerbloom said they are tagged,
although he doesn‘t remember how. For now, he is ignoring genre terms unless
they are also subject terms.
Q10: Joseph Miller said he thinks links between topical subject headings and
names are useful, because often reference librarians are asked for biographies of a
certain kind of person (such as African American abolitionists), but often the
category of persons heading isn‘t on records for such biographies, because it‘s not
about that category as a whole and some people fall into multiple categories. If you
could link individual people to categories they exemplify, you could help users find
them with a simple search.
Page | 19
A10: Ockerbloom‘s wife has a site [http://digital.library.upenn.edu/women/] that
does exactly that, using the Online Books page and other data. He knows the
heading isn‘t always there, but he likes when it is, even if it‘s not entirely accurate
as far as the bib record is concerned. With RDA, this information can appear in the
authority record, or someone might create an auxiliary database that could be
queried at the same time. Schiff pointed out you can say profession/occupation and
field of activity, but ethnicity is not available, so we could say Frederick Douglass
was an abolitionist and can say he‘s associated with the United States, but not that
he was African American, which is why we are still discussing facets, and they are
currently in unprocessable notes. Ockerbloom says if someone we trust can make it
available as linked data, we can use it.
Q11: Sherman Clarke asked Ockerbloom to give an example of things that
LibraryThing is doing that libraries could use. He heard that LC will be linking to
the information users have provided about fictional characters.
A11: Ockerbloom said it is not directly using LibraryThing but likes it as an
example of how amateurs can enhance data. Young added there has been
discussion, but she‘s not clear on the details. LibraryThing has data about fictional
characters that are not currently established in LCSH, because the character has
appeared in fewer than three works. She doesn‘t know the status, but there is
ongoing discussion about unmediated linking from bib records. Schiff pointed out
that in RDA, a relationship between a fictitious character and a work could be
expressed. Clarke said you can do the same for places.
2.13 Open discussion/ Open announcement period
Yael Mandelstam has asked Ed O‘Neill to provide user statistics about subdivision
usage in OCLC. Once she has the data she will share with the SAC list.
Joseph Miller suggested providing documents on paper for SAC attendees. Scott
Opasik asked if Ockerbloom‘s presentation will be available. Joudrey said yes and
Ockerbloom agreed to produce a static version of his presentation inPowerPoint.
Linda Ballinger added that it will be posted on ALAConnect.
The meeting was adjourned at 4:35 p.m.
Tachtorn Meier, Intern
Elizabeth Bodian, Intern
Page | 20