Chapter Seven

Nontextual Elements of an Edition

When the first edition of this Guide was prepared, one member of the Advisory Committee protested against discussing any matters unrelated to establishing a reliable documentary text. Nothing else, he asserted, was properly called “editing.” But even this textual purist would agree that, whatever you call them, considerations unrelated to textual analysis are inevitably part of an editor’s work and have a bearing on how well readers can use the edited texts.

Some of these elements of an edition are obvious. They include editorially provided notes on the provenance or physical location and history of the document that is the basis of the source text. In addition, there will probably be more than a few notes that place a document in the context of its author’s life or the times in which it was created. And, finally, there are a variety of tools that can enable users of an edition to make the best use of it in terms of intellectual access.

Before we address those elements, however, we need to speak about a series of nontextual editorial decisions that can result in errors of omission, not commission. These involve the simple selection of the documents the reader will find in the new edition and the arrangement of the chosen texts.

I. Select Editions and Comprehensive Records

In chapter 1 we touched on the basic choices among a comprehensive edition, a selective one, and a selective one partnered with a collection of surrogate images of the materials not included in the group of fully edited documents. Once an editor finishes the collection of images or original materials that are candidates for publication, another group of questions arises.

These decisions involve the distinctions between definitive, authoritative, comprehensive, and selective series. The terms comprehensive and selective have quantitative meanings only; they refer to an editor’s decision to publish all or only some of the documents that could be considered appropriate to the project in question. The terms definitive and authoritative have qualitative meanings. Modern editors are too realistic to speak of producing a definitive edition of an individual’s writings, one that is not only all-inclusive in scope but also so rigorous in its textual methods that the reader would never need to consult the original materials on which the transcribed, photoreproduced, or digitized texts are based. Where manuscript or typescript sources are concerned, definitive printed texts are a practical impossibility. At best, editors can offer readers an “authoritative” or “scholarly” edition, a collection of accurate and reliable transcriptions of those elements of the sources that can be either translated into printed symbols or adequately described in editorial notes. Some details of inscription, some nuances in the sources, will resist the most ingenious editor. Scholars whose interests demand access to these details will need to consult the original materials or their accurate photoreproductions. Few editions will ever include complete versions of every text related to its subject. Instead, they will be a selection of documents relating to some rational and clearly stated criteria.

Often, decisions about an edition’s degree of selectivity have nothing to do with an editor’s preferences. Increasingly, the wishes of publishers and funding agencies will place strict limits on the amount of material that can be included, the time that can be spent preparing the edition, and even the criteria for selection. On more than one occasion over the decades, the demands of funding agencies forced the editors of The Papers of Robert Morris to modify their criteria of selection in the name of economy and efficiency. The print volumes of The Papers of Eleanor Roosevelt will focus only on her public career as a widow, ignoring the fascinating record of her private life and earlier decades as an outspoken reformer. The decision to limit the book edition this severely was obviously not one that the Roosevelt editors welcomed.

Even in the most comprehensive editions of famous statesmen’s papers, certain kinds of documents are usually excluded. For early American presidents, these include routine documents that were signed by Washington or Adams or Jefferson or Madison as chief executive but that do not represent any intellectual contribution from that leader. These may be ship’s papers, civil and military commissions, pardons, and the like. Customarily, the editors published a sample of each form of document; only if the wording of a document changed would the new version appear in the edition. Calling cards and invitations fall in the same category. These decisions were based on the documents’ routine nature and assumed insignificance. With the advent of electronic editions, editors should consider ways in which these documents can be presented, because editorial assumptions about their unimportance preempt scholars from even considering their role in scholarly work. Historical methodology and interests do change, and editors’ decisions tend to reflect those shifts. One reviewer of a highly “select” edition pointed out that readers “need to feel confident that what has been selected exhausts the range of possibly useful sources and if not, why not” (William C. de Giacomantonio, review of Ethan Allen and His Kin).

Other editions, in order to save space, omit significant documents when their edition can offer nothing new in terms of text. An example of this was the decision of the Papers of John Marshall to publish Marshall’s Supreme Court opinions only when manuscript versions are available. When a printed version of the opinion is the only known version and it is widely available, the text is omitted from the Marshall volumes.

Usually, though, selection criteria will not be this clear-cut. Editors of select editions often supplement the texts that they print in full with calendars of unprinted materials, giving brief abstracts of the contents of such documents and the location of each original. Editors of editions that print only documents created by the central figure of their series work out systems for providing readers with knowledge of the documents that inspired the author’s response. Thus the editors of Ulysses S. Grant’s papers print in full only letters and documents written by Grant; the letters or other materials to which these manuscripts respond are summarized at length in the footnotes accompanying the documents. Still other projects combine the two methods. All, however, hope to provide a “comprehensive record” of their figures’ papers, if not a comprehensive edition of all the texts.

There will be times when a printed format cannot provide adequate and convenient access of this sort. The comprehensive-record technique led the reviewer of one volume of statesmen’s papers to remark despairingly that the calendar had become the text. If the printed edition draws on only one chronological period in its subject’s life, or if it focuses on but a single area of concern in his or her correspondence and papers, the published volumes and the footnotes to the items chosen for print publication cannot serve as the clear “windows” for which editors like Wayne Cutler of the James K. Polk series strive. Editors should remember the needs of an edition’s audience in weighing the virtues of these alternatives to comprehensive publication. Jean Berlin has suggested, for instance, that calendars of letters omitted and letters not found are better suited to the interests of a scholarly audience, while explanatory introductory notes may better communicate similar information to a more general audience (“Selecting the Essential Webster,” 29). When confronted with such problems, the editor may want to consider the publication of separate finding aids or facsimile supplements to the printed volumes.

A. Image Supplements

The technical aspects of using microforms or digital images of nonedited source texts are discussed above, pages 40–41. The methods of traditional microform facsimiles have been easily translated into digital technology for use with either kind of facsimile supplement. Scanned images of original materials or their photoreproductions will provide the facsimile reproductions, while the automated database generates the electronic equivalent of targets, or editorially supplied labels reproducing control file information for individual items, to precede each item’s images and provide the base for the edition’s indexing system.

The editor of a series of scanned images has an enormous advantage here: digital images can be entered into the electronic files as they are collected—to be reordered by the user’s commands, not by the editor’s decisions. When preparing a microform edition, however, editors cannot begin the work of photographing images until collecting is completed and the physical results of that process are put into the order in which they will appear in the microform.

Editors who create “electronic microfilm” must remember that there are currently no standards for digitally scanned images as an archival medium beyond agreement that the TIFF format should be used. Quite simply, there isn’t yet a sure and certain way to provide safe storage for a substantial length of time of the data those TIFF files contain. Thus, an electronic facsimile edition requires both periodic “refreshment” of the images to ensure quality and the presence of a nonelectronic version of the facsimiles for archival purposes.

Microfilm’s continuing role as an archival medium is not the only reason editors must continue to be familiar with this older technology. Microform publication will remain a valuable option for image editions that cannot find a home on an established Web site where their texts and notes can be assured long-term maintenance. Editors in this situation who wish to share the results of their search for documents in image form can publish microfilms along with a copy of the editorial database that provides easy access to the reels’ contents. Should an appropriate Web site for the project become available in the future, the films can be scanned, and the database records can be linked to those online image files.

B. Cautions on Selectivity

Even with facsimile editions, electronic publication, microform technology and digital imagery, and computer-generated checklists or other finding aids to the editorial collection as a whole, the editor is responsible for making an intelligent and fair selection of materials for print. Indexed bound volumes or their online equivalents with easily read text and conveniently placed editorial notes will be used more often and more thoroughly than the untranscribed, unedited supplements. Even the editor whose annotation carries generous quotations from unpublished materials or pointed references to documents in a supplementary facsimile must remember: scholars are as lazy as other mortals. Most are accustomed to the convenience of a book format, and many will avoid reference to unprinted materials or novel electronic formats at all costs. As Jerome McGann said ruefully of himself and his colleagues as they prepared the online Rossetti Archive, “Our imaginations remain dominated by the book paradigm” (“Hideous Progeny, Rough Beasts: Editing as a Theoretical Pursuit,” 12).

II. Organization

Selectivity versus comprehensiveness is not the only choice editors face in determining the formats of their printed volumes or facsimile editions. Nor is that the only preliminary editorial decision that affects a user’s experience and interpretation of the edition: the arrangement of the documents selected for print can be equally important.

One intellectual ancestor of modern American papers editions, the Correspondence of Horace Walpole, grouped Walpole’s letters by his major correspondents. Thus, Walpole’s exchanges with the Reverend William Cole are contained in volumes 1 and 2, his exchanges with Madame du Deffand and her circle appear in volumes 3–8, and so forth. This feature of the Walpole edition is one that American editors—of all traditions—have generally chosen not to imitate. In almost every other modern scholarly edition of correspondence, all items selected for full publication are usually printed in a single chronological sequence.

The advent of electronic editions provides more flexibility. The user of an online Walpole edition who is interested only in the exchange between Walpole and Cole could isolate that correspondence through a search engine and then download it, print it out, and search within those results. Readers who need to read all of Walpole’s letters for a given year could rearrange the files appropriately for that end. Even in the electronic world, however, a selective edition must have rational criteria for the choice of documents, and there must be finite limits to the ways the reader can search and order the edition. In defining these limits, the electronic editor will find that many lessons of print editors are still relevant.

A. Topical Arrangement

When it comes to documents recording a subject’s professional life—whether as an author or a lawyer or a diplomat—documentary editors are likely to create special groupings of related materials for their editions. Legal papers generated by American statesmen traditionally appeared in separate series in both the microform and book editions of the statesman’s papers. The editors of an author’s literary works have little choice but to group together the draft versions of a specific novel, essay, or poem if they are to give their readers any sense of the evolution of an author’s thoughts, but editors in other fields have used topical organization as well. In the Jefferson Papers, Julian Boyd inaugurated a similar practice of grouping documents relating to specific foreign policy issues into special topical sections. The editors of The Papers of Daniel Webster provided their edition with separate subseries for Webster’s career as secretary of state and for his law practice. Within both these subseries, documents are also grouped topically.

Some groups of documents defy a unifying chronological organization. The records of a lawyer’s practice, for instance, would be unintelligible were all legal documents presented in one time sequence. Pleas, depositions, subpoenas, and affidavits that relate to a given case must be grouped together if their contents are to make sense. The principle of organization by case has been adopted consistently by most modern editors of legal materials in both print and facsimile editions (see editions of Adams, Burr, Hamilton, Lincoln, and Webster).

Often the degree of selectivity and topical organization is so high that the resulting publication is titled a “documentary history” rather than the “papers” or “records” of a group. Perhaps the best-known example of this method is Freedom: A Documentary History of Emancipation, which combines topical and chronological arrangement to enable its readers to sift through a vast bureaucratic archive to uncover a wealth of social history. (For a chronicle of the Freedom project and analysis of its organization, see Ira Berlin et al., “Writing Freedom’s History.”) One reviewer aptly termed the series “a history based upon documents” (Brook D. Simpson, “Blacks in Blue and the Fight for Freedom,” 7). Similar considerations shaped the Documentary History of the Ratification of the Constitution, the Documentary History of the First Federal Congress, and the Documentary History of the First Federal Elections.

An edition may also mix chronological and topical organization. The Sanger editors, for instance, decided to segregate documents relating to Sanger’s international work in order to highlight these activities and to bring more continuity to the documents and their notes than they would have had had the editors maintained a strict chronological arrangement. In the first three volumes, international events are mentioned and discussed briefly; the discussion is expanded in the fourth volume, where these documents appear in their own chronological sequence. This was a conscious decision made to correct gaps in the coverage of Sanger’s work and to provide a more cohesive story of her legacy.

Such departures from conventional chronological organization, like departures from any other general editing principle, should be made only after serious thought. This rule is even more important when an editor considers changing the principles of organization or selection after an edition is under way. The editors of the John Marshall Papers recognized the necessity for such a deviation. Selected records of cases that Marshall tried as a practicing attorney appear in volume 5, Selected Law Cases, with papers for each case grouped together and records for related cases in separate sections. Marshall’s opinions as chief justice of the Supreme Court, however, appear in their appropriate places in the chronological sequence of the Marshall Papers, interspersed with personal and professional correspondence, business papers, and other documents.

Records of a given legal case only make sense together; the earliest papers relating to an action may precede the suit by a hundred years, and the final settlement may come decades after the death of the parties in the original action. A judge’s decision, however, can be logically and honestly tied to a narrow period of time, to the days, weeks, or months in which the jurist considered the matter in question. The John Marshall edition simply followed in the honored tradition of documentary editing of allowing individual documents and groups of documents and the needs of the reading audience to dictate organization.

Here, as in any editorial decision, intelligibility and utility were overriding considerations. The needs of an audience for a documentary history are just as exacting as those for a near-comprehensive, chronologically ordered series. A perceptive discussion of this problem appears in Randall M. Miller’s “Documentary Editing and Black History: A Few Observations and Suggestions.”

B. “Chapter” Organization

An early example of the “chapter” format for printed annotated documentary texts was the first volume of The Papers of John Jay in 1975. It proved useful for other editions of eighteenth- and early nineteenth-century materials, including Dear Papa, Dear Charley, an edition of the correspondence of Charles Carroll of Carrollton and his father, Charles Carroll of Annapolis; and The Selected Letters of Dolley Payne Madison. In this format, the documentary texts are generally arranged in chronological sequence. Groups of documents for a given time period are presented within a discrete chapter, customarily opened by an editorial introduction summarizing the activities of the edition’s subject during that time. Additional editorial headnotes may appear throughout the chapters to supply information that could not be effectively conveyed by footnotes.

This organizational method, however, is seen even more frequently in selective editions of nineteenth- and twentieth-century materials, such as the Edison, Eisenhower, George Marshall, Margaret Sanger, and Jane Addams editions. Chapters in these series may be topical as well as chronological. The Sanger editors also made a conscious decision to limit dramatically the number of speeches and articles that would appear in their edited volumes. Most of these are available elsewhere (although in scattered form), and they will all appear in an electronic supplement. To give their volumes a clearer theme, the Sanger editors chose to publish more of Sanger’s shorter personal letters, which provide insights that are otherwise unavailable.

Effective as these editions are, they are not easy or simple productions. Notes in such “chapter” editions should be terse, and their authors must guard against the temptation to summarize and quote the documents themselves. Even in a selective edition, the documents must tell the story. Editors of chapter editions can testify to the delicate balance between being their subjects’ editors and crossing over into the realm of biographers. If the texts and their footnotes do not present a coherent narrative, the editor may need to reconsider the justification for publishing these sources as a documentary edition.

C. Cautions on Topical Groupings

All editors realize that haphazard topical arrangement can make it more difficult for the reader to reconstruct the patterns that will meet his or her needs. Electronic publishing makes such personal reorganization of an archive far simpler, but print editions and microforms do not give users this luxury. Many traditional forms of editorial apparatus, such as name and subject indexes, allow scholars to make their own arrangement of related materials for personal research purposes. Still, the casual destruction of the general organizational pattern can make the volumes less, not more, useful.

The demands of dealing simultaneously with the impatience of sponsoring institutions, varied groups of documents, and the expectations of several audiences for the same edition can create organizational systems that challenge the reader. The Daniel Webster edition’s annotated printed volumes of selected documents appeared in four distinct series: Correspondence, Legal Papers, Diplomatic Papers, and Speeches and Formal Writings. Two of the series included comprehensive calendars while the other two (Diplomatic and Legal) did not. Comprehensive microfilm supplements for three of the four series preceded publication of the first volumes, while the microform of Webster’s legal papers came much later. One reviewer complained wearily: “The lack of any lengthy introduction to editorial techniques and choices, particularly in describing the relationship of the various letterpress series to the microfilm, is confusing; some reviewers of various volumes have written inaccurate accounts of this connection. Will the student or layman understand it better?” (Jean V. Berlin, “Selecting the Essential Webster,” 26).

Just as important, editorial arrangement can distort the documents as effectively as a biased choice of documents, a fact discussed at length by American editors in the history of science. The Joseph Henry, Thomas Edison, and Albert Einstein editions consciously broke with the European tradition of publishing scientific documents in topical groupings. Modern intellectual historians, whether their interest be science, philosophy, or the arts, realize that such a prepackaging of source materials makes it virtually impossible for the reader to understand the context within which each step of the intellectual process occurred. History, of whatever subject, places a premium on time as an organizing principle.

A peculiar set of circumstances led the John Dewey edition to embark on a chronologically arranged series. When the project began its work, Dewey’s widow was still alive, and she proved reluctant to cooperate. Thus, despite suggestions from some Dewey scholars that his writings appear in some sort of “logical” series, the editors had no choice but to begin publication with his earliest published writings, materials for which copyright protection had expired. Even when Mrs. Dewey had died and more cooperative executors took control, the Dewey edition remained true to the chronological organization, which proved more helpful to their goals (Jo Ann Boydston, “The Press and the Project: A Study in Cooperation”).

D. Electronic Aids

Aside from conveniences of formatting and hypertext links, computer technology can be an important element in planning a selective edition. The editors of the papers of Thomas Edison admit that only computerized methods enabled them to analyze accurately the intellectual themes and technological problems in given periods of Edison’s long life so that these could be reflected in their published volumes. Here, as in other areas of scholarly editing, the use of computer equipment does not change the rules—it merely makes playing the game a bit easier. Whatever electronic aids are available, whatever external or internal pressures determine the organizational format and scope of a series, the editor must remember that the arrangement of its contents may determine how widely the edition is used—or whether it can be used at all.

E. Some Conventions of Documentary Organization

Modern documentary editors have adopted certain conventions for the subdivisions of the papers or writings of their subjects. All CEAA and CSE editions, for instance, publish works composed for publication separately from private writings such as letters and journals, a method employed in the Albert Einstein edition. One chronological series of materials appears in the Early Years series, but for Einstein’s mature life, surviving records of correspondence, laboratory notes, and other papers are segregated from the series of Writings, works of Einstein that actually saw print publication. No one quarrels with this division in the case of authors whose works were designed for a wide audience quite distinct from the one to which the letters and journals were addressed. Each group of documents must be studied separately. But further subdivision should be undertaken only for compelling reasons, and any editor who publishes documents in separate series or subseries must be prepared to devise different, though equally clear and cogent, standards of selection for each group.

1. Diaries and Journals

A separate series for diaries presupposes that their contents are so full and interesting that readers will need to study them as an independent work. Some diaries are little more than appointment books that do not stand alone; their only purpose is to illuminate the fuller exposition of the diarist’s life as represented in his or her correspondence and other papers. If only scattered leaves of a diarist’s memoirs survive, it would also be foolish to allot them a separate series. Thus, the editions of James Fenimore Cooper’s letters and Robert Morris’s records of the Revolutionary Finance Office integrate surviving scattered journal entries into the chronological series of correspondence for the same period. Whenever the writer is an intermittent record keeper, making entries only during extended trips, such travel journals may also benefit from incorporation into the general series of his or her writings.

For some diarists, a personal journal represents a life quite distinct from that reflected in correspondence and other records. Many professional authors use journals less as a record of daily events than as a literary daybook in which they jot down ideas for stories or essays and even use some pages to draft their skeletal inspirations into fully developed literary passages. Were daily entries from such journals interspersed with authors’ correspondence for the same period, the editors would destroy the sources’ intellectual integrity to no useful purpose. Readers would have to reconstruct the sequence of journal entries. Modern typesetting and photoreproduction technology allows editors to find creative solutions to special problems created by journals of this kind. An appendix to volume 3 of the Albert Einstein Papers includes parallel publication of the pages of Einstein’s scribbled “Scratch” notebook with editorial transcriptions of the notebooks’ contents. Electronic publication can provide even more options.

There may be other considerations behind a decision to publish journals or diaries separately. The repository owning the originals may withhold permission for publication unless the new edition preserves the originals’ integrity. Editors under pressure to begin publication early may look to a series of easily located journals and diaries that can be prepared for publication while the project completes its search for more elusive correspondence and other papers. And there may be, quite simply, some overwhelming public expectation of separate publication of a set of diaries.

2. “Series” within Series

The founding generation of the Adams Papers editors believed that having a single chronological series of Adams family papers, public and private, would be less useful than having an edition with carefully designated divisions. (The problem of grouping documents is addressed at length in the introductions to the Adams Family Correspondence and The Papers of John Adams.) They assumed that comparatively few readers would approach the volumes with the intention of studying the Adamses as a family; rather, readers would have an interest either in personal interrelationships among family members or in some individual family member’s professional activities in law, literature, or public service. Thus, the Adams edition appears in three series: Diaries and Autobiographical Writings, Family Correspondence, and Papers, the last including letters and other writings bearing on the Adamses’ professional and public careers.

Although volumes of the Adams Papers now routinely carry lists of excluded Adams materials and the editors scrupulously disclose statistics on selection, recent reviewers have challenged some of their basic organizational assumptions. As print publication takes John Adams into public life and diplomatic service, some have become uneasy with the distinction between “public” and “family” letters, a division of his correspondence that is artificial in some respects. Historians of culture and the family argue that modern scholarship has created a large body of readers who are, indeed, interested in the papers of all the Adamses, male and female, public and private, and these readers are ill served by the edition’s divisions.

Practical considerations of staff expertise led other editions to publish a number of separate chronological series. The George Washington Papers, for instance, are divided among Colonial, Revolutionary, Confederation, Presidential, and Retirement series. Work on later series began even before the final volume of Colonial papers went to the printer, and two editorial teams now work simultaneously on the remaining pair of series. Documents within each are arranged chronologically, and when all the series are completed, one chronological run will exist. A similar method was adopted by the James Madison edition, with teams of editors working ahead on series for Madison’s career as secretary of state and president while another group completed work on the records of his last years as congressman. Yet a third series, this one for Madison’s retirement period, has begun. The Jefferson Papers broke its chronological sequence in two several years ago by creating a retirement period series sponsored by a second institution and located in an entirely different state.

A somewhat different example of “subseries” is represented by Series II of the Jefferson Papers. This group of volumes is devoted to very specialized groups of Jefferson texts that stand by themselves, independent of the chronological series of correspondence and public papers. They’ve included Jefferson’s Extracts from the Gospels, Jefferson’s Literary Commonplace Book, Jefferson’s Memorandum Books, Jefferson’s Parliamentary Writings, and Thomas Jefferson’s Farm Book. Placing these assignments outside the work of the permanent Jefferson editorial staff made it possible to enlist experts for comparatively short-term editorial tasks.

III. Contextual Editorial Contributions

The texts in most documentary editions will not “tell their own story.” Some editorial contributions in the form of annotation, bibliographical or geographical supplements, glossaries, and indexing will be needed. “Notes” of one kind or another are usually the first to be done.

A. Source or Provenance Notes

Once selections of documents have been made, once source texts have been transcribed and editors have established reliable texts based on those transcriptions, there is one form of editorial annotation that cannot be avoided: a statement of the location of the original material on which the editorial text is based. This source note or provenance note is the one form of annotation required of every documentary edition. If the edition uses textual symbols, there may be no notes that describe details in the manuscript sources. If the edition forgoes explanations of allusions in the source, there may be no historical (or informational) footnotes. But the editor of every documentary edition, scholarly or “general,” must tell readers where to find those sources and must describe the physical form of those sources.

Often the first words of their own that editors contribute to their editions are those that describe the physical realities of the sources that they edit. While a junior member of the staff may transcribe the words of a letter’s address leaf or a government document’s endorsement, it is the senior editor’s responsibility to design the format in which such information will appear.

1. Formats

Data on source location and provenance may appear in numbered or unnumbered notes adjacent to the document to which they refer or in a back-of-book textual record. The function of these notes can be as varied as their locations. Some editors exploit the source note as a convenient site for any historical annotation that refers to the document as a whole. For instance, if an individual’s name first appears in the edition as a letter’s author or recipient, the source note may be the place where that person is identified. Similarly, source notes sometimes include explanations of editorial decisions that affect the document as a whole: assignment of a date to an undated source, attribution of authorship to an unsigned document, or identification of the unnamed recipient of a letter. Other editors prefer to keep source notes free of such editorial comment. In their editions, a numbered footnote keyed to the heading’s date might explain the reasons for assigning a day, month, or year to the document, while a footnote keyed to the document’s title might explain the methods used to determine the item’s author or recipient.

An edition of diaries or letterbooks may be drawn from comparatively few physically distinct sources, and the location and physical details of such sources are often detailed in an introductory note to the volume or in the back-of-book textual apparatus. Additional notes within the text cue the reader when the editorial text progresses from one source to the next. But the editor whose source texts are many and varied (such as letters recovered from a hundred manuscript repositories) must devote time to designing source notes. Customarily, such notes first state the physical form of the source.

For letters, the editor must indicate both the means of inscription (author’s autograph, handwritten copy, typewriting, printing) and the version of the letter inscribed in this manner (recipient’s copy, retained copy, draft, transcription). Many editors find it unnecessary to give both of these facts explicitly for each item. Instead, the introductory statement of editorial method can explain that all source texts for letters are addressees’ copies unless otherwise indicated, or it may announce that all letters described as recipients’ copies are in their authors’ hand unless otherwise noted.

It’s often necessary to adopt a series of codes indicating methods of inscription. In such systems, “A” usually represents autograph inscription, “L” letter, and “S” signed by its author. Thus, “ALS” would stand for a letter written and signed by its author. Items inscribed after the mid-nineteenth century demand a new series of symbols to indicate typewritten (usually “T”), carbon copy, and even telegram. Variations on these symbols are countless. For helpful samples, would-be editors should consult the lists of such codes in editions whose sources resemble their own in both time of inscription and archival nature. An authoritative description of these codes may be found in “Appendix: Abbreviations Used by Manuscript Collectors, Curators, and Dealers,” in Lewis J. Bellardo and Lynn Lady Bellardo’s A Glossary for Archivists, Manuscript Curators, and Records Managers. Stevens and Burg’s Editing Historical Documents is invaluable for demonstrating the systems of different editions.

The descriptive symbol for the source’s documentary version is usually followed immediately by the name of the owner institution of the original source text or by an alphabetical symbol for the repository. Most projects rely on modifications of the MARC Code List for Organizations’ symbols for repositories, the same ones used in cataloging materials for the project. Many documents require identification of the source text’s collection, not merely the name of its source repository. Not only does this aid the reader in locating the original for comparison, but it can also furnish valuable information about the document’s provenance. When the collection is arranged in a single consistent order (chronologically or alphabetically by author), no further details are needed. If the collection’s arrangement is in various series or is otherwise erratic, it may be necessary to provide numbers for the documents’ volumes, boxes, folios, or record groups. With a computerized editorial system, this information can be imported from fields in the control file when the transcription is prepared. An electronic edition provides the convenience of immediate expansion of any codes and abbreviations.

The source’s version and location are usually followed by certain details of inscription that are not part of the body of the document. These should be carefully described or printed verbatim in the order of their inscription. Thus, for a letter, the editor first indicates its address, then its postal markings, then any endorsement by its recipient, and finally any significant dockets by clerks or later owners of the original manuscript. The editors of the Mark Twain Letters took advantage of the capabilities of computer composition by using the easily recognized icon for an envelope (empty parentheses) to introduce any text drawn from such a source.

Formats and even locations of source notes vary widely. Literary editors who use extensive back-of-book records sometimes choose this area for the location of source information and even nontextual annotation, as well as for the conventional reports of editorial emendations and hyphenation lists. In the Howells Letters and the Frederic Correspondence, the source note is defined as part of the textual apparatus consigned to the back of the book, although informational footnotes follow each document. The Mark Twain Letters edition presents an abbreviated source note immediately preceding the text of each letter, giving nothing more than a code for the owner institution of a manuscript or a short title for a letter’s printed source. The back-of-book textual record carries fuller information.

Some editors use the source note to list enclosures in the original document when these are not printed in the edition. Other editors describe enclosures in footnotes keyed to references to the items in the body of the documentary text. The latter method fails, however, if the edition’s central figure or his or her associates frequently enclosed newspaper clippings, promissory notes, or the like, without mentioning them in the covering letter or report. The editor then has no choice but to expand the source note to list these items. When the enclosures in a given document can be identified but have not survived with their covering letter or report, the editor refers readers to a convenient and reliable source for another copy of these vanished items and summarizes their contents when pertinent. The editorial comment “enclosure not identified” is a perfectly honorable form that spares readers the trouble of hunting for information the edition does not give.

Whatever the form of the source note or its location in the volume, its function, like that of all editorial apparatus, should be consistent throughout the edition. Readers must know that they will always find information on the source’s manuscript version and location in (a) an unnumbered source note preceding or following the document or its historical annotation, (b) the first numbered footnote appended to the document, or (c) the first element of information in a back-of-book textual apparatus for the item. The information in these notes must be presented uniformly, using the same codes and the same sequence of information each time. There is no room for creativity in this part of the editorial apparatus, as variation will only confuse the reader.

B. Contextual Annotation

Most modern documentary editors offer more than accurate transcriptions and notes describing the location and provenance of their source texts. They supply, as well, editorial notes and other devices that clarify wording, references to historic figures and events, or images whose meaning might otherwise elude the modern reader. Even while establishing the documentary texts that form the core of any edition, editors must consider the documents’ need for editorial explanatory or informational annotation, glossaries or gazetteers, back-of-book records, and Web sites, and even the form of the index that provides the ultimate access to the contents of texts and notes alike. As they struggle to maintain the integrity of the editorial texts, they must weigh the advantages of adding editorial comments on context, fact, and allusions to elucidate those texts.

Often the texts themselves, the textual apparatus, and the notes on provenance supply a substantial amount of historical information. Providing legible versions of faded manuscripts, deciphering inscriptions the author tried to obliterate, tracing the history of a manuscript’s or pamphlet’s physical location—all these are scholarly functions that add to the human store of factual knowledge. But documentary editors are usually obliged to go further. Here they venture into one of scholarly editing’s most controversial realms.

In providing information of this kind, editors strive to keep in mind both the author’s meaning and the needs of two audiences: (1) the group or individual to whom the document was originally directed, and (2) the readers who will use the new annotated edition. Part of this responsibility is discharged by providing readable texts of illegible manuscripts or obscure, hard-to-locate pamphlets that give modern readers much the same physical access to these documents’ verbal contents as that enjoyed by their original readers. But this is often not enough. Modern readers almost invariably need additional facts or explanations to understand those words or images as they were intended by their creators and as comprehended by their original readers.

C. Theories and Rationales of Annotation

Many editors have attempted theoretical statements of the goals of annotated editions, hoping that general guidelines would protect future editors from repeating past errors. In 1963 Lester Cappon claimed that proper informational annotation was part of the very “rationale” of modern documentary editions, but that rationale is still not sharply defined.

Among literary scholars, Arthur Friedman suggested his useful “Principles of Historical Annotation in Critical Editions of Modern Texts” more than six decades ago. He focused on the needs of editions of published works, not the private writings of literary figures, concentrating on “notes of explanation” that “attempt to make a work more intelligible by showing its relationship to earlier works,” not those “notes of recovery” that would “supply information that would presumably have been known to the author’s contemporaries, but that has been lost by the passage of time” (118). More than thirty years later, another literary scholar, Martin Battestin, raised points of more general relevance in “A Rationale of Literary Annotation: The Examples of Fielding’s Novels.” Battestin conceded that “no single rationale of literary annotation” would find acceptance. Instead, he suggested three variables that affect the extent and methods of annotation: first, “the character of the audience which the annotator supposes he is addressing”; second, “the nature of the text he is annotating”; and third, and admittedly the most unpredictable, “the peculiar interests, competences, and assumptions of the annotator himself” (60).

The specific problems of documentary editing by and for historians were addressed by Charles Cullen in “Principles of Annotation in Editing Historical Documents; or, How to Avoid Breaking the Butterfly on the Wheel of Scholarship.” Cullen held that the “proper scope” for annotation is determined “first, last, and always [by] the subject of the volume or series.” By subject he meant both the form of the source text and the identity of the person or group that is the series’ focus. Annotation to a group of diaries can be confined to supplying information that would have been known to their author—no other contemporary audience need be considered. Considerably more information on public events might be provided in an edition of correspondence between political leaders, for readers need the facts that will enable them to gauge the response of each letter’s recipient, as well as the motives of its author. The correspondence between a writer and contemporary readers with expert knowledge in a given field demands fairly technical annotation to present accurately the facts and data that would have influenced both.

Much-needed humor was introduced to the debate in Nathan Reingold’s 1987 review essay of the Darwin edition. Reingold offered a charming, half-serious analysis of the three schools of annotation adopted by modern American scholars. These were “the Unwinding Scroll in which bits of the past are disclosed in chronological order while the editors pretend (not always successfully) not to know what the scroll will later contain”; second, “the Omniscient Eye,” whose followers “see all, know all, and often cannot resist telling all”; and last, “the Electron Microscope school,” which comes into play “when literary and historical editors minutely scrutinize physical objects and the signs affixed to them for levels of meanings mere reading of a manuscript cannot disclose (nor sustain in some cases)” (“The Darwin Industry Encounters Tanselle and Bowers,” 17). Had Nate Reingold lived to see Web-based annotation, we would have been treated to equally insightful and witty analyses of online methods.

D. Special Sources, Special Needs

Many of American documentary editing’s traditional conventions of informational annotation assume that an edition will contain the written words of well-educated politically or socially prominent men. Policies of informational annotation, like textual methods and criteria for selection and organization, can be skewed by such assumptions. In providing informational comment, the editor should attend to the needs of the writers and sources being edited instead of relying on guidelines appropriate to the kind of sources commonly edited fifty years ago.

While editors agree that an author’s sex or degree of literacy seldom affects textual policies, this is not true where informational annotation is concerned. For example, editors should ask whether their authors’ sex affects the clarification their texts may demand. Drawing on her own experience in “Gender Consciousness in Editing: The Diary of Elizabeth Drinker,” Elaine Forman Crane pointed out that what was significant in the life of an eighteenth-century woman differed markedly from what would have been important in the life of her male contemporaries. An early (male) editor of Drinker’s diary abridged its entries by excluding “strictly private matters,” precisely what historians of women and the family nowadays consider matters of importance. Crane’s edition of Drinker’s diary included not only a faithful transcription of the complete journals but also informational apparatus that identified the people who figured in her life, whether or not they met the criteria of political or social “importance” in their own day. Beyond this, the editors had to clarify Drinker’s ladylike euphemisms for physical functions, references of special significance in the life of a mother who, “like most eighteenth-century women, was a primary caregiver, and [for whom] health was a constant concern.” Questions of pregnancy, the availability of medicines, and rumors of epidemics may have been of passing interest to Philadelphia businessmen of the day, but were matters of central concern for wives and mothers like Elizabeth Drinker. For an interesting group of essays in this area, see Ann M. Hutchison, ed., Editing Women.

The assumption that source texts will contain words and words alone should also be reexamined. Historians of science were the first to confront the problems presented by images as documents that demanded analysis and annotation as urgently as verbal sources. Reese Jenkins discussed this issue in “Words, Images, Artifacts and Sound: Documents for the History of Technology.” Documents generated by architects and artists soon joined volumes of edited sources. While other series of the Latrobe edition could treat pertinent drawings as “embellishments” or illustrations, images themselves were the documents published in the three volumes of his Engineering Drawings, Architectural Drawings: Latrobe’s View of America. (For an analysis of that series’ editorial methods, see John C. Van Horne, “Drawing to a Close: The Papers of Benjamin Henry Latrobe.”)

Nowadays, digital imaging and the potential of Internet publication present far wider scope for the publication of nonverbal elements of texts. Two of the best examples of the potential of electronic publication of such materials are the online Dante Gabriel Rossetti and William Blake archives: http://www.rossettiarchive.org/, and http://www.blakearchive.org/blake/. While the works presented on these Web sites include the poetry and other verbal works of Rossetti and Blake, their special value is their ability to provide users with the imagery of these two artist’s drawings and paintings, with verbal and nonverbal sources linked. Jerome McGann, founder of the Rossetti Archive, remarked that in creating an edition of the works of men like Rossetti and Blake, “medium and message are so intimately wedded as to be inseparable” (“Hideous Progeny, Rough Beasts,” 9). Although modern technology makes such editions possible, it does not make them easy. McGann warns scholars who will follow the tradition he and his colleagues created that “when technical design issues arose—even in very minor matters—we were regularly forced to reconsider the most basic premises of the project” (12).

E. The Needs of the Modern Audience

Any scheme of informational annotation or commentary should be designed with the interests of its specific audience in mind. The most basic question in this regard is whether the edition will address scholars or laypeople. Scholars who consult documentary editions are more likely to bring a knowledge of the people, topics, or historical era on which the documents focus, but the editor must bear in mind the kind of knowledge each group of expert readers will possess. Military historians are more likely to read the Papers of Nathanael Greene than the William Dean Howells Letters. Thus, they need few reminders of the significance of major battles or minor skirmishes of the American Revolution, whereas literary critics reading the Howells Letters might appreciate an additional sentence or two from the editor pointing out the date and outcome of a specific Spanish-American War engagement mentioned by Howells.

In addition, scholars are more likely than general readers to have access to pertinent reference works, and the amount of information they need from the editor may be comparatively small. An edition designed for a general audience often provides notes and comments that create a more independent, self-contained body of knowledge. The editors cannot expect an audience of laypeople to consult dozens of hard-to-find supplementary reference tools any more than they would leave them to puzzle over an obsolete term that would be easily understood by an expert in the field. Offering too little informational annotation can limit an edition’s audience. By assuming that only experts in a certain field will consult a volume and offering light annotation, an editor may fail to open the edited documents to a wider audience. (See Lillian Miller’s discussion of this problem in her review of The Correspondence of Benjamin Henry Latrobe.)

F. A Practical Framework for Informational Commentary: What Editorial Apparatus Should Explain

Veterans of the process of supplying informational annotation and its supplements are more reluctant than mere observers of the process to suggest what any other editor should explain. Still, they agree that two immutable laws can help narrow the range of what requires annotation:

1. Never explain what the documentary text itself makes clear.
2. Never forget that your edition will have some form of index.

Even an electronic edition will provide something beyond a simple keyword search, whether that something is an index using traditional “controlled vocabulary” for set subject headings or a search engine that enables users to search useful combinations of different fields such as author and date. For an edition without computer-based search capability, the classic index remains the most effective tool known for clarifying and streamlining editorial commentary. Self-evident as these rules seem, they, like all the others that follow, have been broken on many occasions.

Editors may provide informational commentary in a variety of formats: footnotes, headnotes, bracketed interpolations in the text, back-of-book notes keyed to page-line references, and nonannotational supplements such as glossaries, genealogies, and biographical directories. A Web-based edition or a print edition with Web supplements can provide the same variety of conveniences. Whatever combination is chosen from the ones described in this chapter or invented on the spot, the sum of their parts, together with the facts imparted by the documents themselves, should provide the reader with some basic level of information. That information usually falls into the following categories:

1. Information Known to the Document’s Author or Creator

(a) Identity of persons mentioned in the documents when some form of biographical knowledge is required to understand the text. Identification should focus on the person’s role at the time of the document’s creation.

(b) The substance of documents not printed in the edition that are essential to approximating authorial knowledge. Editions of letters written by famous authors, whenever possible, provide synopses of the communication to which each is a reply. If enclosures are not routinely included in an edition, they should be listed and, when necessary to understanding the source, summarized for the reader’s convenience.

(c) Information about places or historical events referred to in obscure or obsolete language or necessary to a modern reader’s understanding of the document. The editor should resist the temptation either to write lengthy accounts of such events or to translate into modern terms every archaic phrase used by an author. Readers of seventeenth-century materials usually have some knowledge of seventeenth-century history and usage. The document’s context will dictate whether the editor need go beyond a one-phrase clarification of a term to a paragraph of narrative prose.

(d) Bibliographical data on published works mentioned prominently in the document, especially when these have influenced the author or are cited by the writer as sources. If the author has provided a comprehensible citation, nothing more is required of the editor.

2. Information Available to the Document’s Original Audience

In many cases, the facts presented to enable the reader to share the author’s knowledge will satisfy this requirement as well. Still there are instances when more editorial comment is needed.

(a) Whether or not a letter or other transmitted document arrived at its intended destination. When endorsements or dockets on the source provide this information, and their texts appear in the provenance note, nothing more need be done. If the facts of transmission are not self-evident, editorial notes should provide what is known from other sources.

(b) The immediate response to the document. This would include action prompted by a public communication of a government official, reaction to a lecture or professional paper, or reply to a personal letter. Selective editions that focus on the writings of a specific figure generally provide more of these summaries and synopses than editions that publish both sides of a correspondence or the broadly defined “papers” of a public figure.

(c) Any other contemporary result of the document or its transmission. Even the source’s present location may be significant here. For example, most editors become suspicious when a document is found in a collection where it does not logically belong. This often indicates that the document served a purpose other than the one its author originally intended and that it was forwarded or even reforwarded to another individual or group for action or review.

The reader will have noticed that these categories focus on factual knowledge. There is a third category that goes beyond the factual and must be approached with more caution.

3. Modern Knowledge of the Document’s Meaning and Significance

By and large, this area of comment and analysis is better left to users of the documentary edition. Responsible editors avoid this category of notes whenever possible, although special circumstances demand the modification of the most cautious policies. The editor’s annotational responsibility generally ends with making clear what the document meant when it became part of the historical record, and documentary editions are seldom the place for a lengthy analysis of later misconceptions of a document’s meaning. Daniel Feller reminded his fellow editors, “Documents endure; historical scholarship does not” (“ ‘What Good Are They Anyway?’: A User Looks at Documentary Editions”).

There are some generally acknowledged exceptions to this rule. One is a letter of recommendation for an office seeker—it is only fair to let the reader know whether the applicant succeeded. Correcting a misattribution or establishing a new attribution of authorship may also demand editorial attention to the history of a document after its creation and original dissemination. For instance, when the editors of the Adams Papers printed three documents written by John Adams as part of “The Constitutional Debate between Thomas Hutchinson and the House of Representatives” (1:309–46), the editorial comment on the first two documents was confined to their historical backgrounds. For the third, the editors also presented comments made by later writers on the essay’s place in the history of American Revolutionary political thought. The document required such special attention because the Adams editors were the first modern scholars to recognize that it was, in fact, the work of John Adams. Thus, they needed to place it in the broader context of Adams’s work and the role his writings played in the development of eighteenth-century political theory and practice.

G. Consistency and Clarity: The Annotator’s Watchwords

While it is difficult to tell someone exactly how to annotate documents, it is somewhat easier to refer them to examples of first-rate annotation. Thirty years after the publication of Lyman Butterfield’s Letters of Benjamin Rush, editors such as Harold Syrett still pointed to the volumes as a model for others to imitate, and readers of the edition seconded this view. In 1952 J. H. Powell remarked of the Rush volumes that their notes were an essential part of the presentation of the manuscripts. Henry Graff recalled, on picking up the volumes three decades later, “I still found that the texts and their notes took me into Benjamin Rush’s world.” Unfortunately, students cannot read Butterfield himself on the subject of annotation, for he never published any guidelines to his own practice.

Butterfield and other exemplary annotators achieve their ends through consistency and clarity. Consistency entails establishing a recognizable pattern of annotation that readers can grasp quickly. If the editor’s introduction promises readers that persons mentioned in the text will be identified at their first appearance, that promise must be kept, and readers can reasonably assume that if there is no footnote identification for an unfamiliar personal name, they need only refer to the index for the location of an earlier note.

Clarity is a more elusive goal than consistency, since it cannot be ensured by the most detailed style sheets or the most explicit statement of editorial method. Informational annotation should be as brief and as clear as possible. Many scholars have discovered to their surprise and chagrin that drafting such notes places the greatest strain imaginable on their literary craftsmanship. The ordeal has few rewards, for if notes are well written, they do not impress readers with their brilliance or wit; they merely supply a name or fact or date so that readers can get on with the business of reading the documentary texts.

As in all aspects of documentary editing, the editor’s most effective weapon is a firm sense of the edition’s source, medium of publication, and audience. At the most practical level, this means that, while preparing informational notes, the editor will constantly refer to the passage in the source. Staying focused in this way can eliminate hours of hunting for information that is not germane to the words or phrases in question or of drafting paragraphs irrelevant to understanding the troublesome passage.

Bearing in mind the edition’s planned publication format saves similar hours of wasted effort. Knowing that the final edition will provide access to glossaries or genealogies liberates the editor from working into the footnotes definitions or lengthy descriptions of family relationships. An editor with a firm sense of the edition’s opportunities for maps or other illustrations can also streamline footnote texts. Electronic publication sharply reduces the need for verbal cues for cross-references—a standard symbol like an asterisk or highlighted text can cue users to links to needed information elsewhere in the edition.

Even when the formal “edition” is print-based, electronic tools can perform many important functions. Editors of the first volume of the Joseph Smith Papers have shown exceptional imagination in creating Web-based supplements to notes that appear in the volumes themselves. The project’s Web site will provide biographical information for persons mentioned in the edition’s volumes, an atlas of maps of the areas through which Smith led his followers in their journey west, and other elements that make the site something close to an online reference library for students of the early history of the Latter Day Saints.

Of course, a clear sense of the edition’s audience is the most important key to fulfilling the mission of consistency and clarity. Readers who consult edited documents—whether in printed volumes, microforms, or electronic editions—do so because they are motivated by some preexisting interest in the subject. These readers are remarkably adaptable creatures who soon detect annotational patterns for themselves. After a few pages, for instance, the reader will sense that the editor has chosen to identify names mentioned in a series in a single note following that series rather than in several notes keyed to each item in the group. If this pattern is maintained throughout the edition, readers adjust their rhythm of reading accordingly. If the pattern changes without warning or reason, readers will be distracted from the text while vainly attempting to figure out what has happened to the scheme to which they were accustomed.

H. Overannotation: The Editor’s Nemesis

Some of the most acute and witty contributions to the literature of scholarly editing concern the dangers of self-indulgent overannotation. Documentary editions should be the beginning of research, not its culmination. Informally, many editors conclude that their duties are often discharged by showing their readers where to pursue further research on a given point rather than by pursuing every lead themselves in the footnotes of their editions. And there is no shame in an editor’s resorting to what Martin Battestin termed “those most pitiable of adverbs, ‘probably,’ and ‘possibly’; or declaring his utter helplessness in that still more humiliating phrase—‘Not identified’ ” (78). Excessive informational annotation is the “occupational disease of editors,” said James Thorpe (Principles of Textual Criticism, 201), and Battestin warned editors to fight off this contagious “impulse to tell all that they have learned rather than what readers need to know” (69).

Symptoms of the disease usually appear in one of two forms. In one, the editor simply furnishes too many footnotes, explaining matters that require no clarification or whose explication goes beyond the needs of any reader or any documentary text. Overly detailed factual annotation prompted criticism of the Jefferson Davis Papers and early numbers in the Madison Papers series. The other form of the plague was ascribed to Julian Boyd, the father of the modern tradition of historical editing. In later volumes of the Jefferson Papers, readers saw increasingly frequent use of a technique whereby Boyd gathered together topically related documents and prefaced them with long editorial introductions—monographs that had little to do with the texts at hand. Valuable as these essays were, some critics suggested that they might better have been published as separate works unrelated to the Jefferson edition.

Boyd’s successors at the Jefferson Papers openly disavowed the practice. The introduction to volume 22 of the edition explained that Charles Cullen and his associates had largely abandoned Boyd’s “file folder” method of topical organization and its corollary of lengthy editorial notes introducing each such segment. “When a file folder system is frequently employed,” Cullen wrote tactfully, “the reader must understand the Editor’s organizational logic in order to make full use of the edition. If editorial staffs change, so does the logic, and readers are increasingly ill-served” (vii).

Automated methods gave editors quantitative assistance in diagnosing their own self-indulgence in annotation. The editors of The Papers of George Catlett Marshall routinely monitor themselves by running a program that anticipates the amount of space that will be occupied by documentary texts and notes in volumes they are preparing. When they find that the ratio of notes to documents has passed an acceptable level, they know that they must review their own work and cut accordingly.

I. Annotational Forms

Choosing occasions for editorial comment and setting the length of such explanations are not the only annotational parameters the editor must consider. The virtues and vices displayed in editorial annotation and explanation appear in a wide variety of physical designs, and even the format of such notes on the printed page can impose interpretation on the documents or mystify readers rather than enlighten them. The wide choice of annotational designs available for inspection and comparison in American documentary editions sometimes seems a testimonial to their editors’ and designers’ artistic impulses. The sections in Burg and Stevens’s Editing Historical Documents entitled “Contextual/Informational Notes” and “Forms of Annotation” (162–97) provide an excellent assortment of formats and methods for all the varieties listed below.

1. Linked Notes

In traditional terms, these are footnotes on printed pages, tied to the piece of text to which they refer by superscript numbers or symbols. In electronic editions, they are notes linked electronically to a specific place in the on-screen text. American documentary editions feature a number of formats for informational notes of this kind. Here are some of the most common methods that have been used for editorial annotation that appears immediately adjacent to the documentary texts (as opposed to a back-of-book record) or is available through a hypertext link.

(a) A format that distinguishes between textual and informational notes. Julian Boyd introduced this convention in the Jefferson Papers, but the example has not been widely imitated. Each documentary text is followed by an unnumbered source note that indicates the source’s provenance, variant versions, and so on. This note is set in a typeface that distinguishes it from the double-columned notes that follow. The only numbered footnotes are those related to textual details. Informational annotation appears in a single, unnumbered note. For the reader’s convenience, the key words and phrases to which portions of this note refer are printed in small capitals.

This device works well only for short documents that require little editorial explanation; it is less satisfactory for longer ones that demand lengthy explication. As a purely practical matter, the reader of a fifteen-page printed text will be hard put to find references to specific words or phrases in five pages of double-columned notes at the close of that document. There are also philosophical objections to combining information that might have appeared in a dozen separate notes. Arranging such data in one note, printed in paragraphs with topic sentences, imposes an interpretive order on the information, and the editor is more likely to go too far in telling the reader what the editor believes the document means.

(b) A combination of an unnumbered source note with numbered notes covering both textual and contextual details. This format can be seen in the John Marshall and Franklin editions. In the Marshall volumes, the unnumbered source note follows the document’s text; in the Franklin volumes, it precedes the text. Although the source note may contain comments on general textual problems, such as the existence of mutilated passages, specific textual matters appear in the notes keyed to the text by superscript numbers.

(c) Bottom-of-page footnotes are generally more expensive in terms of production costs than end-of-item methods, but they are the format of choice for any edition in which the individual documentary texts consistently run more than ten or twelve pages and require frequent annotation. Clearly, it’s inconvenient for a reader to flip back and forth across several pages of text to locate an end-of-item note. (For an example of this system, see the Hamilton Papers.)

(d) End-of-item numbered footnotes. This is the most common convention among historical print editions. It appears in the Cooper edition, the Irving Letters, and the Mark Twain Letters. The source note or provenance note appears first, either as footnote number 1 or as an unnumbered note preceding the numbered footnotes. The source note provides general textual information, but explanations of specific textual problems are mingled with informational footnotes in one sequence of notes for each document. In such editions it is customary to provide bottom-of-page footnotes, rather than end-of-item notes, when annotating the text of an exceptionally long source, although the sequence of such numbered notes still begins and ends with each document.

(e) Headnotes for provenance combined with numbered footnotes. Such formats place provenance information in an unnumbered note that precedes the documentary text. Numbered footnotes provide further textual details, as well as historical explication. This was the original format of the Madison edition, but the changing nature of the edition’s source texts required the device’s modification in the eighth volume. More and more sources were located in variant versions or were found to require lengthy lists of enclosures or discussions of complicated patterns of transmittal or referral. These factors inflated the unnumbered headnote from a few lines to several unwieldy paragraphs.

2. Back-of-Book Annotation

Some editions of the writings of literary figures consign source notes to their back-of-book records, and the Hawthorne Notebooks and the Emerson Works place informational notes there as well, using page and line references instead of superscript numbers in the texts. The relegation of this information to the back of the book has serious consequences. Whenever the editorial texts’ primary function is to provide historical evidence, the provenance of the original must be readily available to every reader.

Back-of-book records of informational annotation serve the reader only when such notes are infrequent and inconsequential for grasping the meaning of the source text. This distinction was recognized by the Thoreau editors. Thoreau’s manuscript journals are literary daybooks, not diaries of the author’s life. The informational notes for these texts appear in the back-of-book records, since they’re few in number and supply little beyond the sources of scattered literary allusions. Thoreau’s letters, however, are filled with references to friends, casual acquaintances, and the events of his daily life, and the volumes of his correspondence carry informational notes immediately after each letter.

3. Introductory Editorial Notes

Any format for footnotes or back-of-book notes can be supplemented or even replaced by editorial introductory passages or headnotes. In near-comprehensive editions, such notes customarily refer to the specific document or group of documents that they introduce. But some editions, especially those directed toward a general readership or those that are the products of substantial selection and topical organization, rely solely on these notes, eschewing footnotes altogether (see, for instance, the series Freedom: A Documentary History of Emancipation). Whenever all or part of a series is organized by topic, each group of documents should be preceded by an editorial introduction that explains the criteria for selection, as well as the scheme of organization of the texts that follow, and it should provide whatever information the reader will need to understand the texts in that section.

In selective editions organized in a single chronological series, editors often insert paragraphs of editorial comment as transitional bridges between sections of documents. Such notes provide a context for the materials they introduce, summarizing the events not documented by the texts and offering the reader the information required to comprehend the texts that follow. In the Webster Correspondence, such notes are printed at regular intervals between the texts of letters.

4. Annotational Formats in Electronic Editions

Here we’ll examine five electronic annotated editions of documents for which at least substantial segments exist in finished form. Among them, they represent a fairly representative sample of methods of annotation and format currently in use.

The John Dewey Correspondence, available both online and in CD-ROM format, is a born-digital series within the broader print edition of Dewey’s writings being prepared at Southern Illinois University. Selections from the comprehensive online edition of letters will be published in book form in the future; consequently, the online letters carry very light annotation. The Dewey Correspondence includes conventional statements of editorial policy and a helpful historical introduction, principles of transcription and search tips, a list of sources and documentary abbreviations. In addition there is an alphabetically arranged section of “identifications” of persons, places, and events frequently mentioned in the Dewey letters and another section containing images of drawings and photographs that were part of the original documents. Linkage among these elements is somewhat limited. A user can move immediately from a reference to a drawing or photo in a letter or its notes to the image. There are no links, however, between documentary texts or annotation and the alphabetical identifications. Similarly, the codes for repositories and short titles of printed sources cannot be expanded on-screen; the reader must look elsewhere for a static list of abbreviations and full titles.

The Dolley Madison Digital Edition, part of the University of Virginia Press’s Rotunda, is a born-digital project that has never had any life in printed pages. Its release was preceded by the one-volume Selected Letters of Dolley Madison, also edited by Holly Shulman, director of the comprehensive online edition. Informational footnotes for documents in this edition are designed to be multipurpose, serving several documents that refer to the same person, place, or event. Thus in the exchange of letters between Dolley Madison and David Bailie Warden reproduced here in figures 6 and 7, you see a box with a list of annotational topics at the right. (Provenance information appears at the foot of the document.) Clicking on the name “Warden, David Bailie” in the list for either letter takes the reader to the same editorial note (fig. 8).

The University of Virginia Press’s Rotunda provides useful contrasts in three other series of personal papers that demonstrate different ways in which editorial annotation can appear online. None of these were “born digital” but are instead electronic versions of printed series published earlier by the press.

Two editions are part of Rotunda’s Nineteenth-Century Literature and Culture series, The Letters of Matthew Arnold and The Letters of Christina Rossetti. The pages of Arnold’s and Rossetti’s letters appear on the screen in searchable text (figs. 9 and 10), but they are essentially on-screen facsimiles of the printed pages. To read a footnote, the user may click on a hypertext link or scroll to the foot of the document; to follow a cross-reference within a document, the user must return to the “Search” screen to locate the document whose text or notes supply the needed information. Cross-references in the notes (such as “See letter no. 21, n. 8” in note 1, fig. 10) can be found immediately by clicking on the highlighted text. The Rossetti edition has one useful feature lacking in the Arnold Letters. The reader can scroll the cursor over abbreviations for sources and short titles and immediately see the full name for a repository or citation for a book.