The idea had been in the air for half a century or more. The LC’s Annual Report for 1902 includes a “Bibliography of cooperative cataloguing … (1850–1902),” which cites articles on this subject from all over the world. In 1852, Charles C. Jewett proposed his famous stereotyping plan, by which the Smithsonian would collect cataloging from U.S. libraries and store it in the form of stereotyped plates, which would be made accessible to any requesting library. The plan failed for technical reasons, and because Joseph Henry, Secretary of the Smithsonian and Jewett’s boss, did not agree that this would be part of the proper function of the Smithsonian.³

In 1876, the ALA was founded. According to Putnam, a “main purpose” in its founding was “a centralization of cataloguing work, with a corresponding centralization of bibliographic apparatus.”⁴ At the first meeting of the ALA in 1876, Melvil Dewey, instrumental in the ALA’s founding, proposed that “the preparation of printed titles for the common use of libraries” be discussed, stating, “There somehow seems to be an idea among certain leaders of our craft that such a thing is wholly visionary, at least, their failure to take any practical steps in the matter would seem to indicate such a belief. Now, I believe, after giving this question considerable attention, that it is perfectly practicable.”⁵

Over the next twenty-five years, the ALA tried a number of different ways to put this “visionary” scheme into effect. Attempts to induce publishers to furnish cataloging for their new books failed to gain the support of librarians and publishers for a number of reasons detailed by Scott and Ranz.⁶ Among them are the following:

1. Not all publishers cooperated; many were unwilling to supply free advance copies of publications for cataloging. This delayed receipt of cards.
2. Preparation of quality cataloging would have delayed listings that the book trade needed promptly.
3. Objectives for entries for commercial purposes were bound to differ from the objectives for entries for library purposes (e.g., there were differences of opinion over what was acceptable content for annotations).
4. Publishers were reluctant to support what was perceived of as another commercial enterprise.
5. Schemes required that libraries subscribe to all or none of the cataloging.
6. The number of titles covered was too limited for the larger libraries, but too large for the smaller libraries to justify the expense.
7. Card sizes in libraries had not yet been standardized.
8. Librarians were undoubtedly uncertain about the permanence of the schemes, any one of which would have required “basic and far-reaching changes in their normal cataloguing practices.”⁷
9. Undoubtedly the major factor was the fact that cataloging rules had not yet been standardized.

The second approach tried by the ALA, after various attempts to enlist the publishers failed, was to try to set up a central cataloging bureau under the auspices of the ALA itself. This was established at the Boston Athenaeum in 1896 and operated until the LC began distributing cards in 1901. The number of subscribers was never high, largely because the all-or-none subscription practice mentioned above was maintained.⁸ Undoubtedly, lack of standardization also continued to be a major factor.

In 1877, a year after the founding of the ALA, a third possibility for the solution to this problem was already being suggested by Melvil Dewey: “Is it practicable,” he asked, “for the Library of Congress to catalogue for the whole country?”⁹ In the next paragraph, he points out that the first step in the solution of the problem will be the development of standard cataloging rules. In making these two suggestions, Dewey outlined the two major ways in which the ALA would contribute to the development of the American approach to a national bibliography.

Heisey and Henderson describe the many codes being followed by American libraries in 1900, when it became apparent that the vision of centralized cataloging of which librarians had been dreaming might be realized by the LC.¹⁰ The ALA had approved a code of rules in 1883, but “they were not detailed enough to provide a universal American standard for cataloging,” and they simply became one among many codes in use in the country.¹¹ This might be compared to the situation today in which those seeking to control electronic resources use various metadata schemes. The three leading codes in use were Cutter’s, Dewey’s, and Linderfelt’s.¹² Heisey observed that “it was the practice, as well as the preference of most cataloguers to use several codes, taking what was most advantageous from each.”¹³ In December 1900, the ALA publishing board appointed the Advisory Committee on Cataloging Rules, chaired by J. C. M. Hanson, head of the cataloging department at the LC, and charged the committee with recommending typography and format for the new cards and suggesting changes in the ALA rules to make them suitable for use in the new centralized cataloging project.¹⁴ The LC had already adopted cataloging rules in May 1898; these rules were based on Cutter’s rules.¹⁵ Cutter was one of the members of the ALA committee—thus, as Dunkin pointed out, the new code, published in 1908, “owed much to Cutter.”¹⁶ However, there was a significant difference. Cutter’s statement of “Objects” and “Means” had disappeared, as had his discussions of the rationale behind individual rules. According to Dunkin, “The new code was a set of rules without reasons.”¹⁷

The rules were not published until 1908, largely because of the arrival in 1904 of a request from the Catalogue Rules Committee of the British Library Association that the ALA consider making the new code a joint Anglo-American code. Exchanges by correspondence delayed the publication of the code by several years, but, when published, it represented agreement by the Americans and the British on all but 8 of 174 rules.¹⁸

The new rules were also designed to take into account the practices of the Library of Congress, which, after all, was a large research library. The committee had decided soon after its formation that the plan for the code should be “carried out for the large library of scholarly character, since the small libraries would only gain by full entries, while the large libraries must lose if bibliographical fullness is not given.”¹⁹ Dan Lacy questioned the rationale that full entries are needed for large scholarly libraries, which may originally have been Hanson’s. Hanson was serving as chair of the committee when it made this decision. Lacy felt that the full cataloging called for in the 1908 rules was the result of the ideals of the library movement then burgeoning in the United States:

Cataloging of an elaborate character suited the economy of the American public or college library of the day, straining to make its necessarily limited collection most readily available and most realistically useful to its many readers. But if it suited the economy of the libraries, it no less matched the aspirations of their librarians, in whom were joined an austere zeal in scholarship not unlike that of Browning’s grammarian and an enthusiasm for public service that placed the reader’s convenience far ahead of the cataloger’s toil. These aspirations were wholly shared by Hanson and his colleagues; there is no evidence that they ever questioned whether the Library of Congress might have a different role, whether it might be called upon to acquire and preserve volumes of material whose infrequent use made unnecessary, and whose mass made impossible, the kind of cataloging suitable for a select and actively used collection.²⁰

As indicated above, from the beginning the potential was present for a clash of objectives at the LC between the need to create cataloging suitable to a large research library and the desire to produce cataloging useful in other quite different libraries in the country. Various reviewers of the 1908 and subsequent Anglo-American codes never fail to note where the LC had forced a decision favorable to it and possibly detrimental to public service in other libraries in the country, so it is interesting to contrast their reactions with the following, somewhat plaintive account by Hanson in his 1907 annual report:

The Library of Congress, mainly on account of the distribution of its catalogue cards to other libraries, had been obliged to make a number of concessions in order to bring its own rules into approximate agreement with those of the American Library Association. No doubt these concessions have served to retard its own work and have at times been the cause of some confusion in its records. On the other hand, the fact that the rules now governing its catalogues have been accepted by the two associations which include the great majority of libraries in the United Kingdom and in the United States represents in itself a great advance in cooperation and uniformity of methods, and will have an influence in its future relations to libraries and students, at home and abroad, the importance of which can hardly be overestimated. It is felt, therefore, that the Library has been fully justified in its policy of making liberal changes in rules and practice whenever such changes served to further a general agreement.²¹

The LC adopted the ALA’s List of Subject Headings for Use in Dictionary Catalogs, which had been published in 1895.²² Prior to 1895, many libraries did not have subject catalogs, relying on shelf classification (and reference librarians) to provide subject access to their collections.²³ One of the reasons for the success of the card distribution program may have been that it allowed libraries without subject catalogs to build them quickly and cheaply and thus provide an added public service. One might posit that the fact that the Library of Congress Subject Headings (LCSH) is now such a deeply entrenched standard in this country is because of the card distribution program that brought its subject descriptors into so many libraries.

The LC’s decision to develop a new classification system, rather than using the Dewey Decimal Classification (DDC), which was then in widespread use, was perhaps the most clear-cut instance in which the LC decided to place a higher priority on its own needs as a large research library over the needs of other libraries in the country.²⁴ Although Young had authorized the creation of a new classification scheme, Putnam was very aware of the service the LC could provide other libraries were he to reverse Young’s decision and switch to the DDC. Miksa states that

the chief difficulty in the consideration was the necessity that any scheme adopted be shaped to the particular needs of the collections of the Library itself. If the Dewey Decimal Classification were to be used, many changes would be required in it. But Dewey was unwilling to allow any significant change. He believed that making alterations would be unfair to those libraries already using his system. Thus he required that it be adopted with only minor changes.²⁵

To Putnam’s disappointment, he had to abandon the idea of using the DDC at the LC.

By 1905, even before the standard cataloging rules had been published, Putnam was able to report considerable success in the cataloging distribution program:

The sale of these cards to other libraries began, you will recall, three and one-half years ago. We have not sought to press it for three reasons: (1) Because the distribution involves to the Library of Congress an expense and some inconvenience not at all reimbursed by the subscriptions received; and (2) because the cards at present cover but a fraction of the existing collection, and (3) because our methods and rules of entry are still undergoing revision, and we did not covet the task of explaining changes or of satisfying subscribers as to inconsistencies. We have not, therefore, sought to push the sales. They have, however, increased each year in almost geometric proportion.⁴⁷

Scott detailed several reasons for this success. First, card catalogs were replacing book catalogs at this period, and the card distribution program came along at just the right time to hasten the transition. Second, the LC was able to set up a permanent card distribution staff, which enabled them to allow librarians to order just the cards they wanted rather than require them to subscribe to all the cards as earlier schemes had. Edlund describes in detail how elaborate the card distribution service was eventually to become.⁴⁸ Third, as Scott puts it, “the entries were legitimatized both as emanating from the national library and as conforming to current cataloging practice.”⁴⁹ One suspects a chicken-and-egg situation here in which the standard cataloging practice, which the ALA had been so active in establishing, legitimized the cards, and the cards, when widely adopted, ensured that the national standard was a widely used standard—and thus a more powerful one. Hanson suggested a fourth reason for the popularity of the cards, already alluded to above. As head of cataloging at the LC, he received many letters concerning cataloging, and from these he was “tempted to conclude that a large proportion of the subscribers have been led to adopt the printed cards because they value the suggestions in regard to subjects.”⁵⁰

Edlund suggested several other factors that may have contributed to the success of the program.⁵¹ In 1904, the LC agreed to publish on the ALA’s behalf a new edition of the A.L.A. Catalog.⁵² This was one of Dewey’s pet projects and consisted of cataloging for eight thousand “best books” recommended by the ALA for a small library. The 1904 edition contained LC card numbers for all eight thousand volumes and, in conjunction with the publication of the catalog, the LC offered to sell cards for the entire set for one lump sum.⁵³ The A.L.A. Catalog and the LC cards appeared on the scene in the midst of the Andrew Carnegie period of American libraries. Edlund points out that between the years 1890 and 1917, the Carnegie Foundation gave more than $41 million for the construction of twenty-five hundred libraries in small towns all over the country. He observes, “Often they were part-time libraries, run by part-time personnel, frequently with only a part-time knowledge of the principles and practices of operating a library. To some of these people, ‘catalog’ and ‘cataloging’ were not exactly household words, so they were prime candidates for whatever assistance the Library of Congress could provide.”⁵⁴ Given these circumstances, it is hardly surprising that “the response to the publication of the catalog was of landslide proportions.”⁵⁵

Last, but not least, a major factor in the success of the card distribution program was the comprehensive scope of the collections of the LC, which were continually being increased by copyright deposit. In Putnam’s words, “A collection universal in scope will afford opportunity for bibliographic work not equalled elsewhere.”⁵⁶ Instead of being restricted to current publications of U.S. publishers, as libraries would have been if card distribution through the publishers had been a success, cards were available for all additions to the LC and, as time passed, for all previously cataloged books. In addition, the beginning of the card distribution program coincided with a massive recataloging effort at the LC as the old card catalog was converted to the new printed cards and as the collection was classified using the new classification scheme.

Putnam himself identified what must have presented something of a paradox to those in charge of collection development at the LC when he stated, “To supplement other collections for research your national library must have the unusual book; to enable its cataloging work to be serviceable to other libraries of varying types, it must have the usual book.”⁵⁷ In other words, the LC should collect everything! Putnam even went so far as to suggest that “it would pay this great community, through its central government, to buy a book for the mere purpose of cataloging it and making the catalog entry available in these printed cards, even if the book should then be thrown away.”⁵⁸ In fact, the LC was never able to implement such a collection policy. Charles Harris Hastings, head of the Card Division for thirty-seven years, apparently tried to push for something similar:

The Card Division also desired to have non-copyrighted books purchased on the strength of orders received for cards, instead of waiting for them to be ordered for the reading-room service, or on the recommendations of the chiefs of divisions. The Chief of the Accessions Division, Superintendant of the Reading Room, and other officials maintained that the Library would be flooded with popular books and suffer serious financial loss if the change was made.⁵⁹

In 1902, Hastings announced to the ALA Annual Conference that “the fact is recognized by those having to do with the ordering of books at the Library of Congress that it, being primarily a reference library, can never hope to buy and never ought to buy many books which may properly be bought by public libraries.”⁶⁰

In 1876, the United States was, according to Frederick Leypoldt (editor of Publisher’s Weekly), “almost the only civilized country … not represented by a national bibliography.”⁶¹ Speaking at the Waukesha Conference of the American Library Association in 1901, Dewey said,

You remember that when the Pacific railroad was built, and as the ends came together to make the connection, a great celebration was held through the country, a thrill that the work was at last done; and I feel today, now that we hear in this able report that printed catalog cards are really to be undertaken at the National Library, that what we have waited for over 20 years, and what we have been dreaming about has come to pass at last.⁶²

The solution to the problem of creating a national bibliography seems peculiarly American, and Dewey’s comparison with the mechanical and technological triumph in Ogden, Utah, singularly appropriate. Putnam, too, saw the triumph as being mechanical in nature. He wrote, “American instinct and habit revolt against multiplication of brain effort and outlay where a multiplication of results can be achieved by machinery.”⁶³ In a sense, the LC cards were interchangeable parts for libraries. Standardization made it possible for the smallest library in the country to have the same quality of cataloging as the largest research library. In this, the card distribution program was profoundly democratic. Every American citizen who used a public library could benefit from the expertise that went into creating the national bibliography at the LC.

Every silver lining has a cloud, however. The card distribution program marked the end of an era when “librarian” meant a person who both cataloged and administered a library, and thus was an incomparable guide for the user through his or her library. With cataloging centralized at the LC, the fears of librarians such as Frederic Vinton came to pass, to some extent: “We fear that the so much desiderated object of co-operative cataloguing (by which each librarian shall have the least possible writing to do) is unfavorable to good librarianship. For myself, I would on no account lose that familiarity with the subjects and even the places of my books which results from having catalogued and located every one.”⁶⁴ Henderson pointed out that the creation of the ALA Advisory Committee on Cataloging Rules to create the 1908 code led to the separation and isolation of catalogers from administrators and stated that “before 1900, cataloging was a concern of all of the ALA’s members, since the issues were discussed in general meetings.”⁶⁵ According to Bishop, “Classification and cataloging occupied the major part of the curriculum in the early years of training in library science. They were definite matters which could be taught, and they were controverted subjects which awakened intense partisonship.”⁶⁶ Today cataloging is practiced mainly at the LC and by a tiny corps of librarians primarily located in large research libraries. Most librarians learn little about cataloging in graduate school and go on to administer libraries, teach children to read, and provide reference service to the public without bothering to learn how to use their own catalogs properly and without bothering to follow cataloging issues or comment on them. When the LC recently considered abandoning the systematic cataloging of trade publications to focus on digitizing their backlogs of rare and unique materials, few librarians other than catalogers took notice.⁶⁷ The loss of cataloging expertise on the part of most librarians resulting from the efficiencies achieved by means of a greater division of labor was probably inevitable. The change would surely have come about eventually under the crush of the information explosion of the twentieth century, but surely there is no harm in lamenting with Cutter the passing of a “golden age,” especially now, when the very existence of human intervention for information organization is under constant threat while most of the library profession has little understanding of the danger the loss of it would pose for their existence as a profession.⁶⁸

Doing the research to write this paper prompted this author to ponder the changing cataloging landscape. The final section of this paper explores the current scene and the possible future of shared cataloging, asking the following questions: Are the same forces operating today as were operating at the turn of the last century? Are they operating in the same way or in different ways? At the turn of the century, the United States was prosperous, powerful, and in an expansionist mode. It was the era of the Progressives, who argued that the business of government was to advance the health and welfare of its people; technologies new at that time were harnessed to serve these goals. Some might argue that, at present, the LC serves a government that is dominated by those who wish to shrink all aspects of government that are not part of the military industrial complex. While funding available for the LC’s technical services remains the same, a change in the internal priorities of the LC now directs more of those funds to digitization projects and much less to cataloging. Apparently, cataloging is now seen as a part-time activity to be done by staff who are also responsible for acquisitions tasks, including electronic resource license negotiation.

In addition, the LC is now situated in an information universe in which more pervasive technologies have come more and more to set their own agendas. The presence of Google on the scene seems to be an indication that there are businesspeople who think that there might be money to be made by competing with libraries in the provision of information to the public, and Google’s popularity seems to indicate that for many ordinary people convenience takes priority over precision, recall, and even accuracy when it comes to information access. It seems possible that the future customers of libraries will no longer be the public at large, but only that small elite consisting of people who do serious research, and in a democratic society it is hard to get funding to support the work of a small elite, even one as important as this one to our future progress and prosperity.

To this author, it appears that the ALA is now dominated by library administrators with shrinking budgets who know very little about the complexities of bibliographic control (other than its expense) and who wonder if the fact that undergraduates are in love with Google might not provide an excuse for libraries to dispense with the information-organization part of their budget entirely.

The publishing industry may still be reluctant to invest in the creation of standardized and detailed cataloging (or metadata), just as it was in the nineteenth century, judging by the fact that Online Information Exchange (ONIX) is still not widely implemented and by the fact that descriptions in Amazon.com are so rudimentary that it is not possible to distinguish one edition from another, or even to find all of the editions of a given work if the author’s name or title varies.⁶⁹ The publishing industry and other content providers also appear to be actively involved in shrinking the commons by extending copyright limits and by more jealously protecting their intellectual property rights, making it more difficult and expensive for libraries, archives, and museums to provide communities with online access to their digital holdings through cataloging records. The old partnership between libraries and publishers in all formats—in which libraries served to popularize published works by making them available to more people and created more customers for publishers by encouraging higher literacy rates—may be breaking down now that publishers have other ways of reaching potential customers directly. Most publishers are essentially for-profit organizations and, as such, probably care little about the fact that those who cannot pay their high fees will no longer have access. In this context, it is interesting to look back at Melvil Dewey’s argument that the government should promote the interests of libraries over those of publishers because libraries deliver more education and civilization to the public for less money than would be the case if publishers alone were responsible.⁷⁰ One suspects that, in the current era, our government no longer places such a high value on educating its citizens that it would decrease the profits of publishers in the way it was willing to in Dewey’s day.

OCLC has largely replaced the LC’s card distribution program as the mechanism by which LC cataloging is shared with the nation’s libraries. If the LC were eventually to abandon the cataloging of trade publications, the question arises as to whether the great research libraries and the remaining public libraries would follow the LC in abandoning cataloging. Would OCLC continue to be viable without LC copy? And without the LC at the center, would cataloging continue to be done in a standard and sharable way? Already, many would argue that the cataloging of audiovisual materials found in OCLC shows less standardization than that of monographs largely because of the lack of a supply of LC cataloging copy for audiovisual materials.

This author has written elsewhere of her fear that the rise of the Internet may threaten the profession of librarianship and the value it places on access to the cultural record for all—regardless of socioeconomic level—in order to ensure an informed citizenry.⁷¹ However, the Internet is a tool that can be used either foolishly or wisely. It also has the potential to allow cooperative cataloging to thrive in a much more efficient fashion in the future. Currently, cataloging practice is very repetitive. Every time a new edition of a work is published, a cataloging record for that new edition is created that repeats much of the information already found in the cataloging records for all the other editions of that work. Newer conceptual models of cataloging, such as Functional Requirements for Bibliographic Records (FRBR), Functional Requirements for Authority Data (FRAD), and Functional Requirements for Subject Authority Records (FRSAR), as well as the related model underlying Resources Description and Access (RDA), are based on the hope that libraries, archives, and museums may be able to raise cooperative cataloging to a new level of efficiency by using the possibly emerging Semantic Web to share in the creation of entity records (or the record equivalent in the Semantic Web, the uniform resource identifier or URI) for works, authors, subjects, places, and the like.⁷² It is even possible that we could share the work of entity description with people who are not librarians, catalogers, archivists, or museum scientists, such as subject experts, bibliographers, and the like. While we would still need to ensure that only people willing to learn how to practice accurate entity identification and how to choose commonly known names for entities as preferred forms should be allowed to have editing privileges, we could collect suggestions for variant forms not yet linked to preferred forms or for corrections to our entity definitions from anyone who took an interest, and we could encourage everyone in the world to link to our entity definitions when citing an author, work, subject, or class; it should be a lot easier for a nonlibrarian to link to the appropriate URI than to have to use the correct string of text, as is currently the case. If these entity records performed the same searching function as our authority records currently do, allowing a user to search for a particular entity using any extant variant of the name of that entity in any language, these more efficiently created catalogs could also perform better than ever before.

It remains to be seen whether we will use our new tools foolishly, to create a new “dark ages” in which much of the cultural record is either lost or hidden from view, or wisely, to advance the welfare of humanity and create a world in which all of its people, regardless of socioeconomic level, enjoy and make use of humanity’s entire cultural record. What is at issue are the goals we wish to achieve as a society and whether we will direct our current technologies to serve those goals or rather abandon those goals in favor of allowing the technologies to set their own agendas. The economic, political, and social factors that predominate in our current society at the turn of the millennium will determine our choice in the same way that our choices were determined in 1900. As always, it is up to us to choose the kind of society we want.