Much of the terminology of modern print publishing, and many practices of documentary editing, are based on completely obsolete methods of book production. While modern technology has introduced its own set of questions and puzzles, some practices of proofreading and collation still in use were developed in the days when printer’s copy was set by hand or by “hot metal” and sheets of paper were pressed onto the inked type to produce galley proofs. All that is left of these earlier techniques is an occasional piece of jargon—and the scholarly editors’ tradition of scrupulous oversight at every stage in publishing a documentary or textual edition.

Computer technology provides its own options for books, microforms, and digital editions. Machine-readable forms of a book edition, appropriately coded for changes of typeface and spacing, are run through computers in one of several systems of computer-assisted composition. The computer makes certain typographic decisions like line endings and justifications, and the record that results then drives the typesetter itself. These typesetters, in turn, summon up the appropri-ate images for letters, numbers, and marks of punctuation, projecting these images onto photographic paper or film, which produces the twenty-first-century equivalent of page proofs, sheets that show what will appear on each page of the proposed volume or online edition.

The most convenient and reliable introduction to changes in book production and the accepted methods of dealing with modern and traditional techniques is the 15th edition of The Chicago Manual of Style (2003). While most students turn to this reference work for advice on annotational forms and punctuation, too few venture as far as appendixes A and B, “Design and Production—Basic Procedures and Key Terms” and “The Publishing Process for Books and Journals.” These sections provide an excellent survey of traditional bookmaking and a sound discussion of modern computerized technology.

While the various methods of computer composition offer great economies in production costs, they also create new problems for the documentary editor. Although they eliminate the possibility of successive human compositors introducing new errors when correcting old ones, computer systems house their own gremlins. They may not be as capricious as the human capacity for error, but they’re still there. They seem to arise most commonly when computer-generated text is converted from one software program to another. Publishers catch most of these before the page-proof stage, but some will survive. Computer-assisted composition means that page proofs must be checked far more scrupulously than formerly because it’s no longer one of several stages when “proof” is available, it’s the only one.

The practice of using collating machines to compare page proofs and editorial copy was once a common solution, but modern documentary editors and their publishers have found new ways to cope with the idiosyncrasies of computer methods, and new rules for proofreading and collation have emerged. Today, for instance, many editions use equivalents of the “compare” function in word-processing software to check the machine-readable versions of page proofs against the text submitted to the publisher (see Herman Saatkamp, “The Editor and Technology”). Computer proofreading of this sort far outstrips manual methods in speed and accuracy.

Whatever the methods used to produce a printed volume, wise editors still demand advance sample pages from their publishers’ designers. The wisest ones submit copy for truly representative samples of documentary texts, footnotes, textual apparatus, glossaries, and any other special features planned for the edition. Inappropriate design and format can be corrected easily at this point. Even with more easily modified computer systems of typesetting, an editor is ill advised to demand a new typeface for footnotes or a revised pattern of paragraph indentation after receiving page proofs.

No modern editor looking for a book publisher should work with a press that refuses to accept machine-readable text and notes instead of hard copy that must be rekeyboarded. Editors who take on the responsibility of providing those electronic records must be sure, in advance, that their word-processing system will produce something compatible with that used by the press’s compositor. Often the press will dictate the word-processing system acceptable to its composition system. If the editor cannot comply, allowance must be made for the costs of converting the edition’s records into a form that can be read by the printing system.

Editors at projects who do not furnish the press with machine-readable copy for a computerized system must be especially effective at communicating their own wishes. They will, for instance, learn to place a check mark or an “OK” above odd usages in the documentary texts so that no publisher’s editor will “correct” such authorial errors. They must furnish publishers with legible double-spaced typed copy, with generous margins for questions from the press or instructions to the compositor.

Details that seem obvious to the editor after years of experience with the documentary edition may demand explanation to those who will publish those volumes. The designer cannot project the patterns of the printed book until she or he knows which elements of the texts or notes require visual emphasis and which can be left to speak for themselves. The editor who submits machine-readable copy that includes in-house codes must supply the compositor with a key to those codes.

A volume’s very size can influence its usefulness. While publishers understandably argue the economics of issuing one large volume instead of two smaller ones for the same texts, the editor must remember that a volume running to one thousand pages will itself be extraordinarily expensive. Because of its bulk, Charles Hobson pointed out, it will “serve a strictly utilitarian function as a reference or research tool. Such a tome is better designed for use at the library table than for holding comfortably on one’s lap while sitting in an easy chair” (review of Papers of Benjamin Latrobe, 230).

An editor’s questions cannot stop with broad considerations of design or electronic systems. If some copy will be rekeyboarded by a compositor, the editor must be sure that the printer’s copy is marked appropriately. Most major university presses and commercial publishers claim to follow the methods described in the Chicago Manual of Style, but all have their peculiar house rules for certain matters. Close and early communication with the publisher’s staff will minimize markup problems.

A. Desktop Publishing

More and more presses now prefer to enter printing codes for books that they publish, repurposing materials in XML or converting word-processing codes, but some editors still produce their own camera-ready copy. Commercial software usually described as “desktop publishing” allows an author or editor to create machine-readable files that generate images of pages of the print edition in final form, complete with appropriate typefaces and page numbers. In truth, of course, this is not desktop publishing, but merely desktop page makeup. The official publisher still discharges the roles of contracting with printer and binder, marketing, and sales.

At first glance, this might seem a dream come true for editors desperate to prevent typesetters and publishers’ copy editors from tinkering with editorial texts and notes, but there is a dark side. Commercial and university presses are not unaware of the savings realized by accepting machine-readable files or camera-ready files. In response to these technological changes, some have reduced or eliminated their own staffs of compositors, proofreaders, and editors, who once could be counted on to review successive volumes of documentary series for style and sense. While documentary editors may welcome their role as desktop publishers, they know that the volumes still need prepublication review from scholars and laypeople who have had nothing to do with their preparation. Thus, computer technology makes the editor’s need to enlist friends and colleagues for such duty more urgent than ever.

II. Microforms

Submission of texts and notes for a microform publication is largely a process of physical arrangement of the items to be photographed and quality control of the resulting product. Editor and publisher should agree on the form of target or introductory note that precedes each item in the microform, and camera-ready copies of these notes can be generated by word-processing equipment once proofreading and verification are completed.

Even with the most sophisticated microfilm or microfiche cameras, materials must be photographed in the order in which they will appear on the published reels to take advantage of automatic frame-numbering devices. This means that the editor must not only complete but execute the plan for chronological or topical sorting in time to meet the microphotographer’s schedule. Although items can later be stripped into the master negatives used to generate the film or fiche copies, splicing can produce an uneven surface that compromises quality of reproduction.

Microform facsimiles must be proofread, not for errors in a compositor’s resetting of characters, but for standard quality of images. Prudent microform editors ask their publishers’ production staff for an advance review of samples of the original or photocopied materials that will appear in the film or fiche. These technical experts can often predict which sources will not reproduce satisfactorily, and the troublesome items can be transcribed well in advance. Such microforms will include images of the source’s target, the source itself, and the editorial transcription.

Most scholarly microforms are accompanied by a separate finding aid containing not only an index to the documents but also statements of editorial method and historical introductions analogous to those required of book editions. Microform indexes are usually generated by the edition’s control files, and even if they are not as elaborate as book indexes, they must be as accurate and clear as the edition’s budget and schedule permit. Richard Ryerson suggested that the underutilization of the microfilm edition of the Adams Family Papers “may be as much a result of the film’s lack of an index as it is of any inconvenience or actual difficulty in reading the text” (“Documenting the Presidency of John Adams: The Adams Papers Project”). The general rules for indexing that appear later in the chapter are geared to book editions, but their principles apply to microform finding aids as well.

III. Electronic Editions

When the second edition of this Guide went to press more than a decade ago, the publication of documentary editions in CD-ROM and DVD format seemed to be the wave of the future. Scholars had just begun exploring the Internet’s ability to provide free or fee-based access to such resources independent of computer-disc players in individual libraries. Early enthusiasm for CD-ROMs and now DVD formats has faltered. In these self-contained electronic editions, the editor was often limited by proprietary textual and retrieval software used by a given commercial publisher. Perhaps more important, libraries soon preferred to subscribe to online editions rather than burden their staffs with the problems of housing and mounting disc-based electronic products. Most editors found greater promise in editions distributed over the Internet or some other electronic network. Here the end users’ browsers determine the appearance of the electronic texts and the ways in which those texts may be accessed. These Web-based editions may be free to all, or available by subscription; mounted at stable institutional sites, or dependent for their life on a commercial publisher’s sales figures.

Literary scholars were pathbreakers in producing machine-readable texts suitable for transmission to colleagues around the world on the Internet. Most early undertakings were directed toward fellow scholars collaborating in a textual venture. Since the 1990s, though, scholars in the humanities have turned their attention to more formal publication and distribution over the “Net.” Perhaps the best known is Jerome McGann’s Rossetti Archive at the University of Virginia, which its creator describes in “The Complete Writings and Pictures of Dante Gabriel Rossetti: A Hypermedia Research Archive.” While texts transmitted in early Internet projects were machine-readable transcriptions rather than scanned images, the Rossetti project provides images of manuscripts and printed works in addition to machine-readable texts. At first, facsimile publication of this kind was limited both by the time necessary to transmit the information needed to record the visual information on a page and by the fact that accurate reproduction of some of these images demanded hardware and software unavailable to some users. More recent technological improvements have minimized these problems to a considerable degree, and Internet-based documentary editions are a far more practical undertaking than many experts imagined only a decade ago.

Although the majority of electronic editions being launched now are Internet-based, some editions still find DVD publication an attractive option. Luckily, the development of standard systems of tagging and coding mean that the processes for publishing in either format are now so similar that we can discuss them as part of the same process, not two distinct methodologies.

As yet, there are comparatively few completed scholarly documentary editions in electronic form to which we can point as models. Notable exceptions are the Abraham Lincoln Legal Papers and The Correspondence of John Dewey. Several projects are works in progress that have already published sizable elements of their planned collections of images, texts, and notes: the digital edition of the Papers of Dolley Madison and the Rossetti Archive at the University of Virginia, and the Blake Archive.

The publishing history of the Lincoln Legal edition is a particularly interesting one. Eventually, the project will have had an electronic existence on DVDs and the Internet, with Web access moving from fee-based to free. Not long after the DVD-ROMs were issued in 2000, the project was able to secure several grants that made the edition available to many small colleges and to nearly every law school in the country. After publication, substantial numbers of “new” Lincoln legal documents and cases surfaced, and the editors found data errors in the DVD edition. Pressing new DVDs to include the additional documents and corrections of the errors was not an efficient option. Increasing Internet bandwidths and improved computer capabilities meant that delivery over the Web became a practical solution to the problem. Work is now in progress to convert the DVD version to an Internet publication that will be freely available and open to frequent correction and updating.

The project’s experience and methods remain instructive. The rationale of the edition’s design and the history of its creation are ably discussed in Martha Benner’s “The Abraham Lincoln Legal Papers: The Development of the Complete Facsimile Edition on CD-ROM.” Digitally scanned images of collected photocopies were linked to their records in a relational database for publication on CD-ROMs, and the Lincoln Legal editors followed a four-step process in designing and producing this edition: (1) converting documents to electronic form, (2) developing a search program and user interface for retrieval, (3) constructing a prototype disc for testing, and (4) pressing the master copy for the product. Every step but the last was performed by the project staff.

An electronic image edition has one obvious advantage over microforms: the images and textual records can be recorded in any sequence. The user’s search commands will reorder them later. This consideration was central for an edition of legal papers, where records for the same case may be accessioned from dozens of sources over a period of months or years.

The unpredictability of the needs of users of legal papers also dictated an electronic medium. In microforms, orderly physical arrangement is a key to usability. When an editor is absolutely certain that readers will be served by a single sequence of images, film is comparatively easy to use, but researchers approach legal records with a variety of interests, ranging from the law itself to social or demographic history.

In the real world, most editors cannot predict how their images will be used. On microfilm, editors must use targets for cross-references to other documents in the series. And there is no hope for reordering what appears in microforms. In an electronic edition, cross-reference links can be made directly. Electronic editions enable each scholar to retrieve and reorder documents at a single workstation, custom designing a collection of documents and notes for her or his needs.

Ease of access in an electronic facsimile edition depends on appropriate editorial machine-readable indexing. The images of the documentary sources are not open to full-text searching as are machine-readable transcriptions. With a DVD or a fee-based Internet edition, or even with an academic Web site with established protocols, the editor may be limited by the chosen publisher’s proprietary software or traditional preferences. Preliminary discussions must concern the number and form of searchable fields as well as the quality of the images to be scanned in for the new publication.

Although an electronic facsimile may not include a separate index in the traditional sense, it should include a thesaurus of indexing terms or descriptors used in its creation. Keywords in targets and notes will be searchable, and a user may be able to create an index “on the fly” from XML tags. Tables of contents or careful listings of document locations or archival organization can also provide easier access.

Electronic image editions, like microforms, demand statements of editorial method appropriate to their format. “Drop-down” lists of subjects or names will help. Indeed, for a very large collection, anything that gives the user some sense of what a collection contains and how it is referenced can be invaluable.

Certain problems of a physical rather than an intellectual nature plague electronic editions. There are as yet no recognized standards for archival storage of this format. Even if the basic data for a Web-based edition exists on tapes or discs, we don’t yet know of an assured way to provide for its long-term preservation. Further, the proprietary nature of some of the software that provides user interfaces may make some editions intellectually obsolete if the original publisher goes out of business—no one else may be able to provide support for the use of the data.

Despite these possible drawbacks, it’s essential that almost any documentary editor at least consider some form of electronic publication. More and more sponsoring agencies insist on a strong electronic component in the projects they fund. This consideration may influence almost every step in a project’s history, from giving serious thought to the underlying document structure and encoding language for transcriptions and annotation to the choice of publisher.

Funders of electronic editions have also become conscious of the need to make these editions interoperable. Many editors already feel the pressure to make their projects’ past, current, and future files part of some shared database. The University of Virginia Press’s plans for the American Founding Era project envisions Web-based publication of the existing print volumes of the Papers of George Washington, the Adams Papers, the Papers of Thomas Jefferson, the Documentary History of the Ratification of the Constitution and the Bill of Rights, the Papers of Alexander Hamilton, the Papers of James Madison, and The Political Correspondence and Public Papers of Aaron Burr, as well as the Web-based edition of The Dolley Madison Digital Edition. Once completed, this multiedition archive will allow researchers to search simultaneously over all the series involved.

Editors who want to leave this option open will need to use common DTDs (document type definitions) in a standardized way. Planning ahead for publication as part of a larger pool of materials will include such mundane considerations as using a subject’s full name rather than initials, working together with other editions to share indexing terminology, or even developing a program that will convert idiosyncratic index entries to a uniform method that can be applied to all the editions in the databases.

As some new challenges for electronic editions appear, others fade in importance. The unpredictability of a reader’s workstation once created serious challenges in this form of electronic publication. The codes used by one piece of word-processing software to produce italic type, superscript characters, and the like bore little resemblance to those used by another manufacturer’s product. Today, documentary texts can be coded in XML, a system widely accepted by the digital humanities community. Indeed, XML tagging is becoming standard for any text that will have some degree of longevity. Editors must also familiarize themselves with developments within the Text Encoding Initiative (TEI), which is creating standards for humanities-based texts.

Luckily, commercial word-processing products now include options that convert these proprietary codes to XML tags for details of physical appearance and inscription. Additional tags can be added to indicate analytical aspects of the text such as proper names or parts of speech, and XML-encoded texts can accommodate hypertext links as well. Computer operating systems now provide a wide variety of XML editors, and most editors will have no difficulty using this encoding language.

A. Standards for Electronic Editions

Beginning editors who search for clear statements of such standards in terms of mandatory requirements from professional organizations will be disappointed. There are, however, several recent books and Web sites that can help put a new electronic editor on the right track. William Kasdorf’s edition of The Columbia Guide to Digital Publishing, while now a few years out-of-date in terms of technical details, remains a first-class introduction. Chapters on electronic publication in Daniel Cohen and Roy Rosenzweig’s Digital History and Susan Schreibman et al.’s Companion to Digital Humanities should be studied carefully. (Remember that these two books are also available online.) Web sites for the Scholarly Electronic Publishing Bibliography (http://epress.lib.uh.edu/sepb/sepb.html) and the NINCH guidelines (http://www.nyu.edu/its/humanities/ninchguide/) provide up-to-date information, while the TEI Web site provides full description of the DTD and how it is used (http://www.tei-c.org/).

The Association for Documentary Editing has defined an electronic edition and listed broad “minimal standards” on the ADE Web site (https://www.documentaryediting.org/wordpress/?page_id=508). These standards do not offer the specifics that some novice editors might hope to see, however. Today’s variety of publication formats and rapidly changing technology mean that all editors cannot adopt the same procedures to produce some sort of homogeneous electronic publication. Rather, they strongly suggest ways that electronic publication can maintain the same intellectual quality seen in print publication, and they are an invaluable source of advice for noneditors considering mounting materials on the Web.

The best operating standard we can offer here is that any electronic edition must work at least as well as a print or microform series. The added expense entailed by these formats can be justified only by improved service to the edition’s audience. A well-designed electronic edition may offer users any or all of these bonuses: (1) access to several versions or views of the same text (image, diplomatic transcription, clear text), (2) options for organizing and reorganizing the edition’s texts in more than one way, (3) annotation that serves more than one document without cross-references or back-of-book index citations, (4) easy and immediate access to a cumulative index for an ongoing series, and (5) inexpensive and easy incorporation of maps, drawings, and other related images.

IV. Documentary Editors and Their Publishers

Whether editors choose book or microform or electronic publication, they are seldom responsible for every aspect of production unless their projects are of limited size and their aspirations for distribution so modest that they control all the manufacture and distribution themselves. For projects of larger scope, some organization or firm will be found to act as publisher and may assume responsibility for various aspects of manufacture and distribution. The most important factor in making the final publication process as painless as possible is the degree to which the editor and publisher view this as a cooperative venture, not a war between competing factions. The best defense against such friction is an exchange of ideas, for misunderstandings usually arise from mutual ignorance. A press that has survived one documentary series is more likely to be prepared for the special problems presented by a second, while the veteran of one documentary edition is more likely than a novice to understand the limitations and capabilities of publishing design and technology.

Editors are responsible for making their needs known to the publisher—and doing so as early as possible. They must not assume that anyone else will understand the special problems of their texts, notes, and apparatus; and if they fail to explain them adequately, they risk an edition that will be a source to be avoided rather than consulted eagerly and conveniently. A helpful guide to editorial responsibility toward publication issues appears in Beth Luey’s “Publishing the Edition.” Confer with the publisher’s editors and designers as soon as a publisher has been chosen. Never be afraid to betray ignorance by asking questions. And never hesitate to volunteer information or ideas to the publishing staff.

Documentary editors unfamiliar with the principles of book or microform design or electronic formats must brief themselves on their chosen medium of publication, formally or informally. By consulting other editions, they can get an idea of which methods are visually and intellectually effective for certain kinds of texts and annotation. Steven Burg and Michael Stevens’s Editing Historical Documents makes such comparisons far easier. While their examples are limited to book and microform editions, they remain instructive in the best sense of the word. When tentative decisions are reached concerning publishing format and design, the editor should consult an expert in the chosen medium to learn whether the plan is economically practical and technically feasible. Ideally, editors choose publishers early in their projects’ careers so that they can direct questions to the designer or production director who will be responsible for their own texts and notes. If this is impossible, they need to recruit their own advisers.

V. Front and Back Matter

Every format for a documentary edition demands editorial contributions equivalent to a book’s front and back matter. Because of the special production problems created by documentary editions in the past, book publishers often preferred to be given copy for the documentary texts and editorial notes as early as possible. This is no longer as serious a consideration, but most editors still prepare front and back matter after documentary transcriptions are completed. It’s far easier to explain what you’ve done after you’ve done it.

Such prefatory sections and appendixes are far more significant in a documentary edition than in scholarly monographs or narrative histories. If readers are to make full use of any edition, they must know the project’s standards for establishing texts and furnishing annotation, in addition to textual records and tables that may be necessary to recover details of the original sources. “Full disclosure” should be every editor’s motto here. The users’ need to know is paramount. The various files and memoranda compiled during the editing process serve a vital function when the editor prepares the explanatory statements and apparatus that allow readers to make use of the final product, whether it be printed volumes, microforms, or electronic texts. Specific groups of documents may require special editorial discussions, but each group published simultaneously should contain the following elements:

1. Statements of editorial method, covering textual matters, informational annotation, and other editorial policies

2. A list of all textual symbols used in the editorial texts and a list of any obscure abbreviations used, such as “ALS,” “D,” and “Dft”

3. A list of short titles for published works and abbreviated forms designating repositories used in the edition or their appropriate equivalents

4. A list of permission policy statements. When the source texts’ provenance so demands, the editor must inform readers that permission to cite or to quote certain items must come from their private or institutional owners. In some cases, the sources’ provenance may also require formal statements that the editor has received permission to reproduce these texts or been granted permission to reprint material in copyright.

Editions that employ clear text in printed volumes often place the first and second of these elements in the back of the book, immediately preceding the record of emendations and inscriptional details. In other editions, all four elements precede the first documentary text.

In electronic editions, it’s impossible to control the ways that readers may use an edition. Many skip carefully drafted statements of editorial policy and guides to search procedures, preferring to go straight to the index or skim documents. All that an editor can do is provide easy links so that the reader who becomes confused will have immediate access to the tools she or he needs.

A. Statements of Editorial Policies

As an edition progresses, in-house statements of procedure and rules for transcription and emendation and annotation accumulate for staff use. When the edition is ready to be shared with its audience, these documents must be polished into final form to accompany the edition.

1. Textual Policies

The statement of textual method indicates which broad option the editor has chosen for textual treatment of the sources. In an edition employing anything but facsimiles or diplomatic transcriptions, the editor will probably need to expand this general statement and elaborate on the treatment of specific textual problems in the source materials.

Whatever textual method is chosen, the editor also makes clear the implications of that method for the documents at hand. Even with a back-of-book textual record, the reader deserves fair warning if clear text produces edited versions of letters that are at substantial variance with the texts read by their recipients. Editorial decisions on what is included in inclusive texts may consistently mask categories of details significant to a group of readers. Of course, the responsibility of the editor who does not supply such a textual record is even greater. If certain items in the edition demanded textual treatment different from the norm, the editor should warn the reader of such exceptions. If the textual notes to these exceptional items do not indicate that special liberties have been taken with them, the editorial introduction should list such documents for the reader’s guidance.

CSE guidelines suggest that the editor also describe the methods followed by the editorial staff in proofreading, collating, and verifying the editorial texts of documents. Among early historical editions, only the Woodrow Wilson Papers offered the readers a similar exposition of editorial practices. The practice is more common now among documentary editions and deserves universal imitation.

2. Annotational Policies

The editor’s statement on annotation presents the general policies on identifying or explaining matters mentioned in the documentary texts. It also points out any supplementary devices such as biographical directories, glossaries, or gazetteers. The editor should never hesitate to be honest about the limitations of the edition’s informational annotation. The preface in each of the twenty-six volumes of the Hamilton Papers closes with the forthright statement, “Finally, the editors on some occasions were unable to find the desired information, and on other occasions the editors were remiss.” Such admissions of human fallibility do not prevent scholars from benefiting from the edition’s documentary texts and editorial notes.

3. Special Statements of Policy

The circumstances that most commonly require additional explanatory sections arise from the edition’s scope or its organization. Selective editions demand a clear statement of the editor’s criteria for choice. Editions arranged by some topical logic rather than chronology require statements of these organizing principles. The absence of such statements remains the cause of the sharpest criticism of selective series.

In addition, many editions’ source texts demand statements concerning the searches that uncovered these original materials. Models of these accounts are the “Primary Sources” sections of the Eisenhower edition’s “Bibliographical” statements for each chronological subseries. These describe the project’s search for materials as well as any peculiarities of arrangement in government records or other repositories (see, e.g., 9:2259–73, 11:1521–23, and 13:1507–11).

B. Textual Symbols and Other Source-Related Tables

The editor can never assume that any symbol or abbreviation used for textual details or the forms of source texts will be self-evident in its meaning. No reader can be expected to guess about such matters. Although few editors would be rash enough to omit a list of their textual barbed wire, some have been careless, offering inadequate definitions of the meanings adopted for the inscriptional forms represented in their editions. Indicating that “L” stands for “letter” is not enough—the editor must explain whether this means the final version of that source, the recipient’s copy. Such definitions may vary according to the peculiarities of the source texts involved, and readers need every clue they can be given.

C. Short Titles and Symbols for Repositories

It’s usually wise to separate the list of short titles and symbols for repositories from the list that explains alphabetical symbols for the forms of source texts. Indeed, when there are a large number of printed sources and repositories in abbreviated form, it may be better to provide a separate list for each of these categories as well.

In electronic editions, of course, textual symbols and short titles can be elegantly expanded. The online edition of the Papers of George Washington volumes, for instance, allows the user to move the computer mouse over each abbreviation to see its expansion or definition in a pop-up box. An electronic edition should provide the user quick and ready access to this information from anywhere in the edition. Born-digital electronic editions such as the Papers of Abraham Lincoln will not use textual symbols or short titles at all. Space considerations online are insignificant, and readers will be spared clicking to a link or scrolling a mouse to find the meanings of “MHi” or “PHi.”

The three sections described above should appear in every volume or other unit of a continuing series. Some editors have assumed that they need be discussed fully only in the first number of an edition, but this places an unreasonable burden on the reader. Few individuals can afford to purchase every number in a series, and no library can guarantee its patrons that volume 1 in each set will remain on the shelf. Even in an electronic edition, these reference sections should be easily accessible from every point in the text to which they apply. Continuing multipart series may create a need for modified statements of editorial method. Editorial practices change over the years in which the volumes or electronic components are published, and an editor is obliged to explain modifications of textual treatment, informational or textual annotation, and format. Successive numbers in a chronologically arranged series may take the editor into new categories of documents that require fresh textual or annotational methods. When this occurs, the introduction to the segment in which the change arises should carry an explanation of the problem and of the solution the editors adopted.

In some traditional print editions, succeeding volumes in a series adopt devices that allow the reader to consult the edition as a whole. The clearest example of such a technique is the Madison editors’ use of their indexes to provide cross-references to pertinent annotation in earlier volumes. Whenever the index to a specific volume becomes part of a broader information retrieval system, the editorial note introducing the index entries should be expanded to explain the method involved. If basic statements of textual method are not republished in succeeding numbers of a documentary series, readers of later units will find only amendments to earlier editorial laws, not comprehensible recitations of current practice. Because these sections should be as concise as possible, their repetition in each volume or addition to an electronic edition adds comparatively few lines, and their inclusion makes the new materials appreciably more useful.

D. Optional Editorial Apparatus

Any edition’s design may demand special features: informational annotation supplemented by glossaries, gazetteers, or discrete sections of biographical sketches, or clear reading texts accompanied by textual records, historical collations, and hyphenation lists. Modern editors of book editions show considerable imagination in lightening the burden on the pages of volumes of printed documentary texts by mounting glossaries, biographical dictionaries, atlases, and the like on Web sites maintained by the edition’s sponsoring institution. The editors of the new edition of the Joseph Smith Papers have been exceptionally inventive in this regard. These supplements are even easier for readers of electronic editions to use, for links from text or notes will take them immediately to the glossary or textual record they need.

Any selective edition not supplemented by a comprehensive facsimile edition of the larger collection from which it draws usually provides some sort of calendar or checklist recording the documents that are not part of the printed volumes. Any selective edition—as well as editions organized by topic rather than by chronology—must provide the reader with a table of contents. Even comprehensive editions organized chronologically do their readers a courtesy by providing such a table.

E. Permissions for Use and Copyright Considerations

Lucky indeed is the editor of an archival edition whose source text is owned by the agency sponsoring the publication project and for which no conflicting claims of intellectual property rights exist. At the very least, all others must obtain permission for reproduction or publication from the individuals or institutions that own the source texts to print editorial texts based on those documents. Editors dealing with sources of the modern era must become familiar with modern U.S. statutes on copyright and intellectual property. Computer-based control files ease the task of identifying the dozens or hundreds of owners of letters in an edition of correspondence. Word-processing software expedites preparation of a form letter seeking permission to publish and integrating the addresses of the owners who will receive the request.

Advice from a publisher’s legal staff, not computer hardware or software, is the best defense against claims of copyright infringement by authors (or their heirs) of the private or public source texts on which the edition is based. Such claimants should be identified earlier rather than later in the edition’s career so that their wishes can be determined while time remains for any necessary negotiations—or for revision of the edition’s scope if agreement cannot be reached. Useful summaries of copyright-related problems appear in Michael L. Benedict’s “Historians and the Continuing Controversy over Fair Use of Unpublished Manuscript Materials.” Volume 10 of TEXT (1997) carries a useful special section, “Intellectual Properties: Copyright and the Status of Texts,” including David Greetham’s hilarious contribution, “Rights to Copy.”

VI. CSE Inspections

In decades past, inspections by a team of scholars delegated by the Council for Scholarly Editions of the MLA followed a carefully prescribed and rigorous pattern. Editors submitted not only documentary texts and informational notes but also draft introductory statements and textual apparatus. CSE guidelines once suggested that editorial texts be given no fewer than five independent proofreadings at the “most appropriate” points in the process of creating printed volumes. At least one (and preferably two) of these proofreadings were to precede submission of copy to the “printer.”

Today, the CSE’s “Guidelines for Scholarly Editions” are far more practical—and less explicit. Recognizing the variety of methods now used to produce machine-readable versions of texts and notes that appear in print, the 2006 guidelines suggest that “editors of scholarly editions establish and follow a proofreading plan that serves to ensure the accuracy of the materials presented.” CEAA/CSE inspections were once generally conducted at the offices of the project that produced the edition, but on-site inspections are no longer the rule. Today the editors and the “vetters” named by the CSE are left to work out the methods most appropriate to the unique problems presented by the texts in question and the technical methods used in their publication.

While this might seem less daunting than the older, more detailed standards, the newer policy in fact places greater responsibility on the editors and inspectors, for they must repeatedly redesign their system of inspection and approval rather than rely on a preexisting rulebook. Editors following these guidelines today often confer with inspectors even before their initial volumes are ready for inspection, so that problems can be anticipated and resolved. Once the manuscript for that first volume has reached the status of printer’s copy, with initial proofreading and verification of documentary texts and textual apparatus completed and final versions of nontextual annotation ready for review, the inspectors begin their official work. If the inspector raises points that require clarification or correction of textual method, the editor can make these adjustments before the editorial texts enter page composition, when corrections and revisions are more difficult and expensive.

Inspectors still read the statements of editorial method carefully and make comparisons of a representative sample of editorial texts and textual apparatus against photocopied source texts provided by the editor or against originals of those sources located at repositories that the inspector can visit conveniently. Editors still must be prepared to answer any questions about textual matters and informational annotation. When a volume is recommended by its inspector and approved by the CSE, the editor may move the corrected, machine-readable version of text and notes forward in the composition process and inform the publisher that the verso of the book’s title page may bear the emblem “An Approved Text.”

VII. Final Establishment of the Documentary Text

Even though CSE inspections usually occur before copy proceeds to page proofs, no scholarly editor pretends that responsibility for guarding a text’s integrity ends at this point. All an editor’s painstaking work can be undone if there is a failure to monitor the progress of the texts through the ensuing stages of production. Editors need to establish policies for proofreading or collating proofs long before receiving them. Publishers do not take kindly to delays in production schedules caused by an editor’s agonizing over the best methods for checking these materials. Whatever combination of proofreading or collation is chosen, there must be a system of clearly assigned responsibility.

In addition to proofreadings and visual collations, editors, like any responsible authors, read transcriptions and page proofs for sense. Oral team proofreading can conceal errors of transcription in homophonic words, and human fallibility is always at work. Completely illogical words and inexplicable omissions have been known to survive the most rigorous series of independent proofreadings and electronic collations. Even though all editions are set electronically before going to a compositor, errors often crop up in random and unexpected ways. Special symbols, single and double quotation marks, and style renderings seem to be peculiarly vulnerable to such computer glitches. Editors need to be vigilant, since errors of this kind may reflect a pattern that might be repeated throughout the page proofs.

A. Textual Notes and Apparatus

Only with a printed facsimile edition will proofreading, collation, and verification of the documentary transcription end with checking the body of that transcript. With all other methods, there will be editorial notes or other apparatus that points to specific details in the source that have been emended or suppressed in the version that the new edition presents. Editors working within the CSE tradition may not be able to complete their textual apparatus until page proofs are available. If the textual apparatus refers to printed lines, only then can the editor begin the work of inserting the line numbers. In projects whose editions use justified right-hand margins, editors check page proofs for the introduction of new, ambiguous line-end hyphenations and list each one. These lists remain tentative until page proof arrives and the final hyphenation list can be created. After page proofs have been checked, the CSE editor can prepare the final version of the back-of-book textual apparatus and hyphenation lists. Subsequent page proofs of the editorial apparatus must be proofread rigorously as well.

B. Nontextual Annotation

Even though there are no published standards for verifying and proofreading nontextual annotation, an editor must maintain the integrity of these editorial contributions as well as that of the documentary texts. Quotations in notes deserve the same high standards of proofreading and perfection applied to the documentary transcriptions. These standards must also be maintained when the editor reviews page proofs.

VIII. The Index

Paradoxically, a documentary edition’s most important tool for accessing its texts and notes—the index—is the last to be prepared in final form. No reader who’s tried to make sense of a badly indexed or unindexed collection of texts will quarrel with the editorial law that no edition of documents is any better than the index that serves it. Because the design of the index is central to an edition’s success, editors find themselves planning that finding aid long before the documentary texts and their notes are ready for publication—perhaps before the first text is transcribed or the first note written. Modern editors realize that they should determine early in their work whether a cumulative index will be the edition’s final publication. This decision will dictate that the edition establish standardized language in all its indexes throughout the multiple-volume series, thus enabling a seamless accumulation at the end.

The wide variety of software products now available for computer-assisted indexing makes the process far less grueling than it once was. To save time and money, editors should talk to colleagues about the software they use for indexing. It’s especially useful to seek out someone working on a project of similar size. Any project’s budget should include money for indexing software, just as it does for word-processing and accounting software. Another invaluable source of expertise on indexing is the Web site of the American Society for Indexing (http://www.asindexing.org/site/index.html). This organization provides professional assistance, training, and software advice to freelance indexers nationwide. The Web site’s section “Software Tools for Indexing” gives some idea of the choices available for various computer operating systems, Web-indexing, and the construction of a thesaurus (a list of terms approved for use in the index). Numerous projects, such as the James Madison and Thomas Jefferson Papers, now use CINDEX, a commercial product that can be networked in an office so that editors can collaborate on indexing work. A tool like this enables easy cumulation of multivolume indexes.

Electronic editions can take advantage of full text searching as well as multilevel indexing markup made possible through internal tagging. The methodology for such a process at the Thomas Edison project is briefly described in Robert Rosenberg’s “Documentary Editing.” Similarly, prebuilt taxonomies can be linked to an edition through a database or some other external system.

Computer-assisted indexing also means that the process of identifying and recording subjects and the form of their entries can begin long before the delivery of page proof. Most word-processing systems include some basic indexing system that will enable an editor to maintain a running list of the location of personal identifications, explanation of terms, and so on. This can become the basis of the thesaurus of approved terms for the index as well as a rough locating device for the references. A system of this kind can be part of the annotational index an editor maintains while preparing the edition.

A. Designing the Index

Documentary editors who have never before prepared an index should not try to pick up the skill on the job. Reliable introductory books on the mechanics of indexing abound. Good choices might be chapter 9 of Beth Luey’s Handbook for Academic Authors; chapter 18 of the Chicago Manual of Style (15th ed.), “The Mechanics of Indexing”; Nancy C. Mulvany’s Indexing Books (2d ed.); and, of course, the American Society for Indexing Web site. After mastering elementary rules, novices should examine indexes in documentary editions whose contents resemble those of their own and compare those indexes with one another and with the entries under “Indexing” in Editing Historical Documents.

Many editors regard the indexes of the Adams Papers as a model. In preparing the index to the edition’s first series, Lyman Butterfield created a thesaurus structure that established a pattern of entries and subentries that proved adaptable to the succeeding thirty volumes of Adams diaries, letters, and public papers. The specific subject entries in the Adams series cannot meet the needs of other editions, but they demonstrate the intelligent planning and anticipation of readers’ needs that are the standard against which others can measure their own efforts.

Admirable as the Adams Papers indexes are, some critics question their treatment of subject entries. The Adams model places many topics as subentries under the name of the individual who is the subject of the volume in question. Readers interested in a topic like suffrage would need to look under both “Suffrage” and “Adams, John: views on suffrage.” A good index doesn’t force the reader to look under both places but would make the subject heading “Suffrage” the main point of entry. Indexes with excessively long entries representing every action and activity of the edition’s subject are more like summaries of the volume than true finding aids.

In fact, the most consistent criticism of indexing in documentary editions, whether the records of an individual or a movement, of a politician or a novelist, centers on the preponderance of proper-name entries. As any indexer can testify, these are the easiest entries to recognize and organize. David Nordloh pointed out that an index of an edition of letters confined to personal-name entries “not only suggests but enforces the notion that an author wrote to people but not about anything” (“Supplying What’s Missing in Editions of Selected Letters,” 42).

Accurate and helpful subject indexing demands the knowledge and experience of the scholar or scholars who compiled the edition; it cannot be delegated to a junior staff member, much less to an outside indexer. Only someone intimately familiar with the materials to be indexed and with the needs of the people who will use the edition can design an effective system of subject entries, and only someone with that same knowledge can match the terms that appear in the documents with the appropriate index entries. The indexes to volumes of the Letters of Delegates to Congress series, for instance, routinely contain entries like “Delaware River, defense.” No outside indexer could be expected to know that the names of posts like Fort Mifflin and Fort Mercer should be included under “Delaware River.” The index of a documentary edition of a work compiled or written for publication often includes the sources cited by the author. The absence of index entries for such pamphlets and books in Jefferson’s Parliamentary Writings challenged readers trying to determine when and how Jefferson employed his own primary sources.

The thought and hard work demanded by good subject indexing serve the editor as well as readers. A properly designed index reduces the number and length of editorial notes. The most obvious example of this is the elimination of “see above” references to earlier identifications of individuals when the index provides a key to these notes. Most commonly, the index entry in a book edition for such a note carries a keyword such as “sketch” or “identification” or prints the page references in boldface or italic type. Most important, an index with complete subject entries enables readers to gather facts or trace ideas for themselves, tasks that might otherwise demand editorial summaries of one kind or another.

Even though the rules for indexing one series cannot be transferred wholesale to any other edition, a few guidelines can be drawn from the experience of American editors:

1. The index should cover all the materials in the unit it professes to index. If a selective edition of correspondence includes a checklist or calendar of letters omitted from the printed volume, this list’s entries should be indexed. If the editor uses devices such as a biographical directory or gazetteer to streamline annotation, data recorded there should be indexed. A volume’s introduction and its appendixes deserve index entries, as do any lengthy descriptive notes on illustrations. There’s no excuse for a partial index. Even if some of an edition’s elements, such as biographies, are already in alphabetical order, the reader deserves an index that presents a guide to every page in the volume, with the exception of elements like the title page, table of contents, acknowledgments, and bibliography.

2. The editor should generally provide one index and one index only for each unit or group of units published simultaneously. This general rule was often ignored in some early editions of the papers of famous American lawyers, with a separate index for legal terms and yet another index for names of legal cases. Both the recent Daniel Webster Legal Papers and the John Marshall Selected Law Cases volumes have a single index.

3. Index entries and subentries should reflect the nature of the documents in the edition. With an edition of letters, it’s usually wise to indicate which citations in a personal-name entry pinpoint the pages where that individual’s correspondence appears. A string of undifferentiated page numbers doesn’t suffice. This can be done by using different typefaces (bold or italics) or subentries for “to” and “from.” A volume of legislative papers, for example, might require an index that breaks entries down into subentries that allow the reader to trace the evolution of specific pieces of legislation or of committee action on bills or resolutions. The editors of The Correspondence of John Bartram, the pioneering American botanist, were roundly criticized for an index that identified Bartram’s human associates but omitted plant names and collecting locations, matters of considerable interest to botanists and historians of science.

4. Index entries should be phrased to serve the needs of the readers of the edition in question. The index to a volume of legal papers, for example, can use technical phrases drawn from the law, although such subject entries would be pretentious and baffling in the index to the personal correspondence of a poet. Editors of the Journals of South Carolina’s legislature in the State Records of South Carolina series modeled their indexes on those of the existing J. H. Easterby and Ruth S. Green edition of the Journals of the Commons House of Assembly so that scholars needn’t master a new retrieval scheme.

5. Whenever a documentary edition focuses on a central figure or organization, index entries under that name need to be broken down into intelligible subentries. The reader can assume that every entry in the index is in some way related to that central figure. Certain topics, however, can be located best under the entry for the edition’s subject: opinions, literary quotations, writings, physical descriptions, opinions, and so on. Efforts to avoid this approach have been embarrassingly unsuccessful, as have experiments in limiting index entries for the central figure to “mentions,” which produce useless columns of undifferentiated page numbers.

Editors are divided over the proper arrangement of such subentries. Some impose a strict alphabetical order. When the index entry and its subentries stretch on for a page or more, this may be cumbersome. Others, whose texts are chronologically arranged, present the subentries in the order in which the topics or events to which they refer arise in the edition. Readers with some knowledge of the subject’s career often find this a more convenient device.

6. Even though a book index’s final form may not be prepared until page proof is available, editors should keep the index’s design in mind while preparing the edition itself. Unnecessary cross-references can be eliminated from the notes if the index will provide easy access to such material. The same principle holds for an electronic edition, where there may be a wider variety of access tools available to the reader.

7. A properly maintained running annotational index expedites final indexing immeasurably. When such a preliminary index shows that the “John Smith” mentioned in the text of a letter of 4 March 1866 is “Smith, John C. (1820–1895)” while the “Jack Smith” in a letter a week later is “Smith, Jackson F. (1833–1903),” the person entering data for the final index has no reason to hesitate or to key in an erroneous entry.

The most basic rule is that editor indexers must put themselves in their readers’ places. Their goal is anticipating the entries and subentries users will expect to find for the specific information they seek as well as to supply cross-references for the unexpected facts and ideas the editors have uncovered in their work.

IX. After Publication

When the last bit of corrected page proof for the index or textual apparatus has been proofread or mechanically collated, editors can do little but wait for the reviews of the printed volumes. While some reactions from readers are fair and helpful, others are ill informed and illogical. Angus Easson identified three kinds of reviews inspired by scholarly editions: the “occasioned” variety, in which the reviewer produces an essay on the author represented in the edition; the “literary critical” essay, which predicts the ways in which the world of literature and criticism will be changed by the new edition; and last, and rarest, the “bibliographical” review, which actually discusses “the principles and practices of the edition.” Ideally, Easson, commented, “The reviewer should be capable or have experience of being an editor, though yet not be a bitter or frustrated one” (“Reviewing Editions: Letters, Journals, Diaries,” 51).

Editors of microform and electronic editions face a similar postpublication situation. Creators of textual archives distributed over the Internet, in fact, can expect to be assailed by a never-ending series of reviews as the nature of those archives are expanded and improved. At least electronic editors have the advantage of being able to correct errors and omissions brought to their attention in this fashion.

The most frequent criticism of documentary editions in the past was the slow pace at which editors produced printed volumes. Consumers of microforms and electronic texts will be just as impatient. More realistically, observers might wonder that any editorial staff survives to produce a single tome, microfilm reel, or XML-encoded text, much less to publish ten or twenty or a hundred or thousands. Yet, somehow, dozens of editors have done so, and this Guide attempts to share the lessons drawn from both their mistakes and their inspired decisions.

What this book cannot share is a paradoxical satisfaction of documentary editing. If editors do their jobs well, they inspire work that will supersede much of their own. The documents that they publish will lead other scholars to search for sources that escaped their attention, and their notes will encourage others to dig a bit deeper for the historical truth that evaded them. What seemed comprehensive or definitive in their volumes will later prove to be only a landmark step toward that ultimate goal. The honest editor is, if anything, pleased by such results. Documentary editions are properly regarded not as the end of scholarly research but as its beginning. The scholar who finds documentary editing a congenial discipline must be a pioneer at heart, ready to establish a foundation of evidence on which others will build.

Publishing the Edition

I. Book Editions: From Hot Type to No Type