[DCRM-L] BSC comments on Future of Bibliographic Control

[Randal Brandt]

Following are the comments that were submitted to the LC Working Group on 
the Future of Bibliographic Control on behalf of the Bibliographic 
Standards Committee. Thanks to Larry Creider, Nina Schneider, and Deborah 
Leslie for pulling everything together, and thanks to all of you who 
commented on the earlier draft. This was a group effort.

Randy

****************************************************************************************************
The Bibliographic Standards Committee (BSC) of the Rare Books and Manuscripts
Section (RBMS) of ACRL welcomes the opportunity to comment on the Report on the
Future of Bibliographic Control drafted by the Library of Congress (LC) 
Working Group
(WG) on the Future of Bibliographic Control. The Bibliographic Standards 
Committee is
especially encouraged by the fact that access to rare and unique materials 
is recognized as
one of the five central themes within this report.

The Bibliographic Standards Committee strongly agrees that, indeed, there 
are many
institutions and organizations that have the expertise and capacity to step 
forward and
play significant roles in the bibliographic future. Only if we are allowed 
and encouraged
to do so will this be successful.

Although the Bibliographic Standards Committee is most interested in the 
second theme
of the report relating to the exposure of rare and unique materials, we 
realize the
recommendations of the entire report will impact rare book and special 
collection
repositories just as significantly.

What follows are specific comments on many aspects of the report.

In general, the report ignores the impact of the vast quantity of materials 
on the Web and
the issues of selection and providing access to them, except by talking 
indirectly about
making use of metadata.

Page 6: It is gratifying to see that the Library of Congress has learned 
that announcing
major changes to the library community without advance preparation, as 
happened with
their series decision, is not the way to introduce change. The ability of 
the cataloging
community, together with OCLC, to discover ways to cope with that decision 
is an
indication that the Library of Congress can indeed rest easy in the 
knowledge that there
are organizations and institutions who are willing to step forward and 
volunteer their
expertise in cataloging standards.

Page 7, Paragraph 2: The Bibliographic Standards Committee agrees that 
users would be
better served if access to a variety of materials were provided in the 
context of a unified
philosophy of bibliographic control. Unfortunately, this desire is most 
likely impossible
to realize given the commoditization of information and proprietary 
considerations.

Page 7, Paragraph 3: The Bibliographic Standards Committee strongly 
disagrees with the
statement that "Consistency of description within any single environment, 
such as the
library catalog, is becoming less significant than the ability to make 
connections between
environments." Both are important; minor inconsistencies are tolerable both 
within and
between databases. Major inconsistencies need to be remedied lest they 
result in chaos.
Introducing a database with no authority control into one with authority 
control
eliminates any authority control and adequate precision or recall in both 
databases.

Page 7, Paragraph 4: What is meant by "cataloging?" Perhaps this needs 
redefinition as
much as "bibliographic control." The Working Group seems to use "cataloging" to
denote all that is old and outdated and "bibliographic control" to denote 
all that is new
and good. It is "necessary to embrace a view of bibliographic control as a 
distributed
activity, not a centralized one." Fortunately, this has been happening for 
a few years
thanks to the Program for Cooperative Cataloging (PCC) and OCLC's master record
concept.

Page 8, Paragraph 1: The Bibliographic Standards Committee is very worried 
that the
Working Group accepts the commoditization of information without a nod to 
the notion
of information as a public good. Does this mean that the Library of 
Congress will begin
charging for its services? Should we expect to see pop-up ads on 
Classification Web in
the near future?

Page 9-10: Redefining the Role of the Library of Congress. As mentioned 
earlier, the
Bibliographic Standards Committee agrees that LC might want to reconsider its
responsibilities to the nation's libraries. There are many experts in 
public, academic, and
special libraries who would be willing to work with LC to create 
partnerships and
opportunities to improve the universe of bibliographic control. This will 
be a new world
for LC and a learning experience for all involved, but a public recognition 
that the locus
of expertise has given way to many loci is necessary. As LC is aware, the 
Bibliographic
Standards Committee is very involved in the creation and maintenance of 
standards for
the rare materials community.

Page 10: The shift of bibliographic control of primary resources within LC 
does not mean
that traditional cataloging practices must be abandoned. The traditional 
practices need
revision and need to be coupled with other means, but they will continue to 
be necessary
for materials of all formats. As we will see in Recommendation 2, 
traditional cataloging
practices will be necessary to make some of these resources available. 
These resources do
not leap into malleable electronic metadata without costly human 
intervention and
intellectual activity.

Page 11, Section 1.1: Eliminating Redundancies. Descriptive metadata does 
not account
for most of the cost of bibliographic control, so although improved sharing 
is a good idea,
it will do little to reduce costs. Even classification could be more 
automated than it
currently is. The real, barely reducible, cost comes with controlling 
names, titles, and
subjects.

Page 13-14, Section 1.2: Copy cataloging and loading of LC authority files 
are not the
only reasons that libraries have reduced staff. Budget cutbacks and lack of 
qualified
professionals, as well as the soaring costs of the materials and licensing 
should also be
considered.

Page 15: Overall, the Bibliographic Standards Committee agrees with 
Recommendations
1.2.1.1-1.2.1.3. LC could make better use of PCC-produced data. If the 
recommendations
in 1.2.2 to examine original cataloging programs and sub-programs at the 
Library of
Congress are to work, LC will need to do a better job than it has in the 
past of identifying
and working with other entities. As mentioned above, LC will need to 
explicitly abdicate
from the library community's expectation that it is THE source of knowledge 
in these
areas and direct some inquiries to other institutions or groups.

Page 16: Recommendations 1.2.3 and 1.2.4. The problem is not simply the 
number of
PCC participants; the problem is with institutional barriers to expansion 
of PCC
participation. For example, when NACO and BIBCO catalogers move from libraries
where they have been trained in NACO and BIBCO procedures and made 
contributions
to libraries that are not members, their expertise is lost. The 
Bibliographic Standards
Committee also recognizes that libraries need to expand the number of 
certified librarians
within their institutions. NACO certification, and perhaps BIBCO 
certification, should be
attached to catalogers and transferable with the cataloger. There should be 
investigation
of ways OCLC can encourage small, specialized libraries, to achieve Enhance 
status,
which is necessary for BIBCO membership. OCLC in particular should consider
amending its loading algorithms and rewards for upgrades and corrections to 
avoid M-
level records with 20 libraries attached whose catalogers have done work in 
their local
catalogs but found it too time-consuming to make the changes in the "master 
record."

Page 16-19, Section 1.3: The section on collaboration in authority record 
creation is
excellent, particularly in its recognition of the inescapable amount of 
human intellectual
effort that is devoted to authority work. A great deal could be done here 
by making
participation in cooperative authority processes easier, as mentioned 
above. The
Bibliographic Standards Committee also notes that rare materials libraries 
and special
collections are often excellently positioned to be able to contribute to 
the creation and
maintenance of national authority files. NACO participation among special 
libraries
should not only be encouraged, but should be facilitated by broader 
training and
mentoring programs.

Page 18: Again, the Bibliographic Standards Committee is willing to work 
with LC to
increase collaboration on authority data and controlled vocabularies.

Page 19-21: Enhance Access to Rare and Unique Materials. On the whole, this 
is an
excellent section. However, some things might be added. First is that 
increasing
cooperative collection development by research libraries will mean that 
more current
material, particularly from foreign countries, will become "rare" and 
require original
cataloging because no one else holds these materials. Research libraries 
need to commit
to the fact that creating unique collections will require more resources 
for bibliographic
control of those materials.

Page 19, Paragraph 3: "Few models exist, however, for how such trade-offs 
might be
made." See Mark A. Green and Dennis Meisner, "More Product, Less Process:
Revamping Traditional Archival Processing," American Archivist, 68 (2005): 
208-263.
Although principally about archival processing, this article also touches 
on access and
description and discusses the implications of the various trade-offs that 
must be made.

Page 20, Recommendation 2.1: Rare book and special collection repositories have
recognized the need to make the discovery of their materials possible. The 
Bibliographic
Standards Committee has gone a long way towards this goal in codifying 
rules for
various formats of materials and in emphasizing access to a greater number 
of items and
empathizes with LC in prioritizing these materials.

Page 20, Recommendation 2.2: "Streamlin(ing) cataloging for rare and unique 
materials,
emphasizing greater coverage and access" is an oxymoron. One can streamline 
some of
the processes, but poor access will hinder users from finding the materials 
they need
almost as much as providing no access at all. When thinking about the 
different levels of
2.2.4 and the "some level of access" of 2.2.1, institutions need to 
remember the way in
which broadsides were ignored not so long ago and the way pamphlets were bound
together and given an assigned title with little or no access to the 
individual pamphlets.
Individual items of these types often have great research (and monetary) 
value and are
used in serious scholarship. By not providing full access, libraries hinder 
discovery by
both scholars and collection development staff who must decide how best to 
spend
limited resources on new acquisitions. The Bibliographic Standards 
Committee notes that
there is tension between recommendations 2.1 and 2.2. There is danger in 
merely shifting
unprocessed materials to under-processed materials by adopting an 
uncritical approach to
"some access." The Working Group cites the ARL White Paper on "Hidden 
Collections"
(on p. 19) and would be well-advised to consider the approaches examined 
there more
carefully.

Page 20, Recommendation 2.3: Integrating finding aids and databases and 
metadata
records into the discovery tools for rare and unique materials is a 
wonderful idea.
However, these systems need to have the ability to limit to or exclude such 
materials.

Page 21, Recommendation 2.5.1: Is it necessary to share metadata for unique 
materials?

Page 21, Recommendation 2.5.2: Success will depend on OCLC's ability to offer
federated searching on institutional records and to make it easier for 
finding aids or
images, for example, to be loaded into OCLC.

Page 21-22: Position Our Technology for the Future. It would be foolish to 
adopt a
replacement for MARC that does not retain the advantages of MARC (e.g. 
subfields for
searching) or that does not address the major defect of MARC, namely its 
inability to
handle hierarchical relationships. Would it not be better to work with Web 
developers to
increase the relevance and ranking of library catalogs and standards and to 
index MARC
records in such a way as to increase visibility in search engines? 
Catalogers and
programmers will need to work together to make this possible.

Page 23-24, Section 3.2: Standards. The report states that "it is through 
consistent
application of standards that the full value of bibliographic data can be 
released across
many potential use environments" and that standards are in reality, a 
business issue.
Standards not only remove barriers, they also impose barriers, particularly 
those relating
to cost. Standards require conceptualizing data in a certain way. A better 
argument for
standards is the utility for the users of bibliographic data, which goes 
along with
recognition of their limitation. We need to recognize the limitations and 
shortcomings of
standards along with their undeniable value and not blindly endorse 
anything simply
calling itself a "standard."

Page 25, Recommendation 3.2.1-3.2.2: Suspend work on RDA. This makes sense, 
even if
not for the reasons given. Some in cataloging community think that it goes 
too far, others
think it doesn't go far enough. Until ALA and LC agree on a format for the 
recording and
display of data, and detailed statements on encoding, existing catalogers 
will find it very
hard to implement RDA. The Bibliographic Standards Committee also wonders 
how the
recommendations in 3.2.2 will, if at all, affect Descriptive Cataloging of 
Rare Materials.

Page 26, Paragraph 2: It is unlikely that machine applications will ever be 
the primary
users of bibliographic data. They may be one of the major manipulators of 
such data, but
the users remain human, with human objectives. It is enough to say that we 
need to
structure encoding schemes that support such manipulation.

Page 26, Section 4.1, Paragraph 3: "... most users now conduct their 
research in multiple
discovery environments: search engines, online booksellers, course 
management systems,
specialized databases, library catalogs, and more." This is similar to the 
what users have
always had to do, consulting catalogs, periodical indexes, newspaper indexes,
bibliographies, printed catalogs, microfilm, etc. The difference is that at 
one time,
researchers had to travel to different institutions or use different 
machines. Most of this
can now be done from one computer, which gives the illusion that these 
resources should
all work the same way. Unified searching may be an unattainable goal.

Page 27, Paragraph 2: The report's assertion that library users value 
features and data that
help them make sense of results by ranking, organizing, and clustering, may 
or may not
be true. Recall and precision of results is just as important, if not more so.

Page 27, Paragraph 3: It is very disturbing to learn that the Working Group 
considers that
a library catalog should be designed to ingest or interact with records 
from sources
outside of the library cataloging workflow. Unless there are stringent 
guidelines and
intense oversight, any idea of authority control or standards will be 
negated. If LC is
concerned with the overwhelming responsibility and staffing issues of 
updating and
maintaining its records, mixing controlled and uncontrolled bibliographic 
data would be
foolhardy decision. Rather than abandon its controlled catalog, LC should 
consider the
alternative of using its catalog as a base.

Page 27, Paragraph 4: "Many libraries have chosen to produce metadata to 
satisfy the
needs of their most sophisticated users, despite the fact that such users 
are but a small
percentage of their total user base. They do so under the increasingly 
dubious assumption
that all users will benefit from the greatest detail in cataloging." This 
statement
contradicts the whole tenor of Recommendation 2 and promulgates an 
atmosphere of
anti-intellectualism. Ideally, we, as information professionals and 
librarians, should
encourage discovery and deeper understanding in everyone who makes use of a 
library.
Nevertheless, the fact is that discoveries and research are made by the 
small number of
users who make intensive use of primary sources and secondary literature. 
This is as true
of scientists as of humanists.  Such intensive use of information requires 
complex tools to
study the very complex reality we live in. What is required in the library 
catalog is detail
sufficient to distinguish resources so that the user may discover and 
select them. The
larger the database, the more sophisticated the user, the more detail may 
be required.

Page 28, Recommendation 4.1.2.1: It is reassuring to see that LC recognizes the
importance of maintaining the "integrity of library-created data."

Page 28, Recommendation 4.1.3.1: Algorithms need to go a long way before 
they are
useful in suggesting works that might be useful to patrons. For example, 
commercial
search engine algorithms fail to distinguish between popular and scholarly 
material.

Page 28, Recommendation 4.1.3.2: If implemented, this will take longer than 
creating
original controlled vocabularies and their variations.

Page 28, Section 4.2: Realization of FRBR. What does the statement "FRBR 
suggests
alternatives for analyzing intellectual content for bibliographic control" 
mean? FRBR
proceeds from what users, do or are supposed to do, and posits ways of 
constructing
relationships and displaying those relationships.

Page 29, Paragraph 2: Developing a means to exchange work-level data will 
take thought
but is hardly less feasible than many of the other suggestions the report 
adopts, such as
recasting LCSH in a hierarchical structure. The cataloging rules are part 
of RDA. In the
report, the Working Group calls for the suspension of work on rules that 
will support the
creation of authority records using FRBR, but then attacks the FRBR model 
for not
having such cataloging rules. The real problem will be finding the 
resources to apply
FRBR retrospectively and to materials whose metadata is inadequate for any 
but the most
basic purposes.

Page 29, Recommendation 4.2.1.3: There are at least as many problems with 
the concept
of Manifestation as Expression. Nonetheless, the fact that these 
determinations are
flexible and subject to development as scholarship intensifies on a 
particular group of
Works, with the related Expressions and Manifestations, allows for the 
model to stay
alive.

Page 30, Recommendation 4.3: Although the Bibliographic Standards Committee is
particularly interested in description, subject access is an important 
aspect of discovery.
This section, and the diagnosis of problems, ignores the basic fact that 
LCSH is complex
because reality is complex. Navigating the thesaurus and the Subject 
Cataloging Manual
is difficult, but so too is describing the world of knowledge. Any attempt 
to do justice to
a resource's subject is bound to be difficult. The length of LCSH strings 
is more a
problem of catalog displays. Oddly, LCSH strings are quite useful for 
keyword searching
which can then allow retrieval of other items with the same or similar 
subject strings.
LCSH is not, and should not be, designed for novices. Such individuals are 
best served by
starting with a keyword search and then building on the subject headings or
bibliographies to find other materials.

Page 31, Recommendation 4.3.2: De-coupling of strings is possible, but 
hardly necessary
when keyword searching of subject strings is so effective.

Page 32, Recommendation 4.3.3.2: "Apply terms from any and all appropriate 
sources of
controlled subject headings in bibliographic records to augment subject 
access." How
does this fit in with the streamlining mentioned in Recommendation 1? How 
does this
reduce the cost of cataloging? How will it help the user?

Page 32: The Desired Outcomes for section 4.3 do not flow from the 
recommendations.
The result of the recommendations on LCSH will be more complex subject 
analysis, not
less, and certainly not more intuitive. It will not be easier to update and 
to apply.
Terminology may be more current but certainly not more consistent. The 
application of
the recommendations will require substantial resources that may or may not 
be justified.
Rather than test FRBR, the recommendations concerning LCSH should be tested for
feasibility and user benefits before they are put into practice.

Page 35, Recommendation 5.2.1.2: This recommendation is excellent, but 
until libraries
make clear that there is a demand for graduates with these skills and that 
cataloging and
bibliographic control are the backbone of a library, the recommendation 
will languish.

Above all, there needs to be a commitment on the part of the institutions 
involved to
invest in the expertise required for control of these items: technical 
knowledge of papers,
bindings, printing and photographic processes, reproduction technologies, 
languages,
scripts, subject expertise in areas ranging from chemistry to 
denominational schisms, as
well as electronic systems, encoding schemes, and programming ability. Only if
institutions are willing to value this expertise will there be the 
necessary personnel to
realize these outcomes. In this report, the Working Group never really 
commits itself to
the notion that no matter how many shortcuts can be found, how much electronic
manipulation of metadata is done, libraries will still have substantial 
investments in the
human aspects of bibliographic control. Until this fact is recognized, 
libraries will not be
able to make the changes that will lead to effective implementation of the
recommendations, let alone implement some of the others that are needed.

Respectfully submitted (on behalf of the ACRL/RBMS Bibliographic Standards
Committee),
Laurence S. Creider, New Mexico State University
Nina Schneider, Clark Memorial Library, UCLA
Deborah J. Leslie, Folger Shakespeare Library
Randal S. Brandt, The Bancroft Library, UC Berkeley


__________________________
Randal Brandt
Principal Cataloger
The Bancroft Library
(510) 643-2275
rbrandt at library.berkeley.edu
http://bancroft.berkeley.edu