College and Research Libraries


Book and Serial Union Catalogs: 
A Symposium 

The following three papers were presented at the meeting of the ALA Board on Resources 
of American Libraries, Los Angeles, California, June 25, 1953. 

By C. S U M N E R S P A L D I N G 

Library of Congress Book Catalogs: Proposed Expansion into 
Current A u t h o r and Subject Catalogs of 

American Library Resources 

Mr. Spalding is chief, Catalog Maintenance 
Division, Library of Congress. 

WE H A V E ALL NOTED with regret the absence of Halsey William Wilson at 
the meetings of our 72nd Annual Conference. 
I am informed that there is probably no one 
of us who can remember an Annual Confer-
ence of the A L A at which Mr. Wilson was 
not present. Just as so many of the great 
monuments of the American bibliographical 
scene, the CBl, the many indexes, the card 
service, and the Union List of Serials mark 
high the place that Mr. Wilson and the 
H. W . Wilson Company hold in the realm 
of library services in this country, so also was 
it Mr. Wilson's vision that foresaw not only 
the development of the L C Catalog but also 
its potential development into a tool which 
would hold a key to the resources of Ameri-
can libraries. In his Proposed. Plan for Print-
ing Library of Congress Cards in Cumulative 
Book Form published in 1946, he suggested 
that this catalog should also show the holdings 
of other American libraries. From the begin-
ning, the Edwards Brothers catalog and the 
subsequent L C Catalog have included the 
entries prepared under the cooperative catalog-
ing program and, by showing the name of the 
library preparing the copy, have thereby par-
taken in a limited measure of the nature of 
a union catalog. In the summer of 1952 the 
administration of the Library of Congress 
deemed the time ripe for a serious investiga-
tion of the full potentiality of the LC Catalog 

as a current catalog of national library re-
sources. A committee was appointed to this 
end and it is largely from the report of this 
committee that my remarks this afternoon 
are drawn. This report has been received 
by the Library administration but has not 
as yet been given active consideration. It was 
thought, however, that the profession at large 
should know its contents and should consider 
the proposed expansion of the catalog in terms 
of its desirability, in terms of the particulars 
of its various features, and in terms of its 
financial practicability. 

The committee approached its assignment 
pragmatically. That is to say, it concerned 
itself primarily with the nature, quantity, and 
quality of the bibliographic raw materials 
presently available, the characteristics of the 
bibliographical tools now in existence, and 
the methods and procedures of demonstrated 
effectiveness. It sought to reveal what seems 
to be presently possible with the means at 
hand in order that a useful beginning might 
be made in the near future which might, as 
time goes on, be developed, improved, or ex-
panded in accord with improvements in inter-
library cooperation, methods, bibliographical 
interrelationships, and the like. 

In developing its proposals for the expansion 
of the L C Catalog into what would in effect 
be a current national union catalog, the com-
mittee felt that the Catalog's present character 
as a display of and an index to L C printed 
cards should in no way be altered. It was 
recognized that this would mean a composite 

JANUARY, 1954 15 



catalog exhibiting one set of criteria for L C 
cards and another set for other cards but it 
was felt that this theoretical inconsistency 
will not have any substantially adverse effect 
on its usefulness as a tool for serving many 
varied bibliographical and bibliothecal pur-
poses. 

A determinant of many of the character-
istics of the publication as proposed was its 
relationship to the National Union Catalog 
( N U C ) on cards and to the proposed pub-
lished version of the same. The basis for 
expanding the present LC Catalog would be 
the reports received by the N U C of titles 
held by American libraries. These reports 
are of two general types: those which repre-
sent current cataloging and those which are 
records of the entries in entire catalogs or 
regional union catalogs, or of entries in large 
sections of the same. Since, as will be seen 
below, the committee is recommending the in-
clusion of only recent imprints in the expanded 
catalog, it is proposed that only the group 
representing current cataloging be searched 
for eligible titles since the yield of unreported 
titles of recent imprint from the other group 
would hardly justify the expense of handling. 
The present average annual receipts by the 
N U C of reports of titles currently cataloged 
by other libraries and of titles cataloged by 
L C but for which cards are not printed is 
about 400,000 cards (plus or minus 15 per 
cent). These would constitute the source 
from which new titles and locations of titles 
would be obtained. « 

The committee did not consider that this 
proposed expansion of the present L C Catalog 
should depend in any way on whether or not 
the N U C is to be published. The base im-
print date of the entries to be included in the 
expansion of the present L C Catalog would, 
however, provide a convenient imprint termi-
nation date for a published N U C . If and 
when the latter is published the scope of the 
current catalog as to imprints included should 
be carefully reconsidered. 

SCOPE 

Physical Form: The present L C Catalog 
includes the following types of materials: 
books, pamphlets, serials, and photocopies of 
the same (Books: Authors and Books: Sub-
jects); motion pictures and film-strips 
(Films); Maps and Atlases, and Music and 
Phonorecords. Reports of holdings of other 

libraries of any of these materials would be 
included.1 

Bibliographical Form: The only biblio-
graphical categories that would be excluded 
r^e telephone directories, talking books for 
the blind and books in raised characters, cur-
rent general daily newspapers, reprints of 
articles appearing in indexed or abstracted 
periodicals and journals (unless cataloged for 
L C printed cards), and fiction issued by the 
standard publishers of the American book 
trade. The committee rejected telephone 
directories as a category generally not re-
tained in collections (except for the latest 
issues). Talking books for the blind and 
books in raised characters are already under 
special and well organized controls in this 
country. Current general daily newspapers 
are generally not cataloged and are easily 
locatable. The content of reprinted articles 
from periodicals and journals is likewise easily 
available if these publications are covered by 
general or specific indexing or abstracting 
services. It was felt that reports of holdings 
of fiction issued by the standard publishers of 
the American book trade could be eliminated 
as a category because: 1. entries for about 
95 per cent of these works would be available 
in the form of L C entries, 2. the receipts 
of reports of holdings of these works is frag-
mentary in the extreme, 3. the works in ques-
tion are normally widely available, and 4. 
as a category, fiction would have a low rank 
in research importance as compared with other 
types of materials. 

The committee considered other categories 
of works which L C would not ordinarily cata-
log but thought it best not to recommend 
excluding them. It reasoned that just as L C 
may waive its general policy and catalog some 
of this material if it is deemed exceptional, 
so it may be assumed that cards for such ma-
terial received from other libraries (which 
generally observe similar policies in catalog-
ing) have been cataloged and reported only 
after a screening has found the material to 
be of some particular consequence. Very 
little such material is currently reported to 
the N U C and it would be of only very minor 
significance so far as workload is concerned. 
In respect to serials and the implications of 
New Serial Titles to this publication, it was 

1 F r o m this point on all statements of qualification f o r 
inclusion will be made on the assumption that other 
qualifications f o r inclusion have been met. 

16 COLLEGE AND RESEARCH LIBRARIES 



felt that it would be premature at this time 
to adopt a policy of excluding serials. 

Language: Entries for works in all lan-
guages using the Cyrillic, Greek and Roman 
alphabets would be included. Entries for 
works in languages using the Cyrillic alphabet 
and for which no L C card has been printed, 
would appear in transliterated form. Entries 
for works in languages using oriental char-
acters or alphabets would be excluded, at 
least until such time as satisfactory solutions 
are found to the personnel, editorial, and 
typographical problems which they present. 

Subject: Although the L C Catalog and the 
N U C receive entries for works on all subjects, 
it is especially to be noted that the principal 
responsibility for acquisitions in the field of 
agriculture and medicine, so far as the na-
tional library complex is concerned, is that of 
the Department of Agriculture Library and 
the Armed Forces Medical Library. These 
two libraries issue the major bibliographical 
publications in these fields, the Bibliography 
of Agriculture and the A F M L Catalog. The 
committee was of the opinion that the most 
effective place to record holdings of American 
libraries in agriculture and medicine would be 
in these publications, in which case they should 
be excluded from an expanded L C Catalog. 
Until such time as the Bibliography of Agri-
culture and the A F M L Catalog are expanded 
to include the titles and locations reported by 
other American libraries, the committee felt 
that holdings of American libraries in all 
subject fields should be recorded in the ex-
panded L C Catalog. 

Imprint Dates: The committee felt that 
the only prospect of an expanded catalog 
which could be self-sustaining or close to 
self-sustaining would lie in confining the titles 
and locations to be added to those which are 
currently being reported to the N U C ; and 
these only if confined to imprints of very 
recent years. 

All reports to the N U C of imprints begin-
ning with 1952 have been segregated from the 
other cards of the N U C supplement and are 
therefore ready, so to speak, for processing for 
inclusion in the proposed expanded catalog. 
Until such time as the N U C may be published 
the committee did not feel that even the cur-
rently received reports of pre-1952 imprints 
should go into the expanded catalog; these 
should more properly be kept for inclusion in 
such a published N U C . It furthermore 

seems to be wise to begin the operations 
involved in an expanded catalog on a scale 
which would not require an excessive increase 
in workload and in the cost of the publication. 
Therefore, the committee recommended that 
the reports of imprints of 1952 and later be 
included in the expanded catalog. 

If 1954 were to be the first year of ex-
pansion of the catalog on the basis outlined 
above (i.e., to include imprints of 1952-54), 
more cards, more new titles, and more ex-
pense would be involved than in any year for 
many years to come, due to the fact that in 
one year a 3-year accumulation of reports 
would be processed and published. It is esti-
mated that in 1954 200,000 locations would 
be added to the present catalog. Of these 
about 97,000 would be new titles. If reports 
were to continue to be received at present 
rates it would be another ten years before the 
volume of locations in any year would equal 
this total (assuming that 1952 were to con-
tinue as a base imprint date) and it is antici-
pated that the number of new titles would 
level off by i960 at around 75,000-80,000. 
However, it must be borne in mind that if 
this project were to be undertaken success-
fully, it is possible that participation in this 
union catalog endeavor might expand con-
siderably and thereby increase the effective-
ness, size, and cost of the catalog. 

T w o alternatives to retaining 1952 as a 
base imprint date year after year would 
be: 

1. to limit the c o v e r a g e to three years by 
picking up each year the current y e a r and 
d r o p p i n g the oldest year. T h i s w o u l d result 
in I955> f ° r example, in 27,000 f e w e r locations 
and 11,000 f e w e r n e w titles than if 1952 i m -
prints had not been d r o p p e d out. Under this 
alternative the annual number of locations 
w o u l d run about 113,000 of which about 54,000 
w o u l d be n e w titles. T h i s is to be c o m p a r e d 
with the estimate of about 200,000 locations 
and about 80,000 titles which could be added 
about the year 1964 if 1952 w e r e retained as 
the base imprint date. T h e cost differential 
w o u l d be roughly p r o p o r t i o n a l ; 

2. to retain 1952 as a base imprint date but 
also to pick up, after publication of the N U C , 
all titles new to the N U C , regardless of im-
print date. T h i s alternative w o u l d , of course, 
result in a considerably larger c a t a l o g in the 
years subsequent to publication of the N U C . 
N o estimate of the degree to w h i c h it w o u l d 
enlarge the catalog is available at the present 
time. 

JANUARY, 1954 17 



EDITORIAL REQUIREMENTS A N D PROCEDURES 

Main and Added Entries: Cards for works 
within the imprint scope of the expanded 
catalog would be sorted out from the reports 
of holdings of other libraries received by the 
N U C and would be sent to the Cumulative 
Catalog Section which is responsible for the 
preparation of the present L C Catalog. They 
would be searched in the author catalog of 
the Card Division at the Library of Congress. 
Those for which L C printed cards are found 
would be recorded as additional holdings on 
the card in the Cumulative Catalog Section's 
file. Those for which no L C printed card is 
found would be typed on a proportional spac-
ing typewriter with suitable small-sized type 
in enough copies for author and subject needs. 
Cross references for necessary added entries 
would also be made at this time. Whenever 
necessary, headings would be adapted to con-
form to A L A rules for choice of entry and 
established L C headings. 

Essential added entries would be made even 
if not traced on the entries as received. Full 
descriptive information as contained on the 
contributing library's card would be given. 
Editing of description would be confined to 
standardizing paragraphing, capitalization, 
punctuation, and abbreviations. Tracing of 
subject headings and of added entries, both 
as edited for the expanded catalog, would be 
given. L C and Dewey Decimal class num-
bers would appear when available. The mini-
mum requirements for inclusion of any entry 
would be author (if any), title, date, and 
location. 

Subject Entries: Insofar as possible all new 
titles would be entered under subject headings 
in Books: Subjects or in the subject indexes 
of the parts for nonbook materials. The de-
gree to which subject coverage would be possi-
ble is largely a function of the willingness and 
ability of the contributing libraries to furnish 
on the cards they send to the N U C the tracing 
of the assigned subject headings. The cards 
presently received by the N U C are found to 
have traced subject headings on only 57 per 
cent of the cases where such subject headings 
would be expected. It is not proposed that 
the editors of the Catalog attempt to subject 
catalog titles without the aid either of as-
signed subject headings or of assigned classi-
fication numbers. In the absence of the 
former, however, an effort would be made to 
convert class numbers into an equivalent 

subject heading. 
Since it is the policy of the present Books: 

Subjects part of the Catalog to include under 
suitable headings all works, even those not 
normally entered under subject headings in 
card catalogs, the cooperation of contributing 
libraries must also be enlisted to supply a 
notation as to literary form, nationality of 
literature, autobiographical nature of the 
work, and the like in the case of titles which 
are not ordinarily assigned subject headings 
and whose titles do not clearly reveal the in-
formation essential to supplying suitable form 
headings for the purposes of the published 
catalog. 

The committee recommended that the gen-
eral editorial policy for subject entries be to 
integrate all such entries as closely as possible 
with the L C subject heading system. L C 
forms of subject headings would be used and 
effort would be made to bring all subject 
headings into conformance with L C policies 
of assignment of subject headings. Unless 
Books: Subjects and the subject indexes for 
the non-book materials are kept within a 
unitary system, it is felt that the utility of 
the subject organization of materials would 
be seriously diminished. Studies indicate that 
in 90 per cent of the cases the subject head-
ings supplied by contributing libraries are 
either in accord with L C form and practice 
as received, or may be converted to agree 
with L C form and practice with little diffi-
culty. 

Preparation of Entries: There appear to 
be only two alternatives as to typewriters that 
will produce cards which will come close to 
matching the printed cards in point size and 
in condensation of type, important considera-
tions in the economics of printing. One of 
these, the Coxhead composing machine, is 
now used in preparing added entry-cross refer-
ences for Books: Authors. It has the ad-
vantage of matching more closely the type of 
the printed cards in type style, in point size, 
and in variety of type face. It has the 
disadvantage of much greater cost and of 
considerably slower operation. The second 
alternative is the I B M 8-point Textype type-
writer. Its advantage is a much lower cost 
and a considerably faster rate of operation. 

LOCATIONS 

Additional locations of titles reported prior 
to the closing of an issue would be indicated 

18 COLLEGE AND RESEARCH LIBRARIES 



on the entries in the Books: Authors part 
and in the parts for nonbook material by the 
standard N U C symbols. In the case of typed 
cards the first symbol would be that of the 
library which either first submitted an entry 
or whose entry was selected as copy. No 
attempt would be made to keep the locations 
in alphabetical sequence. Locations received 
after the first appearance of an entry in the 
catalog would be published only when the 
entry comes up for republication in the next 
stage of the cumulative pattern. The com-
mittee did not feel it could recommend in-
clusion of locations additional to the first one 
on entries in the Books: Subjects part of the 
catalog because of the increase in the search-
ing and typing costs that this procedure would 
involve.2 It was thought that for most pur-
poses it would suffice that additional locations 
of desired titles could be obtained from 
Books: Authors. 

Locations for U.S. federal, state, and local 
government publications were thought to be 
unnecessary. Federal documents are avail-
able at depository libraries, state documents 
at state libraries, and municipal documents 
are generally not collected by libraries except 
the obvious local libraries, the Library of the 
Bureau of the Census and the New York City 
Municipal Reference Library. American doc-
toral dissertations do not require secondary 
locations since they can be obtained from the 
libraries of the institutions granting these de-
grees. Locations of analytical entries for 
periodical articles would not be included inas-
much as the latter are easily located. 

FREQUENCY 

The present Books: Authors section of the 
L C Catalog appears each year in nine monthly 
issues, three quarterly cumulations, and an 
annual cumulation (except that in every fifth 
year a quinquennial cumulation is published 
in lieu of an annual). In considering fre-
quency the problem is to balance the need 
for up-to-date information about titles against 
the cost of repetitive printing in cumulations. 
The committee felt that the best balance 
would be had if new titles reported by outside 
libraries were published in the quarterly and 
larger cumulations. Publication only in an-
nuals seemed to provide an unsatisfactory 
answer to the demand for up-to-date informa-

2 Estimated to be about $9000 in the y e a r 1 9 5 4 . I n the 
following y e a r s it would be less in proportion to the 
smaller number of cards to be processed. 

tion. Semi-annual publication was rejected as 
inefficient. On the other hand, it was ques-
tioned that the need for monthly publication 
was great enough to justify the cost of print-
ing and reprinting. 

Cost: Present estimates indicate that if the 
less expensive I B M Textype equipment is 
used, the Books: Authors section of the Cata-
log could be expanded as described at a sub-
scription price of about $200 per annum. As 
compared with the present catalog the ex-
panded catalog would contain roughly twice 
the number of entries and three times the 
number of locations at about double the 
present cost. It would be a remarkably com-
prehensive list of the current holdings of 
research interest of American libraries and 
would thus be a vital tool for the use of 
scholarship and research; it would have the 
potentiality of simplifying interlibrary loan 
work at the borrowing end (by showing what 
libraries are in possession of desired books) 
and at the lending end (by cutting down on 
the number of requests received for books not 
located in the library), and by making pos-
sible a more even geographical spread of 
interlibrary loans; and it would have still 
greater value than the present catalog as a 
tool for cataloging, reference, and acquisi-
tions purposes by virtue of its greatly ex-
panded coverage. 

It is more difficult to estimate the probable 
subscription price of Books: Subjects due to 
the more uncertain effect of the increase in 
the price on the number of subscriptions. It 
is probable, however, that the subscription 
would have to be at least $250 per annum.3 

It should be emphasized that the expansion 
of the author catalog is not dependent on a 
concurrent expansion of the subject catalog. 
If the subject were not to be expanded, how-
ever, the cost of expanding the author catalog 
would increase slightly, by about 2 per cent. 
This would be due to the fact that certain 
costs which could be otherwise shared by the 
two catalogs would then have to be borne 
entirely by the author catalog. 

This, then, is a plan of what could be done. 
Many of its features are susceptible to change 
according to the needs of the potential sub-
scribers, with or without consequent effect 
on the probable subscription price. More im-
portant than the details of the make-up of 

3 U n l e s s 1 . quarterly issues are discontinued, a n d / o r 
2. Books: Subjects is converted f r o m a catalog to an 
index. 

JANUARY, 1.954 19 



the catalog, however, are these basic questions: 
Is the expansion of the L C Catalog into a 
current catalog of American library resources 
a desirable development? If so, is the time 
ripe for this development? If so, would the 
present subscribers be willing to pay twice 
what they now do in order to get this tool? 
If not, to what extent should subsidy be 
sought? In what direction? With what 
prospects of success? 

It is up to you and your confreres who are 
present or potential subscribers to the Catalog 
to answer these questions. If we are really 
to expand the Catalog in 1954 we need to 

know the answers fairly soon. 
I hope that you will be willing to give this 

matter serious thought and that you will bring 
the proposal to the attention of your local and 
regional library organizations so that later, 
when the time comes for the Library of 
Congress to distribute a prospectus and a 
questionnaire, opinion may have reached a 
state of crystallization. In the meantime, the 
Library of Congress will be glad to receive 
your comments, questions, suggestions and 
the like. If you care to write, I suggest that 
you address your letters to John W . Cronin, 
Director of the Processing Department. 

By C H A R L E S W . D A V I D 

T h e Reproduction of the National Union Catalog 

Dr. David is director, University of Penn-
sylvania Library. 

THERE IS A F A S C I N A T I O N about the con-cept of a vast and all-inclusive bibliog-
raphy of the printed output of the mind of 
man. W e are all familiar with the unsuc-
cessful attempts of some of the earlier bibliog-
raphers, who felt themselves adrift on a vast 
ocean of print, to produce such an index. As 
an example, though by no means the earliest, 
I may cite the Mare Magnum of the learned 
Florentine abbot, Francesco Marucelli, who 
died in 1703, leaving his work incomplete in 
111 manuscript volumes. 

In spite of failures the dream persisted, 
and what was acknowledged to be impossible 
by individual effort was undertaken by co-
operation. Little encouragement is, perhaps, 
to be derived from the experience of the Inter-
national Institute of Bibliography (founded in 
1895 at Brussels) and its Bibliotheca Uni-
versalis. But the Gesamtkatalog der Pruessi-
schen Bibliotheken of the Prussian State 
Library, which got under way in 1898 and 
began to be printed in 1931, has a much more 
encouraging record. Though wrecked by 
World War II, it would in a better world 
have been an undoubted success. Though at 
first legally and practically limited to Prussian 
libraries, it was inspired by the concept of a 
much broader inclusiveness, and in the end 
it was expanded to include all the important 

research libraries of Germany, and even 
Austria was in process of being drawn into it. 
Had disaster been averted we would in fact 
have had a major portion of a manageable 
world bibliography. 

Contemporaneously, American librarianship 
began to press for an expansion of the Union 
Catalog at the Library of Congress which 
would bring it to what was optimistically 
called "completion," and some minds began 
to entertain the thought of its publication in 
this expanded form, so that it could be made 
generally available to research libraries in 
various parts of the country. The key to 
success in such vast operations was of course 
to be found not only in the magic of coopera-
tive effort but also in the technical revolution 
in communication with which we are all at 
least in some degree familiar. 

In 1947 two Philadelphia librarians went 
still further out on a limb and in a brief 
article in the Journal of Documentation out-
lined a plan for a Cumulative World The-
saurus to be printed and widely distributed 
among research libraries all over the world.1 
In the light of the current rise in the cost of 
every library and publishing operation, the 
most charitable judgment which could be 
passed upon that article would seem to be that 
it may have been far in advance of its time. 

1 D a v i d , C . W . , a n d H i r s c h , R u d o l f , " A C u m u l a t i v e 
W o r l d T h e s a u r u s , " in Journal of Documentation, I I I 
( 1 9 4 7 ) , 43-45-

20 COLLEGE AND RESEARCH LIBRARIES