Even though the control file’s ultimate purpose is “control” of the materials identified as being appropriate to the edition’s base of sources, editors use the database well before they send out the first request to an owner repository. The control file’s first function is making possible an efficient and comprehensive collection of the edition’s raw materials. The mere need for a particular documentary edition sometimes means that comparatively little is known about the source materials on which that edition will be based. If all variant versions of the correspondence or papers of a given writer, group, or organization were in one convenient location, it would be unnecessary to fund a project to locate and catalog them. If an archival collection were already adequately arranged and described in a published finding aid, there would be no need to appoint an editor to assume the burden of creating one. No authoritative edition is possible without a thorough canvass of source materials that may deserve a place in it, and no documentary edition is any better than the collection of materials on which it is based. Long before scholars begin to edit, they become tracers of lost documents and masters of archival management.

Few editors escape becoming experts in some aspects of the collection, care, retrieval, and cataloging of documentary materials. In all likelihood, they will be called upon to create an entirely new archive, which may include copies, original manuscripts, microfilm reels, and digital images that together can be called “the Papers of X,” “the Records of Y,” or “the Writings of Z.” It is the editor’s responsibility to make sure that this archive is as complete as possible and that intelligent cataloging procedures enable the project’s staff to find what they need. If they are truly diligent, they might receive a tribute like the one paid by Debra Newman to Robert Hill and his cadre of Marcus Garvey’s editors: “While conducting research for a guide to records for black history among the civilian records in the National Archives, wherever I found Garvey records, there was evidence that Hill’s staff had already seen them” (review of The Marcus Garvey Papers).

Most editors must first identify the location of documents from a variety of sources. This phase of editing may be the most physically and intellectually taxing, but the rewards of discovery and victory over the vagaries of time and chance, the accidents of war and natural disasters, and the carelessness of heirs are often the sweetest that editors taste. The work begins with a survey of what is already known about the body of documents to be gathered. The editor begins as a conventional enumerative bibliographer for the individual, organization, or group that is the subject of his or her edition. The basic reference tools for the study of American history and literature will come into play: the appropriate sections of the monumental Guide to Reference Works, initiated under the editorship of Constance Winchell and continued by Eugene Sheehy and Robert Balay, and the Harvard Guide to American History, rev. ed. (1974).

A. The Editor as Enumerative Bibliographer

Modern editors now have access to such effort-saving tools as the cumulative electronic editions of abstracting and indexing services (e.g., America: History and Life and the MLA International Bibliography) for searching monographic literature since the early 1950s. If you’re lucky enough to have access to a library subscribing to JSTOR, you can review the full text of hundreds of scholarly journals going back to their first issues. Unfortunately, most other electronic tools are not completely retrospective: for publications of earlier decades, the bound volumes of the printed equivalents of the series must be searched as well. Once the background bibliography is completed and the works listed have been located, footnotes and back-of-book bibliographies are combed for citations to original materials. The editor should also enlist the cooperation of scholars who have shown an interest in the project’s subject, for even if these men and women have published the findings of their earlier research, their notes on sources and their experience in canvassing the same repositories that the editor may visit will be of invaluable assistance. The results are a file that notes the date, description, and location of every document that earlier scholars have unearthed.

The control file database will be used to create the lists of the known pertinent holdings of given repositories. Notes of likely sources are entered in the database without the accession number that indicates that a document actually exists as part of the office’s collection, but even at this stage, entries in the field labeled “location” should be based on the system to be used throughout the project’s life. Traditionally, these were the location symbols for repositories around the world now known as the MARC Code List for Organizations and conveniently available online at the Library of Congress’s Web site. Use of these coded symbols is less necessary now than in the precomputer age of editing. Before the availability of electronic tools, this standardized system avoided the possibility of duplicating symbols, and it was easily mastered by the editorial staff. Many modern editors still use the codes for initial data entry, expanding repository names to their full form in an automated database. Others create a master list of repository names and import them into database fields.

Editors still need to be familiar with the principles of the code, since many of the reference sources they consult use the old abbreviations. The MARC codes usually consist of three elements: (1) an uppercase letter designating the repository’s state or nation, (2) one or two letters pinpointing the state or city, and (3) a code for the repository’s name. Thus “N” stands for the State of New York; “Bu” represents the city of Buffalo; and the addition of “Hi” for historical society designates the Buffalo and Erie County Historical Society. For a list of standard MARC codes currently in use, see http://www.loc.gov/marc/organizations/. Should a project need to devise symbols for libraries absent from the MARC code list, the same principles can apply.

Editors should also check earlier editions of the correspondence, journals, or personal or organizational papers to be edited and index their contents for use in the collection phase of the new edition. Although such printed versions may be flawed, some will be the only ones that survive. These citations, too, are entered in the edition’s database, following the same pattern used for entries recording located documents. When earlier printed texts prove to be the only ones available for materials whose originals have been lost, these records will also remain in the permanent file with their own accession numbers. When printed texts are matched to the originals that were their sources, the bibliographic information for the printed version is simply added to the “notes” field in the original’s permanent record. Don’t be tempted to discard such information: it may be necessary to refer to printed texts for manuscripts that have suffered physical damage or fading. Beyond their value as records of lost or now incomplete originals, these earlier editions also provide clues to patterns of correspondence: the identities of correspondents as well as geographical and chronological patterns of an individual’s letter writing. Back runs of periodicals often provide printed versions of documentary material, and an editor should give special attention to those journals that customarily publish special sections of printed documents or routinely include portions of the collections of a historical society or other special library.

Journals that consistently publish articles and documents related to the editor’s subject are also excellent locations for notices of the inauguration of the editorial project. Unfortunately, editors of many scholarly journals have adopted strict policies on accepting such advertisements, but editors should make every attempt to advertise the project as early as possible. In the past, it was customary to place notices in journals and in such newspapers as the New York Times and the Washington Post. Today, Web sites like H-Net and other Internet lists that provide forums for specialized audiences are perhaps even more important for advertising the existence of a new project and its needs.

B. Defining the Scope of the Search

The editor’s preliminary files will indicate the names of the individuals and organizations whose personal papers or institutional archives are likely to contain other materials useful for the project as well as those collections known to include specific sources of interest. Next comes compilation of lists of certain and probable repositories of documents for the new editorial archive that the project will create.

1. Manuscripts

The resources for a “manuscript search” in repositories around the world are in a process of productive explosion triggered by the use of the Internet. We know that a good deal of the advice we offer here will be out-of-date by the time you read it, but we’ll do our best to provide a good start. For American manuscript material, you’ll have to consult several guides, some online and some on paper. Each has special limitations and individual virtues, and the editor may have to consult all of them.

a. Online Sources

ArchivesUSA (http://archives.chadwyck.com/) is a Web resource that is part of the publications of ProQuest. It’s fee based, and not all academic libraries subscribe. It’s invaluable, though, and editors who don’t have access to it themselves are well advised to enlist the assistance of a friend or colleague who does. It provides easy access to the complete National Union Catalog of Manuscript Collections (NUCMC), which began periodic publication by the Library of Congress in 1962. The volumes were based on data furnished by each manuscript repository, and their thoroughness and accuracy vary greatly. In the 1980s it became clear that a substantial number of the records provided to the National Union Catalog merely duplicated records that were being provided to the online OCLC and RLIN services. It was decided to end print publication of new NUCMC records with volume 29, and all NUCMC volumes are currently out of print. ArchivesUSA provides easy searching of all NUCMC records, which represent more than 103,000 collections.

Beyond NUCMC records, ArchivesUSA provides access to NIDS (National Inventory of Documentary Sources in the United States) finding aids for over 400 repositories (over 54,000 collections) and collection descriptions submitted directly to ArchivesUSA. In addition to these search capabilities, ArchivesUSA provides up-to-date information on addresses, phone numbers, and Web sites for libraries and manuscript repositories. This part of the service is based on DAMRUS, Directory of Archives and Manuscript Repositories in the United States. There are also more than 5,000 links to online finding aids.

OCLC (Online Computer Library Center) is known to most researchers through its WorldCat service, available at most academic libraries. This database of online library catalogs contains most catalog records for manuscript collections generated since the demise of NUCMC. The version of WorldCat now available online for free (http://www.worldcat.org/) does not have the sophisticated search capabilities of the fee-based WorldCat used by academic librarians. It’s worthwhile to find a cooperative librarian to search the database for your project.

UNESCO Archives Portal is designed to provide information for archivists and archive users: http://www.unesco.org/webworld/portal_archives/.

Another useful Internet source is the RLG (Research Libraries Group) ArchiveGrid. This is an important tool for searching through documents, personal papers, and family histories held in archives around the world. Thousands of repositories have contributed nearly a million collection descriptions to ArchiveGrid, which contains collection descriptions as well as linked finding aids: http://www.archivegrid.org/web/index.jsp.

b. Print Sources

The online aids listed above duplicate many—but not quite all—of what you’ll find in these books:

The “Hamer Guide,” Philip M. Hamer’s A Guide to Archives and Manuscripts in the United States, appeared in 1961 as part of the NHPC program. It lists 1,300 repositories. Descriptive entries for each source vary, but entries are usually generous enough to make clear the major focus of each archive and to list its major collections by name. When printed catalogs for a repository or one of its collections exist, these are noted.

“NHPRC Directory” is the commission’s Directory of Archives and Manuscript Repositories in the United States (2d ed., 1988). Although it covers thousands of repositories, its descriptions of individual collections are spare. The entries’ contact information for libraries and archives and statements of policies for access and photocopying are long out-of-date. A further limitation is the Directory’s policy of citing only manuscript catalogs and calendars published after the appearance of the Hamer Guide without repeating information from that earlier commission publication. Still, the entries in the Directory do give generous cross-references to fuller descriptions of institutional holdings that may appear in the Hamer Guide, NUCMC, American Literary Manuscripts, and other printed sources.

The revised edition of the Harvard Guide to American History, edited by Frank Freidel, appeared in 1974. You may be surprised to find us citing a paper source that’s more than thirty years old, but you should still check this guide’s important sections on primary sources under the major headings “Materials of History,” “Printed Public Documents,” “Unpublished Primary Sources,” “Microform Materials,” and “Printed Historical Works,” all in part 1 of volume 1. See also “Collections of Original Sources” (1:278–79).

J. Albert Robbins’s American Literary Manuscripts: A Checklist of Holdings in Academic, Historical, and Public Libraries, Museums, and Authors’ Homes in the United States (1977) reports the holdings of only 600 repositories, but these are described in great detail. The entries are based on canvasses of the holdings of manuscripts written by 2,750 men and women deemed of special significance in the history of American literature. The entry for each individual includes Library of Congress codes for institutions holding his or her manuscripts, the size of these collections, and notes of any existing finding aids. Sadly, the ALM has never been updated.

Few significant American figures confined their correspondence or their professional activities to the United States, and their heirs did not limit the sale of family papers to American collectors. Editors must be equally cosmopolitan in searching for surviving papers. The best printed guide to published catalogs and other finding aids for materials in foreign repositories is the excellent section “Foreign Archives of Interest to American Historians” in the Harvard Guide to American History. It remains the most logical and comprehensive survey of such literature, although it hasn’t been updated since the 1970s. To supplement the Harvard Guide, editors should consult recent issues of American Archivist, as well as pertinent sections in the Winchell-Sheehy-Balay Guide to Reference Works. Beyond these paper-based sources, modern researchers have the luxury of consulting online guides to major repositories in other nations, with many of these Internet resources available both in English and in the language of the nation where the institution is located. See, for example, the Web site for the Archivo General de Indias in Seville: http://www.mcu.es/archivos/visitas/indias/indias.html.

c. Dealers and Collectors

Unless an edition’s subject is a recently deceased and obscure figure, the editor will not be the first to show an interest in collecting relevant documents. Tracing the course of materials that have passed through the hands of dealers and collectors of rare books and manuscripts is one of the most time-consuming and frustrating tasks an editor must confront. Editors should gauge carefully the time allotted to searching for the records of such transactions and accept the fact that they will never unravel every mystery concerning such sales. Records of sales of documents fall into two categories: auction catalogs and dealers’ listings. Catalogs list only items to be offered for sale to bidders at public auctions. Before the advent of the Internet, such catalogs received a wider distribution than the periodic issues of listings prepared by individual dealers and circulated among special libraries and dealers’ longtime customers. Dealers’ listings generally quoted the prices asked for individual items, and sales were concluded by written or telephone contact between the purchaser and the dealer. Most major auction houses and dealers now offer their wares on the Internet as well as in paper-based catalogs or listings. Editors can create a search list for most of these from professional directories like those of PADA (Professional Autograph Dealers Association), ABAA (Antiquarian Booksellers Association of America), and ILAB (International League of Antiquarian Booksellers), or even the advertisements in Manuscripts magazine. For earlier decades, when these listings appeared only in type or print on paper, different strategies must still be used for the two categories of sellers.

The wider circulation of printed auction catalogs ensured that they are more likely to have survived. Consulting such catalogs is simplified by the existence of American Book-Prices Current (ABPC), which has provided an indexed, comprehensive listing of manuscripts and rare books sold at auction in the United States since 1895. ABPC listings after 1956 also cover materials sold at London auction houses, and the guide now covers the auction sales in Stargardt, Germany, British auctions outside London, and major Canadian and Australian auctions. Occasional sales in Switzerland, Germany, and France are also included. The last print edition of ABPC, volume 110, appeared in 2006. In the future, the source—at present including all records since 1974—will be available only online at subscribing libraries. For pre-1975 listings, researchers must use the earlier volumes and their cumulative indexes.

The ABPC is only a starting point for the editor. Its descriptions of items cannot reproduce completely the information offered in the original auction catalogs, nor does it reproduce any photographic facsimiles that may have accompanied the original auction listing. Using its references, the editor should obtain the auction catalogs from which the ABPC’s notices were drawn. Here George L. McKay’s American Book Auction Catalogues, 1713–1934: A Union List (New York, 1938) is an invaluable finding aid.

Although post-1956 ABPC listings include London auction sales, the editor whose subject maintained a considerable correspondence abroad or whose papers are popular among European collectors should weigh the need for a search of foreign dealers’ records. Unfortunately, the autograph and rare book trades of the Continent have felt no need to create regular national listings comparable to the American ABPC or the analogous British series, Book-Prices Current and Book Auction Prices (the latter are now sadly defunct). A notable exception is Germany’s Jahrbuch der Auktionpreise. Some American auction houses maintain in-house files of European catalogs, and the Grolier Club of New York City and the British Library in London boast generous holdings of these records. Before venturing outside the records of the American market, the editor should consult a dealer who specializes in manuscripts and books relating to the project’s focus. Expert advice can save months of wasted effort, particularly if the advice is “Don’t bother. It won’t be worth your trouble.”

Printed dealers’ listings, because of their irregular publication and limited circulation, have a higher mortality rate than auction catalogs. Editors who can identify dealers who consistently specialized in materials of interest to their work may obtain access to their back files of lists and offerings. Failing that, any repository with a large collection of the papers or writings of a prominent figure will be on the mailing lists of dealers who frequently offer these materials for sale and usually keep these circulars on file, available to scholars. Similarly, an individual collector with an interest in the editor’s subject will have a place on those same mailing lists, and he or she may be able to provide access to such listings.

The editor whose manuscript materials fall in the period 1763–1815 can consult Helen Cripe and Diane Campbell’s American Manuscripts, 1763–1815: An Index to Documents Described in Auction Records and Dealers’ Catalogs. This monumental work indexed entries in auction catalogs before 1895 (the year of ABPC’s inauguration) as well as dealers’ listings through 1970. Entries are indexed by date, name of correspondent or author, and name of dealer. Even editors whose projects fall outside the 1763–1815 period should consult American Manuscripts: Part 3, which contains the only extant published inventory of dealers’ listings (with repositories owning these items); its surveys of periodicals in the autograph- and book-collecting trades are invaluable. Like ABPC, American Manuscripts is only a starting point, providing nothing more than the dates and titles of manuscript materials. The bibliographical information offered in the volumes merely leads to the rare and scattered copies of dealers’ lists and catalogs, where full descriptive entries appear.

In planning a search, the editors enter the text of copies of notices from auction catalogs and dealers’ listings in the project’s database, just as they would for earlier printed versions of documents. These records, too, can reveal patterns of correspondence of which earlier researchers were completely unaware, and they will ensure that the editor pursues all fruitful leads in consulting guides to manuscript repositories and in making on-site searches of collections. Once the search is completed, the editor will often find that these listings are the only surviving evidence that X wrote to Y on a given day, and notices that offer précis, lengthy extracts, or photographic facsimiles of these “lost” documents will be an important addition to the edition’s files.

d. The Internet and Manuscripts

Beyond the more traditional methods for identifying manuscript repositories with materials for an edition, modern editors test their skills at retrieving information on the Internet. Aside from the searches we’ve suggested through fee-based and public databases, online dealers’ catalogs and auction catalogs, editors should try the basic Web search, using their favorite search engines. Before you enter your quest in something like Google.com, though, think through your strategy. If the subject of your edition is as well known as Eleanor Roosevelt, you’ll need to refine your search terms to something like “Eleanor Roosevelt letters.” If you’re looking for the surviving correspondence of a more obscure figure, then search under that person’s name (within quotation marks). You’ll be pleasantly surprised at the results. Many local historical societies and smaller specialized libraries haven’t added records of their collections to WorldCat. A reference to a collection on their Web site may be the only public announcement of its existence.

In the early 1990s, Daniel Pitti made his mark in the archival world by designing the electronic format known as the EAD (Encoded Archival Description). This gave archivists and manuscript curators around the world the opportunity to convert existing finding aids (usually typed descriptions of collections not included in card catalogs) to a form that could be quickly and easily mounted on the Web. The very existence of many of these finding aids was often a well-kept secret, and they were, of course, available only at their owner repositories. Overnight, they became accessible to scholars around the world for careful review and analysis without a visit to their home libraries. One curator at Yale is said to have remarked happily, “We began to get photocopy requests for materials we didn’t realize we owned!” Whenever your preliminary research for your edition indicates that a given institution owns material in your field of interest, go to that institution’s Web site to see what online finding aids are available. If you have difficulty finding such information, e-mail the appropriate librarian or curator and ask if there is a finding aid for collection X, Y, or Z, and whether it is available online. Even if the tool hasn’t yet been mounted on the Web, the curator may reward your initiative by xeroxing and faxing it to you.

Finally, never expect 100 percent success. With all the tools we’ve recommended, with all the online finding aids at your disposal, there will be some manuscripts that you don’t find. Editors of documents from earlier centuries can never retrieve records that have centuries to scatter. Editors of more recent vintage will never locate the heirs of every correspondent of their subject or persuade all those heirs to give permission to publish manuscripts that they own.

2. Printed Documents

Most individuals significant enough to warrant a documentary edition of their writings have written works for publication for editors to collect and evaluate. At this early stage, documentary editors can confine themselves to the basic methods and terminology of enumerative bibliography and the preparation of reliable and comprehensive listings of printed works and their various editions. Obviously, the editor of the papers of any public figure or organization will compile a bibliography of the printed works produced by that author or group. The editor’s preliminary inventory of “known” writings will form a skeletal enumerative bibliography. To expand this list, the editor will probably need to consult a wide variety of reference works and published catalogs that apply to the period and the field in which the subject was prominent. Perhaps the most convenient source for guidance to such reference tools is G. Thomas Tanselle’s Guide to the Study of United States Imprints (2 vols.). Although Tanselle’s volumes long preceded the age of the Internet and World Wide Web, they remain an invaluable aid for navigating older printed bibliographies.

For the eighteenth century, as an example, scholars routinely consult Charles Evans’s American Bibliography: A Chronological Dictionary of All Books, Pamphlets and Periodical Publications Printed in the United States of America from the Genesis of Printing in 1639 Down to and Including the Year 1820 (13 vols.), which serves as a chronological dictionary of imprints through 1800, as well as Roger P. Bristol’s Supplement to Charles Evans’ American Bibliography. Both “Evans” and the Bristol supplement are now available online at many libraries as the “Evans Digital Edition” or “Early American Imprints, Series I: Evans 1639–1800,” and the convenience of searching in this form can hardly be overstated. Students of American bibliography of the early nineteenth century consult American Bibliography: A Preliminary Checklist for 1801–1819, compiled by Ralph R. Shaw and Richard H. Shoemaker (19 vols.). This resource is being mounted online as Early American Imprints, Series II, Shaw-Shoemaker (1801–1819). The series was continued for the 1820s and 1830s as A Checklist of American Imprints, compiled by Shoemaker and other scholars. To date, this segment is available only in “paper” form. Holdings listed in these printed guides should be supplemented by reference to RLIN and OCLC.

Often an editorial project needs to buy, not merely to locate, a specific book or pamphlet. This is incomparably easier nowadays thanks to online services such as BookFinder.com that link multiple booksellers’ databases.

Bibliography for the editor does not stop with an enumerative listing. Catalogs of rare books and pamphlets use the vocabulary of descriptive bibliography, which pinpoints the physical differences among various editions, impressions, states, or issues of the same work. Editors should master this terminology to determine whether they have located all the pertinent variants for such printed documents. For the novice, the best introduction will be John Carter’s ABC for Book Collectors. Try to get the seventh edition, with corrections, additions, and an excellent introduction by Nicolas Barker. Once you’ve mastered that, you may want to look at the more sophisticated standard scholarly works in this field, such as Fredson Bowers, Principles of Bibliographical Description; Philip Gaskell, A New Introduction to Bibliography; and chapter 4 of David C. Greetham, Textual Scholarship: An Introduction.

Planning a search of periodical literature is exceptionally challenging. While most magazines publish a cumulative index—or at least a table of contents—newspapers present a greater problem. For an author’s contributions to newspapers, earlier biographies and the writer’s correspondence will offer clues for a search for pertinent issues. The editor whose study focuses on the pre-1820 period has the great advantage of access to Clarence Brigham’s History and Bibliography of American Newspapers, 1690–1820 (2 vols., 1947), plus his Additions and Corrections (1961). For the later period, students can consult the work of the U.S. Newspaper Program, whose lists of volume-specific newspaper holdings for more than 77,000 libraries are now part of the OCLC database as well as being available on microfiche. The Harvard Guide and pertinent sections in Tanselle’s U.S. Imprints offer the best discussions of finding aids for the precomputer era. In some regions, surviving Works Progress Administration guides supplement these sources, and checklists of newspapers in microform will allow the editor to identify periodicals that can be obtained in microfilm or microfiche on interlibrary loan. All of these sources should be supplemented by reference to the most current volumes of A Guide to Reference Works and its periodic supplements.

Readex, the publisher of Digital Evans and Early American Imprints online now offers three series of “Early American Newspapers,” running from 1690 to 1922, which can be searched through one interface. While all these Readex products offer extraordinary convenience for searching, they are also very expensive, and comparatively few libraries subscribe. Here, as with other fee-based online tools, the editor may have to enlist the assistance of friends or colleagues with access to these research wonders. Somewhat more generally available is American Periodical Series (APS) Online 1741–1900 (ProQuest), which contains 1,000 magazines published between 1741 and 1900. “Making of America,” a full-text archive of nineteenth-century books and journals at the University of Michigan (http://www.hti.umich.edu/m/moajrnl/browse.html) current-ly has a fairly limited number of periodicals online, but new items are constantly being added, and it’s well worth your time to check the Web site for relevant additions.

In searching full-text versions of magazine and newspaper runs for your subject’s contributions, familiarize yourself with the search options each project offers. Many don’t provide detailed tagging that identifies instances when a given personal name appears as the author of a piece rather than its subject. For these searches, the editor must find some form of the subject’s name likely to be associated with authorship rather than a mere “mention.” The Andrew Jackson Papers, for instance, searched the online Early American Newspapers for “Sincerely Andrew Jackson” rather than merely “Andrew Jackson” to pinpoint letters Old Hickory had submitted for press publication. Similar closings for letters or the phrase “By So and So” will serve other editors.

No two editors follow the same line of attack, but all will find it useful to consult colleagues who have completed a search similar to theirs. In searching for printed documents, as for manuscripts, the law of diminishing returns applies. The editor should first make contact with the repositories most likely to have the needed materials. Only after these returns have come can the editor decide whether the first canvass was fruitful enough to justify a second and wider appeal.

C. The Canvass for Source Materials

When the results of these preliminary inventories of primary and secondary sources and unpublished and online finding aids have been entered in a database, the information can be sorted into patterns that serve the editor’s purposes. A report arranged by repository (and, within repository, by collection, if known) gives a convenient list of items that should be ordered in image form from specific libraries or archives. A report arranged by correspondents can open a new range of searching if certain individuals appear frequently on this list. After consulting both budget and timetable, the editor can begin the process of locating the original materials and, where appropriate, of obtaining hard copies or image files for the project’s archives. Some institutions contain so much material that only a personal visit from editors or other staff members will ensure a complete search, but a good part of the canvass will probably be conducted by mail. For convenience, we’ll include both traditional and e-mail in our discussion.

The scope of the mail canvass will be dictated by several factors: the era in which the project’s figure lived, the nature of the subject’s prominence, and the history and geographical distribution of the personal or organizational records involved. At one end of the spectrum are the papers of men or women who (1) maintained meticulous files of their own papers, (2) spent a relatively obscure life unlikely to make their manuscripts attractive to autograph dealers, and (3) died within the previous decade, before their fame could spread and their papers could be dispersed. Here it can be assumed that such a figure’s personal archive represents copies of all or nearly all the materials to be located and that there has not been time or reason for recipients’ copies of letters to be scattered or destroyed. The search can be confined largely to the subject’s correspondents (or their heirs) and to the files of the businesses, educational institutions, or government agencies in which that figure performed his or her professional work.

There’s a correspondingly greater challenge for the editor of the papers of a more conventionally famous individual, whose personal papers are likely to have been dispersed during the decades or centuries after his or her death. Here the editor’s contacts can’t be limited to repositories where there are known to be collections of papers likely to produce materials for the project. A “blind search” will be necessary, one that extends to institutions where there is no reason to suspect that such documents survive. Again, editors respect the law of diminishing returns in conducting such canvasses. Letters of inquiry go first to libraries known to own materials pertinent to the project. Here the editor knows that images or photocopies must be generated, bills should be paid, and questions should be answered. It makes sense to begin this chain of correspondence as soon as possible. It will also take longer for librarians at institutions that have a dozen manuscripts to locate and photocopy or scan those items than for librarians whose collections contain no pertinent manuscripts to respond to a blind search letter with a polite note indicating that they can be of no help. Requests to foreign repositories should also be included in this category. Clearly, it will take longer to obtain photocopies from abroad than from a neighboring state, and even digital imaging may be more time-consuming as curators and scholars in different nations (quite possibly using different languages) may need a bit more time to clarify and understand each other’s needs and procedures. Not even existing projects are spared this labor. In the course of a long-running project such as one of the Founding Fathers, it is inevitable that repositories will purchase new documents subsequent to the first search. A second mail canvass will then be necessary. The Jefferson Papers Retirement Series, for instance, conducted a second search after its creation in 1999. The original search for Jefferson materials had been conducted fifty years earlier, and this second review produced dozens of “new” Jefferson documents for the years 1809 to 1826.

Editors of the records of formal groups or organizations must be even more resourceful. They must identify the officers and other leaders of such bodies, whose names frequently appeared as signatories on official correspondence. Many of the organizations’ letters and other outgoing communications that have made their way to academic libraries or other repositories will be indexed under the names of individual “authors,” not the organization involved. Editors must naturally search collections of the papers of such individuals as likely sources of materials for the organizational archive as well as establish contact with local chapters or branches of the organization.

1. Phrasing the Search Letter

Whether an editor seeks the records of an individual or a group, the form letter used for the blind search will be the basis for the entire mail canvass. (A sample of such letters appears in the appendix.) It should include (1) a description of the project’s scope and an indication of its scholarly credentials in terms of sponsorship or endorsement or a statement of the qualifications of the individual editor, (2) a clear statement of the types of materials needed by the project, (3) a precise statement of any special requirements for images or hard copies, and (4) clear proof of the editor’s businesslike knowledge of library procedures.

The last might consist of including a stamped self-addressed envelope for the addressee’s convenience or providing return e-mail address, postal mail address, and telephone and fax numbers in the “signature” of an e-mail message. When the form letter is modified for an institution where materials are known to exist, the editor should not only list the materials in question but also name the sources that indicate that these documents are part of the repository’s collections. It is also wise to make clear that the editor’s interest is not confined to the list offered, since additional acquisitions or improved cataloging may have enlarged the list of the repository’s holdings.

It’s also wise to indicate just why you think that each repository may own materials of interest to you. If you don’t have references to specific documents in their holdings, mention the collections that sound promising to you or the collecting interest of the library or archive that might have led the institution to acquire relevant materials. When the letter is a truly “blind” inquiry, a vague phrase like “knowing the broad interests of your institution” can’t hurt.

If your edition has a subject that involves the modern era, it’s never too early to ask about literary rights and copyright restrictions. Most libraries automatically inform you of their general policies, but some may be careless about informing you of special limits on the use of specific collections or documents. You’ll display your credentials as a knowledgeable researcher when you indicate in your first inquiry that you know you may need help from the donors or heirs of donors of materials at the repository in order to obtain appropriate permissions to reproduce or publish the documents. Should your later research lead you to a copyright holder of whom the repository is ignorant, be sure to let the librarian or curator know.

Editors must weigh the advantages of using traditional (i.e., postal) mail or e-mail for their searches. Most libraries and manuscript repositories have some kind of Web site that provides e-mail addresses for appropriate staff members, and this information is usually up-to-date and reliable. While most repositories are happy to take e-mail inquiries, a few may prefer the formality of postal mail or still lack Internet services. It’s easy enough to e-mail or telephone a curator or librarian in advance with a polite question about such preferences before sending out the full-scale search inquiry. Remember, too, that you’ll make a better impression on addressees if you call them by name rather than opening your inquiry with “Dear Librarian” or “To whom it may concern.”

The reasons for these elements of a project’s mail appeals are self-evident. A request from an editorial project is only one of dozens of time-consuming inquiries a librarian or curator will receive each week or even day. The letter should convince its recipients that the project is a worthwhile endeavor whose request merits the time and trouble it will demand. A specific description of the required materials spares curators unnecessary work and ensures that the editor receives all the needed information and documents from an individual repository. Include any details you have at hand, even shelf marks (if known) for boxes or folders. A list of known materials assists the librarian in locating items in the same collections that may not be cataloged. A statement of the editor’s realistic attitude toward library procedures assures correspondents that the letter comes from an experienced professional, not an inexperienced or amateur researcher.

Before drafting search letters, editors must come to some difficult decisions about just what standards their project’s images require. Some projects seek negative microfilms that can conveniently produce positive paper prints or digital scans for their files, while many libraries prefer to keep the negative of any microfilm they make as a security or record copy. In such a case, the library may be willing to furnish a direct-image duplicate negative. When a project calls for a facsimile edition, the degree of resolution and the reduction ratio in any films or digital images made for the project should meet appropriate, consistent standards. Most microfilm cameras and scanners can be adjusted, but instructions and resettings should be determined in advance. The editor with facsimile publication in mind should also attempt to obtain as many of the project’s hard copy prints as possible in positive rather than negative form, since these will reproduce more clearly in the next generation of duplication.

The modern option of obtaining digital images complicates further the form of copies that you request, and it is a problem for all projects, not merely those planning an electronic publication of image files. Most repositories are happy to provide digital images of unbound manuscripts, since they can retain duplicates for their own files, retiring the original manuscripts from service. Thus your request for scanned images fits nicely into a library’s own preservation program. However, you must make clear your minimum requirements for the scanned images to suit your edition. The library’s routine processes may not suit your needs. The Papers of Abraham Lincoln project, for instance, plans an online publication of high-definition color images of Lincoln’s correspondence and other manuscripts. Project staff visit most repositories holding a large body of Lincoln material in order to create digital images themselves, since many libraries don’t own cameras or scanners that could generate images that meet the project’s standards. While most editions won’t face a challenge of these dimensions, they remain responsible for making their needs known clearly and succinctly when requesting digital images, xeroxed copies, or microfilms.

2. Scope of the Search

Once the text of the basic search letter is determined, editors must decide just who will receive one of the inquiries. Word-processing equipment eases the burdens of a mail canvass, for much of this software was originally designed for business offices, where such form letters are routinely required. The preliminary control file’s findings on institutions with pertinent holdings will provide the list of repositories that will become the basis for addresses for a word processor’s mail merge operations. For blind search letters, the word processor can reproduce the body of the letter (for postal or e-mail transmission) as many times as necessary, and most can import the appropriate inside addresses from a database. For those libraries known to have material for the project, the editor simply adds a list of specific items.

The scope of the blind search depends not only on the project’s budget and schedule but also on the prominence of the figure or organization central to the project’s work. For an exceptionally celebrated figure, that search may extend to every institution listed in the current editions of the American Library Directory and Directory of Special Libraries and Information Centers that admits to having a manuscript or special collections division, as well as to all historical societies in the United States. For lesser-known subjects, the blind search might be confined to institutions whose collecting focus or geographical location makes it likely that their holdings include items of interest to the project. In general, it is better to cast the collecting net too widely than too narrowly. A veteran of a decades-long search for the Documentary History of the Ratification of the Constitution remarked, “A vacuum cleaner is more efficient than a broom and a dustpan.”

3. On-Site Searches

The editor’s preliminary research inevitably produces a list of repositories whose holdings require a personal search. Responses to the project’s mail canvass often reveal that the resources of an institution are so limited that its staff cannot even search for a half-dozen known items. Others will not have the equipment needed to provide good quality images but are willing to allow editors to use their own cameras or scanners to capture those images. The wise editor first turns for advice to other scholars who have visited the institution in question, learning from their personal experience of a given library’s procedures.

Beyond this, no personal search should be undertaken without first confirming that a visit is convenient for the institution and its staff. Budgetary considerations, local holidays, and even the opening of a new exhibition can force a site to modify its days or hours. Collections may be temporarily unavailable or housed in a separate facility with appointment-only access. Aside from such practical considerations, documentary editors must remember that their searches place far greater demands on a library than those of more conventional researchers. The editor who arrives unannounced at a historical society to canvass three dozen collections will receive a cold welcome. And any scholar who plans to bring special equipment, such as a laptop computer, digital scanner, or personal photocopier, must ask in advance whether a library’s wiring or rules on the handling of rare materials permit their use. Most major research libraries routinely accommodate such modern tools of the scholarly trade, but smaller institutions may not.

The effective editor appears promptly and as well prepared as possible. Upon arrival, the editor asks whether the library has unpublished finding aids to the collections to be consulted. (Any published aids have, of course, been consulted in advance.) Editors cannot assume that understanding the workings of one institution’s catalog means that they have mastered the peculiarities of catalogs found elsewhere. Direct questions about any peculiarities in filing or access methods help the editor find everything there is to find. The definition of a cataloged or an indexed manuscript collection, for instance, can differ from one repository to another. Some archivists may consider a group of manuscripts to be completely indexed if there is a file giving alphabetical access to the names of authors of pieces in the collection. In that case, the editor will need an alphabetically arranged enumeration of the subject’s most constant correspondents. Working from this list, the editor can consult index files, alphabetically arranged “miscellaneous” folders, and collections of autographs for letters written by these men and women to or about the subject.

Since most editors conduct searches to obtain images of some kind, whether photocopies, scanned images, or microfilm, the editor should ask the curator in advance if there is a preferred method of marking or flagging items for reproduction before beginning work. If the appropriate marker, usually a piece of labeled acid-free paper, can be placed immediately with the appropriate document, the original materials will be spared unnecessary handling, and the editor will not have to go back through the collections to insert these markers. Editors who bring their own cameras or scanners will ask the curator for specific rules for handling original materials before capturing a single image.

A running list of materials located during the on-site search should be kept whether or not markers are placed. Ideally, when an editor visits a library, her or his laptop computer will contain records from the project’s database for documents known to exist at the repository. When a document already has an entry in the database, that record is modified to indicate that it has been located and that it has been scanned by the editor or that arrangements have been made to order an image. If the editor has Internet access to the “home” database, the updated information can be transmitted immediately. If this is impractical, the updated records in the laptop can be loaded into the database when the editor returns home. If laptops can’t be used during a site visit, the information will have to be recorded manually and added to the project’s files as soon as possible. Using a laptop or pen and paper, an editor creates a list of items of interest to the edition at every institution visited. This list notes each item’s location (including collection name and any box, volume, or folder number) and pertinent information about the document’s physical appearance. It should also indicate whether it is an original or a photocopy, a recipient’s copy or an authorial draft of a letter, a fair copy or a preliminary authorial draft of a document. The list also notes the document’s length and details, such as address leaves or endorsements. This will serve as the basis of the reproduction order itself, and it will allow the editor to check the completed order. Since it may be necessary to consult the original manuscript or printed materials again, the detailed list will make it easy to retrieve the items a second or third time. Of course, the editor with special requirements for document reproduction should explain these to the staff of the library or archive before placing an order.

The editor should accept as immutable law any special policies a library has for photoreproduction or scanning. Some libraries will not permit an editor—or any other patron—to use a personal copier or digital scanner to obtain images. Others demand the eventual return of copies of their materials because some researchers have donated such photocopies to other research centers without the permission of the owner institution. Other restrictions on photocopying or scanning bound volumes and fragile manuscripts exist for the protection of these source materials. Curators of rare material are usually only too happy to explain the reasons for these and other rules.

Finally, never assume that a librarian or archivist will remember what your project has done at his or her repository. When you need to do follow-up inquiries, be sure to let the library know what documents you’ve already found and what collections you’ve combed earlier. There’s turnover in the staff of libraries, archives, and museums just as there is in editorial projects. Earlier correspondence or orders for copies or digital images can be misfiled. One project ruefully supplies a cautionary tale from the early 1990s. The editors learned of a new collection acquired by a library where their searches had located 750 letters, all copied for the project and processed five years earlier. The editor who wrote to the library about this new resource neglected to mention the earlier search and those 750 photocopied documents. A conscientious archival assistant not only searched the collection in which the editor expressed interest but went on to send copies of the 750 letters the edition had known about for five years.

a. A Word on Research Etiquette

Curators and librarians, like documentary and textual editors, respond favorably to kind and courteous treatment and intelligent questions. The librarian who answers a blind-search letter or e-mail message deserves a personal note of thanks. The curator who provides images or extends hospitality to on-site searches deserves even more praise and gratitude. Editors who make themselves memorable for good manners will be kept in mind when an institution acquires a new collection or catalogs an existing one and discovers unsuspected riches for their projects. A long-term project may have to conduct several canvasses, and the editor who needs to visit a repository more than once has special need for a warm cooperative welcome. Should the project need to hire searchers on a freelance basis to canvass distant collections, these deputies should be drilled in the project’s practical requirements and in the standards of proper scholarly behavior. If the editor is new to the area of scholarly research, John C. Larsen’s edition of the Researcher’s Guide to Archives and Regional History Sources provides a good introduction to professional vocabulary and procedural traditions.

Luckily, as libraries and other research institutions mount image “editions” of more and more of their own collections on the Web, archivists and academic librarians have been reminded of the intimate connection between the history of editing and archival management. Schools of library and information science now offer courses that include introductions to editorial methods (see Elizabeth Dow, “Archivists and Scholarly Editing”). It’s much easier than it once was to find enthusiastic partners among librarians and archivists as more and more of these professionals recognize that editors are their colleagues in making reliable versions of documentary materials accessible as widely as possible.

4. Private Collectors and the Marketers of Rare Materials

Dealers in rare books and manuscripts and collectors in this field often demand far more persuasion. Prudent editors exploit the results of their canvass of sale records for manuscripts and printed materials beyond planning the project’s search and supplementing the document file. Reviewing these records may reveal a few dealers who specialize in the project’s subject. At the least, consideration shown to these dealers can ensure the editor a place on their mailing lists for future offerings. The fortunate will find among such dealers a friend to offer advice and even introductions to private collectors and other dealers who’ll lead the editor to more original materials. Some editors have established practical quid pro quo relationships with dealers and auction houses. The editors supply valuable historical and contextual information as well as verify authenticity of manuscripts for dealers, while dealers and auction house experts repay the favor by notifying editors promptly when materials of interest to the scholars come on the market. If the project’s subject has been a popular one for collectors, it will pay the editor to pursue this tactic. A dealer or auction house may be willing to open its old records so that the editor can consult inventories of collections sold decades ago with no published catalog. Dealers will sometimes disclose the identities of institutional purchasers (though not of private buyers) of specific items. In some cases, the editor may find it worthwhile to consult books about patterns of collecting in his or her field. Working from published catalogs of auctions and from dealers’ listings, an editor may be able to guess which collectors (or their agents) were likely to have purchased certain items; the editor can then begin discreet and tactful approaches to those collectors or their heirs.

Rare book and autograph collectors frequently choose not to advertise the existence, much less the extent or nature, of their collections because they have no wish to be troubled by the inquiries of scholars or visits from thieves. Dealers who wish to stay in business are conscious of their clients’ concern for privacy. Dealers promise confidentiality to those individuals who purchase rare materials, and they will not and should not violate that promise to assist any researcher. When an individual collector shares the contents of a collection with an editorial project, that individual’s wishes should be honored to the letter. When a collector directs that his or her name not be given as the owner of any documents used in the edition, then the source line for these texts should read “in a private collection.” If there is public access to project files, correspondence with private collectors should be kept where security can be assured. Access to many private manuscript collections has been eased by the Manuscript Society Database, which provides access to hundreds of private collections owned by society members.

Some collectors and dealers voice concern that print publication of a manuscript’s text will lessen its market value. The editor can often persuade them otherwise by pointing to what Katharine Kyes Leab calls the “imprimatur value” that original material receives by being chosen as the source for all or part of a definitive edition. Happily, there is one group of collectors unlikely to be concerned about devaluing their collections by authorizing publication: specialists in the collection of stamps and other materials relating to postal history. Frequently, items purchased because of their philatelic interest (envelopes and address leaves) are accompanied by the letter covered by that significantly franked or postmarked material. Not only are postal collectors more willing than autograph specialists to share their collections with editors, but they are also potentially valuable consultants. Their expertise in the mechanics of the transmittal of correspondence can save an editor countless errors in dating materials and in assessing the significance of details of postal markings.

Journals directed to collectors should be supplied with notices announcing a project’s creation, but the editor need not wait for responses to these advertisements before establishing contact with dealers. During an edition’s life, its staff must continue to review current auction catalogs and dealers’ listings. Major auction houses now mount their catalogs on their Web sites—including illustrations that may include good images of manuscript material. When such notices appear for an auction house in the vicinity of the edition’s offices or in the neighborhood of a friendly editorial colleague, a member of the staff or other deputy can usually view the document while it is on public display before the auction. Although auction houses frown on the use of cameras during these viewing periods, few object to an editor’s taking notes or even discreetly making a handwritten transcript of a letter.

Even the editor who obtains a transcript of a manuscript being offered for sale prefers a copy of the original for the edition’s files, and when it is impossible to have a transcription made, it’s even more important that the editor persuade the manuscript’s purchaser to provide such a copy. Since dealers and auction houses will not disclose a private purchaser’s identity, this means a new pair of form letters: one addressed to the dealer and the second directed “To whom it may concern” and accompanied by a stamped, self-addressed envelope, to be forwarded to the item’s purchaser. (Samples of such letters appear in the appendix.)

When collectors of modern manuscripts fail to respond to such appeals, the editor should not curse them for insensitivity. Collectors of manuscripts of recent date are a sophisticated lot who understand the niceties of copyright law. Although they can claim physical ownership of the inscribed paper of interest to the editor, they hold no literary rights to the text embodied in that document. A responsible collector cannot authorize further dissemination of the document without the permission of the author or the author’s heirs or any others with copyright claims.

II. Cataloging and Control

Even while the search continues, photocopied documents or scanned images will arrive via e-mail or the U.S. Postal Service, and the editor must master the art of cataloging and control of documents. The most comprehensive search is wasted if the editor or a project’s staff can’t retrieve individual items or groups of items they have collected. The editorial staff should also guard against cataloging documents that do not fall within the project’s scope or miscataloging pertinent items. The process of authenticating documents by handwriting, signatures, and literary style begins now. To these methods of identification the editor should add a sense of logic and historical context, assigning dates or authors to undated or unsigned materials. As Lester Cappon reminded editors, “Authenticity of the document is the cardinal rule, axiomatic by the very nature of historical method” (“A Rationale for Historical Editing,” 57). Even more basically, an editor can never assume that a document is authentic because it comes from a well-known repository: at even the oldest and best-known libraries, there are dark archival corners that have never been reviewed carefully by modern curators.

In rare cases, the editors are not only the publishers of texts of original source materials but also the curators of these documents during the project’s life. Some large-scale documentary projects, such as the Adams Papers at the Massachusetts Historical Society, the Benjamin Franklin Papers at Yale, and the Thomas A. Edison Papers, have borne responsibility for files that include both photocopies gathered throughout the world and original manuscript collections held at their sponsoring institutions. It’s more usual, however, for editors to serve as manuscript or rare book curators when their editions are archival ones, limited to a specific group of manuscripts or printed works owned by the institution that sponsors the project. Editors who are not archivists or rare book librarians by training must learn the rudiments of these professions so that the unique materials entrusted to their care do not suffer. Even editors who do not have this responsibility are wise to learn the tricks of these trades, for that knowledge will assist them in understanding the methods of the repositories to be searched and in devising a system of cataloging the new editorial archives created by the edition’s research.

Should an original collection be uncataloged before editorial work begins, the curator of the institution’s other holdings should be consulted to make certain that the methods used to arrange the editorial collection are compatible with those used for other materials at the repository. Editors who must plan the arrangement and description of these materials should consult the Society of American Archivists’ online Standards for Archival Description Handbook (http://www.archivists.org/catalog/stds99/index.html). This Web site provides links to online and print resources outlining approved methods for cataloging various kinds of manuscript, photographic, and artifactual sources in formats accepted by the professional archival community. Be warned that standard archival methods may not meet all the needs of a nonarchival editorial project. Editors who deal with variant versions of the same documents from several sources must often invent systems for identifying the source of each copy or original.

A. Accessioning Materials

The first step in establishing physical and intellectual control over an edition’s documents file is to assign each document its identifying number, usually called the accession number, which frequently indicates the order in which the document or its copy was added to the project’s files. For many projects, assignment of these numbers begins while the editors are still planning a search for materials: identifying numbers are given to every document to which a reference is found in the secondary literature or other sources. Projects that need to maintain separate files for discrete groups of materials often use two or more sequential series. One series may contain the numbers 01, 02, 03, and so on; the second can begin with 001, 002, 003; and the third, with 0001, 0002, 0003. For photocopies, these accession numbers are traditionally stamped or written by hand on the back of the reproduction. If a document is represented by more than one page of photocopy, sequential page numbers are a convenience, and the leaves of the copy might be labeled 017-1, 017-2, and so on. Of course, no such stamps or written numbers should be imposed on original materials; identifying numbers on the documents’ individual folders must suffice. If all or part of the edition’s files of documents are stored as digital images, with no corresponding “hard copy” prints, then the accession number is the image file’s identifying number with numerical extensions indicating additional pages.

When the project’s collection includes such materials as letterbooks, notebooks, ledgers, or other formats containing dozens of items of varying dates and titles in one physical whole, the best initial method is to assign the same accession number to the entire volume or leaflet and to keep the copied leaves or digital images together rather than to separate individual items. The chronological and alphabetical sections of the project’s control file will furnish access to individual entries, and the integrity of the original will be preserved. Once editorial work begins, patterns that divide these volumes into distinct and independent sections often become apparent. The editor may then wish to add numerical extensions to the original accession number (10001-1, 10001-2, etc.) so that each discrete section can receive separate cataloging in terms of date, authorship, or topic.

The staff member responsible for accessioning materials must also check copies or scans for completeness. If the copies result from an on-site search, the searcher’s descriptive list is the basis for comparison. Should the copy arrive as the result of a mail canvass, the cataloger must check closely for possible omissions—margins cut off by careless photocopying or the absence of endorsement or address leaves whose existence can be deduced from “show-through” on another page of the copy. Any questions concerning such imperfections should be raised immediately in a letter or e-mail to the curator of the original material.

B. Labeling and Arranging Folders

Editors who maintain purely digital image files of their source materials have the luxury of designing and redesigning the virtual arrangement of their documents at will. Other editors, however, need to house some paper-based copies or microforms (with accession numbers) of these source materials on shelves or in file drawers, and the original physical labels and arrangement of the materials will be difficult to reorganize. Doing it the right the first time is the goal here.

Even in the twenty-first century, editorial projects are likely to mimic the methods developed more than half a century ago for labeling and filing these hard copies. Documents or their facsimiles are usually assigned control numbers before cataloging information is entered into the control file. Once an item has been assigned its identifying number, a file folder is prepared with a label (now often computer generated) containing the following information: accession number, date, and a brief version of the document’s title. This label is placed where it can be easily seen when the editor flips through folders on shelves or in drawers. Thus the physical arrangement of the information on the label in this manual system must reflect some organizing principles that can be reported by the control file database. Whatever element of information determines the document’s location on the shelf appears at the top of the label. For collections focusing on correspondence, this is usually the date. For collections of technical material arranged topically, a key word in the document title might be the cue for filing. When several items exist for the same date or topic, a system of subclassification will be necessary. For correspondence, the most convenient subheading is usually the name of the correspondent who sent or received each letter to or from the project’s subject; for subclassifications within a topic, either the document’s date or the name of its author suffices. Even the physical shelving of photocopies can spark debate. At one project, there was a lively argument between right-handed and left-handed staff members, because it is easier for the first group to see labels placed in the top right-hand corner of a folder. The southpaws, of course, preferred folders with the same information placed on the folder’s top left-hand corner. The project preparing the image edition of Papers of the War Department, 1784–1800, used tab-end folders with color-coded numbers representing each manuscript’s date so that any chronological misfiling was easily evident.

Computerized databases can modify the sequence of steps in processing an accession to the edition’s files. For instance, many projects find it more convenient to enter an item’s cataloging information in the control database as a first step in accessioning. The database assigns the accession number and generates an appropriate label to be printed out and attached to the folder that will contain the item, and even makes additional labels to attach to the versos of photocopies in lieu of stamped numbers. While the notion of a “master” file is irrelevant with computerized methods, the database must be able to generate a report that will serve as a shelf list, one in which the sequence of records matches the physical arrangement of the folders.

As noted above, many projects find it more convenient to use laptop computers to accession documents “on the fly,” entering records with accession numbers at the libraries where the search is being conducted. These records will include a special field to indicate that the document photocopy or digital image has not yet been added to the project’s “home” files, and that flag will be removed once processing is completed on receipt of photocopies or images. The advantage of this method is obvious, since each document becomes part of the overall database record for purposes of analysis and study as early as possible.

Computerized systems offer a special bonus to enterprising editors who plan a microform or other facsimile edition. These editors deliberately create an accession number field large enough to accommodate not merely the sequential identifying number assigned each document but also the number of characters that will be needed to identify the item’s final location in the microform (reel and frame numbers). This allows the editor or the microfilm technician to substitute the microform-location information for the accession number in the document record, facilitating the final preparation of the index. Thus, if the editorial archive were expected to contain no more than 1,000 items, but the editor projected publication of twenty microfilm reels of 1,200 frames each, the accession number field would be set for seven characters, not the four needed to identify 1,000 pieces. When microfilming is completed, accession number “0000022” can be changed to “21/1002” to indicate that the document’s image appears on frame 1002 of reel 21. The database can then be sorted alphabetically, topically, or chronologically to create index entries carrying accurate references to the new microform.

Before the first document folder is filed, there must be guidelines for chronological arrangement of materials at the project. Most editors follow the principles of archivists and file undated material to which it is impossible to assign even a month or year at the very end of the collection’s chronological series. There is less agreement, however, on the arrangement of partially dated material and documents that bear inclusive dates. Some projects file documents that bear only a fragmentary date (a year or a month of the year) at the end of the period recorded in that date. All materials dated simply 1771 would follow items dated 31 December 1771; all materials dated March 1771 would follow items dated 31 March 1771. Other projects adopt precisely the opposite course, placing partially dated materials at the beginning of the period into which they fall: here materials dated 1771 would precede those dated 1 January 1771. Whichever method the editor adopts, it must be followed scrupulously. The same general principle applies to documents with inclusive dates. Some projects file such items at the first date that appears: a ledger sheet with entries for 1 January 1771 through 2 March 1775 would be placed at the beginning of the year 1771. Conversely, others would file the item at the end of materials for 2 March 1775. The decision here may reflect the nature of the document. If the range of dates results from a document containing a clear range of dates (a letterbook or a ledger or account book), then the earliest date is a logical label, since it marks the beginning of the document’s creation. If, instead, the range of dates reflects the period within which a document is known to have been composed, as with the draft of public papers or literary materials, then the last date, the point of final composition and perhaps even publication, is a more logical filing point.

The editor should choose the method that best serves the individual archive. Whatever method is selected, it should be made in advance, recorded for staff reference, and stuck with through thick and thin. Editors can set up a computerized database to allow for range dates so that the first and last dates are reflected, but the editors will still have to know where hard-copy photocopies are filed. Even during the initial processing of materials, the editorial staff will be able to supply some details of authorship and dating that do not appear on the document itself. When handwriting identification enables a cataloger to identify as John Smith the correspondent who signed himself “John,” the full information should appear on the folder. And if the cataloger, through the use of a perpetual calendar or the application of common sense, can establish that a note dated “Wednesday, February 3” was written in 1808, then the full date should be placed on the folder. Traditionally, information supplied by a member of the editorial staff appeared within square brackets in these records, for example, “John [Smith]” and “3 February [1808].” Remember that these square brackets can perplex a computer’s sorting mechanism. You may need to create an “inferred” field next to the date field and use a “yes/no” toggle between the two when an editorially supplied date is entered.

C. The Expanding Functions of Control Files

The control file begins to grow while the editor plans the project’s search, and it will shape the edition’s form and be reconfigured by the edition’s progress. Control files undergo constant revision and improvement. An examination of newly accessioned manuscripts and photocopies allows an editor to assign dates to undated materials, to correct the dates on misdated items, and to determine the author of an unsigned letter or the intended recipient of one whose envelope or address leaf has been lost. For any date or title that is corrected or supplemented, the item’s control file record should provide a cross reference from the date, author, or recipient once erroneously attributed to the item to a record with the accurate date or title. This prevents confusion over just which items from a given repository have been cataloged so that two orders are not placed for the photocopy of a manuscript that was cataloged under two dates or descriptions. If the editor prefers, he or she can record this redating information in a field for repository notes. Whenever editorially supplied information is added to a control file, the corrected letters, words, or numerals should also be entered in the same format as used on the label on the document file folder. And, of course, the folders of hard copies must be rearranged to mirror their new cataloging.

D. Making the Best Use of Twenty-First-Century Ways and Means

Arrangement of data in control files no longer poses the physical and intellectual challenge to documentary editors that it did even twenty years ago. While twenty-first-century editors should never ignore the hard-won wisdom that comes from the experience of earlier scholars in the field, they must always remember that the technology available today makes much of that wisdom obsolete. An added challenge for the modern editor is readiness to take advantage of what computer hardware and software provide. At a very basic level, it’s far easier now to enter and manage detailed information about the location of the originals of documents in a project’s files. For manuscript collections, numbers for boxes, volumes, or file folders can be recorded—and those numbers can be sorted easily at a later date. The same holds true for locations in multi-reel microfilm sources. For print materials, a modern control file can even record a book or pamphlet’s call number for easy recall when an owner library may need to be contacted.

Frequently, a project’s limited funds or limited space rules out the creation of hard copy from microfilms. Here, location information should carry not only the code for the owner repository but also the full title and reel number of the film where the editorial staff can consult the text of the original. If the film’s frames are numbered, references to the frames should appear in the control file. The editor whose archive will include a substantial number of filmed reels should establish a separate system of control for these spools. Each can be assigned an identifying number, and this can be included in the location information in the database control file. If this is cumbersome, the films themselves can be arranged in a rational and consistent order that will make their retrieval easy: this might involve shelving the reels in the order of the MARC location symbols for the sources of the films’ contents. All of these issues are far more easily addressed with automated methods, and the editor who anticipates these problems can include appropriate fields in the control file record in advance.

1. Subject Indexing

While every edition’s control file demands fields for dates, authors, physical descriptions of the original, and repository locations, a computerized database may tempt editors to add subject headings to this list of control file essentials. Facsimile publication projects may need to take special care in designing fields for subject entries, since the database will be the foundation of the finding aid or search engine that accompanies the facsimile edition, whether it be microform or digital images. With no publication of word-searchable texts or contextual annotation, users will need all the help they can get from a basic subject index.

Even rudimentary subject indexing adds precious minutes to creating a document record. For a small, short-term project, the ability to do some imaginative searching of the control file for proper names may provide enough subject access for the edition. For a small edition of letters, reliance on basic search capability might make it wise to include in the control file fields for any and all place-names provided by the author in a dateline or in a letter’s address. Even at this basic level, subject indexing demands that the editorial staff maintain a list of “approved” versions of proper nouns and subject headings to be entered in the “subject” field. In indexing parlance, such a list is a “thesaurus,” and designation of standard entries is an exercise in “controlled vocabulary.”

More elaborate subject indexing demands even more scrupulous planning. Not every topic or every proper name should be indexed; veterans in this field advise establishing a firm limit to the number of subject entries that will be made for a given document. The initial computerized index done at the time of cataloging will be only a rough one to guide the staff in later editorial work. In the words of the Joseph Henry editors, “The object is not to produce a definitive index at this pre-editorial stage. The purpose is to establish a rough content control, in addition to traditional controls.” Most editors find it prudent to place a limit, albeit a generous one, on the number of fields for subject entries at this initial stage. As the process of collection and analysis progresses, an editor can design a more comprehensive index and implement it electronically.

As much as possible, though, it’s better to anticipate the needs of an index rather than to improvise during the process of cataloging. Since most scholars come to editing with some knowledge of the subject of the edition, they’re well qualified to predict the subject headings most likely to be used in the index and to design a thesaurus or list of terms to appear there. Thought should also be given to the forms in which proper names are entered. Here a running authority list can easily be created. Editors should remember that they will not be the first to consider subject access to the information in their disciplines: librarians and archivists have anticipated them in most areas and created and published standard sources for thesaurus construction and name-authority lists. Examples of these are the Library of Congress Subject Headings and the Art and Architecture Thesaurus, both conveniently available online.

Once the subject index’s design and authority lists have been established, the cataloger merely adds a step to the entry of basic information such as date, document title, and details of provenance. Some projects find it necessary to investigate relational databases at this point. Editors of large collections that will receive intensive indexing, such as the Edison archive, design systems of codes for subject headings and personal names to expedite data entry. This system of codes may in turn demand a separate file linked to the control file. Once the limits of indexing have been established, the cataloger—or a senior staff member responsible for subject indexing—can enter codes for themes and individuals whose importance warrants a subject entry or can retrieve the full terms themselves from the project’s thesaurus and authority files.

Beginning editors should not be daunted by our outline of possible subject-indexing demands. A subject-oriented content-access system can be initiated during the cataloging process itself or later in the project, before the selection of materials for publication. Many editors find it best to postpone subject indexing or eliminate it altogether. Early scheduling of an index provides an initial subject guide to the project’s collection, but implementing this system will slow the process of cataloging somewhat. Even this small delay may be unacceptable for a project that must catalog as quickly as possible. When little is known about the patterns of an individual’s correspondence, for example, it is imperative that materials be cataloged with the greatest haste so that planning of the search can progress. For most projects, however, a rough content control file can be created simultaneously with cataloging control. If your project needs to plan the initial search based on subjects or if your project will need to retrieve documents later based on subjects, then you should consider topical indexing at an early stage.

III. Conclusion

With every passing year, documentary and textual editors find new ways to use computer-based technology and the Internet to perform traditional functions like accessioning document files and to indulge in luxuries like subject indexing and full-text searching. Still, the work of collecting materials and creating control files remains anything but a neat linear progression. Although editors may need to assure sponsors and funding agencies that their work moves along in a steady, easily defined fashion, they make no such pretense among themselves. While initiating a project, an editor periodically reviews the results of work to date before moving on to the next task, and the results of that review may mean reevaluating steps in research or cataloging that seemed settled and done weeks or months earlier. As in other matters, the new editor’s best source for both caution and encouragement in planning a search and creating control files may be the experience of editors who have survived the ordeal in similar theaters of scholarly operations.

Putting the Control File to Work

I. The Search for Documents