Producing Print and Digital Editions

One of the primary challenges in creating a digital or hybrid dissertation is finding a workflow that can output a final product in multiple formats. University libraries almost all require submission of some form of print material to deposit in the institutional repository. In the case of George Mason University, a Word template provides guidance on preparing a properly formatted document. It helps candidates meet the requirement for content and format of the title page, other front matter, tables of content, section headings, figures, and more. While useful in meeting prescribed standards, these documents are not particularly accommodating of digital components such as interactive maps, dynamic visualizations, cross-linking between sections, or other elements common in digital platforms. Additionally, the precise formatting of the deposited dissertation is rarely appropriate for a digital platform and the textual content of a Word or PDF document cannot simply be pasted into a web page. Likewise, content written for a web page cannot easily be migrated into the format required by the library.

For these reasons, producing a digital or hybrid dissertation requires a workflow that does not prioritize one output over another because transferring from one to the other late in the process is extremely difficult. To solve this problem, my workflow relied heavily on a composition tool called Scrivener that facilitates writing, reorganizing, and citing without pre-determined output settings. Once the written content and citations are finished, Scrivener allows output to a variety of formats, including Word documents, HTML, and Markdown (among many others). This allowed me to export the content to a print edition to send to my committee members for comment and to prepare for submission to the library. I also exported the text to a Markdown file which could then be translated into an HTML file using Pandoc. I used this approach rather than a direct output to HTML because the latter includes style elements unnecessary for my needs. Finally, I copied and pasted the HTML content into the appropriate sections of the website.

This workflow is particularly useful because Scrivener exports footnotes to suit the output format. The Scrivener > Word process exports footnotes in the standard DOCX format. The Scrivener > Markdown > HTML process creates the required tags to facilitate in-page hyperlinking for footnotes on web pages. When changes to the content or footnotes were required, I made them in Scrivener and re-exported to both outputs, eliminating the need to translate between print and web formats.

However, this workflow does have some drawbacks. Comments and proposed changes made by external reviewers in the Word document cannot be automatically accepted and must be made in Scrivener instead. While not as efficient, this process does require close attention to the nature of each comment or change, which may be beneficial to some writers. The most significant drawback is that the Scrivener > Markdown > HTML process removes font styles entirely, requiring them to be re-added when copied into the web pages. This issue is most evident in footnotes, where resource titles are often italicized. A direct export to HTML would retain these styles but also include additional unnecessary styles. In future projects, I would likely refine the workflow to find a middle solution between these two approaches.

The two output formats both required further processing to complete the finished products. For the print edition, I followed the guidelines provided by GMU’s University Dissertation and Thesis Services. For the digital edition, I copied each section of the text into the appropriate exhibit page on the Omeka Classic site. Where appropriate, I replaced static images present in the print edition with interactive content or galleries of images. I organized the site using the Exhibit Builder plugin because each exhibit can contain sub-pages, HTML content, custom embedded content, integrated Omeka content, and built-in navigation aids. The site has eight exhibits, each containing the text for a single module. It also features pages for the network visualization and allows browsing collections and maps on the site. The search function provides results from the exhibit pages, individual items, and collections. The ability to search across the written content and collection items allows users to pursue their own interests on the site, such as a family name or locality represented in the claim documents. The site uses a customized version of the Berlin theme originally built by the Omeka development team.

Dissertations are a unique form of scholarship because they may take on a variety of forms (digital dissertations or performance art, for instance) but must always be accompanied by a highly specialized print document that meets the standards defined by the university and library. The length and content of the official dissertation object deposited in the library is unspecified, which means that the repository records for non-traditional dissertations might include only an artist statement or a summary text. In the case of this hybrid dissertation, the entire text is presented in two matching forms: one print edition in the library repository and one digital edition on the project website. The workflow I developed to produce these two editions is not the only possible solution but was an effective way to reduce duplication of effort and circumvent difficult migrations between web and print formats. Future scholars seeking to produce digital or hybrid dissertations that take non-traditional forms while still meeting university requirements may find this workflow a useful template upon which to build their own.