METHODOLOGY

Tasks, Organization & Digital Technology Employed

All DMDE documents are controlled by DocTracker, a database developed and built by the DMDE to collect, edit, track, manage texts, and output XML pages for online publication. DocTracker shepherds documents from the time they are acquired, through delivery of the finished files to the publisher.  The following are the steps DMDE editors take when processing a document for publication:

Capture

The initial step, after locating material(s), is to capture the document.  In the early years of the project, most documents were photocopied, while documents from larger collections, such as the Library of Congress—which had been microfilmed—were digitized.  Today, editors make digital images of all new materials, resulting in large files that may be visually manipulated to facilitate the transcription process. All images are initially saved on the DMDE’s shared drive, provided by the University of Virginia.

Entry into Production Database

Editors first enter the documents into the ‘DocLog,’ whether or not they are intended for publication.  ‘DocLog’ captures basic metadata: a unique identifier, author/receiver; document type (text or table); date; source; rough transcription; image, etc.  Following this, a document can be promoted" to the edition, thereby placing it in the queue for production.  Documents may not be promoted  for a variety of reasons: some may be the file copies of letters that are published; some are third party correspondence used for research purposes; others are non-epistolary materials such as recipes, newspaper articles, poems, programs for balls and charities, and other ephemera whose publication status has yet to be finalized.

Once "promoted," the editor assigns the document to the appropriate volume based on date.  The editor then adds metadata on the ‘Control’ tab, such as document genre (correspondence, legal, financial, etc.), manuscript type (receiver's copy, file copy, published copy, etc.) keywords, partial transcription, summary, and editorial (e.g. dating an undated document based on internal evidence) and bibliographic (e.g. noting if a document is torn, missing pages, etc.) notes.

Transcription & Verification

The editorial methodology for transcribing documents includes the retention of archaic and misspelled words, abbreviations, and original punctuation and style (underlining, superscripts, capitalization, dashes, etc.), with the following exceptions made for readability and clarity of meaning. The date and origin of a letter, if included, are located at the top of the transcription, regardless of where they appear in the original.  If the letter does not include a complete date, the conjectural version is placed in the letter’s header, with the conjecture included in square brackets. Illegible or missing material, whether from damage to the letters, tears, or illegible handwriting, are rendered with angle brackets and ellipses, thus: < . . . >.  Unclear characters or words, for which the transcriber has been able to deduce an unequivocal reading, are displayed in green; readings that are more uncertain are displayed in red. If there is no punctuation at the end of a paragraph, a period is included for clarity. Dashes of various lengths have been standardized to an em dash, except where a very long dash has been employed by the author to indicate the end of a sentence, in which case the dash has been rendered as a period. Following print convention, dashes used to represent missing portions of a name are given as two em-dashes: Mr. S——. In cases where two punctuation marks appear together, only the first has been rendered in the transcription. For example, instead of “I received your last letter. —” the transcription reads: “I received your last letter.” Similarly, a double period underneath a superscript element has been rendered as a single period (for typographical reasons, this period follows the superscript in the online version).

In the past, transcriptions were made in Microsoft Word, then printed and proofread. The transcription was then tandem-read and corrections were noted in the paper file.  Thereafter, the transcription was entered into the production database. Today, the DMDE is transitioning away from paper files, so transcriptions are entered directly into the database, which also stores the document images. As the transcription is entered, the editor applies the visual XML tags via the scripted buttons.  The editor then notes any specialized, or ‘hand-tagging,’ necessary, some of which may be added after the documents are output to XML files, perhaps by an editor more familiar with TEI and/or XML. The transcription is then proofread and later tandem-read; each step is recorded on the ‘Workflow’ tab to ensure quality control and to drive the reporting features of the database. When in the document is placed in the production queue, it is marked for review by the senior editor.  Before the final submission of the edition to the publisher, the entire edition is reviewed one final time by the editors.

Annotations

While applying the visual XML tags, an editor will also begin to markup the document for annotation.  If the document contains a previously identified person/place etc., then that identification, or “ID,” is linked to the corresponding text in the document.  If there is no existing identification, then the editor creates a blank “ID” and links it to the appropriate word in the document, thereby signaling that the annotation needs to be researched and written.  Ideally, the DMDE strives to identify all people, places, titles, and organizations mentioned in a document, bearing in mind the audience and time constraints.  Editors generally limit their time researching an annotation to two hours.

After the annotation is researched, drafted, tagged, and linked to the appropriate document(s), fact checking and approval(s) are conducted. Next, the newly written and revised (since the last publication) annotations are reviewed by the copyeditor.  The edits are made in the database and the final submission of documents and annotations is output for transmittal to Rotunda, with one further step in the digital publication's workflow remaining.  Indeed, after receipt of the edition, Rotunda edits the submitted digital "manuscript," smoothing the work of many hands into one consistent publication.  Rotunda’s edits—which are entered into the separate XML files, not in the DMDE's production database—must then be synchronized with what is online, the final publication.

Editorial Essays & Introductions

In addition to the annotations, the DMDE provides "Editorial Notes," essays that range in length from roughly 500 to 4,500 words.  These notes provide DMDE readers with context and scholarly discussion of topics that require more explanation than can be offered in an annotation.   Over the years, essays have been written on a variety of topics, from the popularity of autograph collecting—and Dolley’s involvement with the hobby—to more detailed essays on her land sales and connections to her slaves.  Editorial essays are linked to relevant documents (and vice-versa) and are also accessible through the table of contents. The scope of the Editorial Notes is at the sole discretion of the editor-in-chief, Shulman.  Similarly, the DMDE has published an extensive introduction, which is updated with each new volume and includes sections on editorial methodology, technical explanations, project history, and acknowledgments.  The goal of the introduction is to present and explain the edition to the general public.