Assessment

The project team found the preservation activities to be a source of significant learning and experience. While we began the project with many hypotheses about what challenges we would find, we discovered many areas that were either unexpectedly resource-intensive or surprisingly easy to address. These challenges and some opportunities to circumvent them are defined in the companion preservation guidelines document. They are both technical and social. Like all preservation challenges, they do not have one solution. Rather, authors, publishers, and preservation services must collaborate to make the best decisions based on the nature of the work at hand, existing resource limitations, and established standards for digital preservation. Decisions often require tradeoffs that affect the user experience, functionality, or level of confidence in long-term preservation. This section will outline those challenges at a high level and connect them to examples or patterns observed during the project.

One area of widespread concern is the embedding of features that depend on third party services. Video from YouTube, Vimeo, or other video services, social media feeds, data visualizations, or even whole web pages that are embedded within a publication present recurring challenges to preservation. We hope that our recommendations—such as hosting resources locally or procuring copies of third party resources, providing clear descriptive and rights metadata, and avoiding HTML inline frames (iframes) that present content on third-party websites —will mitigate some of the risks, but there are two important tensions here in scholarly communication that scholars and publishers should keep in mind when thinking about preservation:

First, there is a tension between different registers of scholarly communication. Many scholars write in blogs and journals where the freedom and convenience of embedding YouTube videos might outweigh the need for long-term preservation. Scholars who study audiovisual media often make liberal use of embedded third-party resources in online publications. Some of these publications fall into conventional conceptions of the scholarly record; some do not. But as Lavoie and others have discussed, the concept of the scholarly record has itself evolved in the age of digital publication:

The scholarly record, by virtue of its transition to digital formats, is now much more mutable and dynamic than in the past; it is made available through a blend of both formal and informal publication channels; and the scholarly record’s boundaries are expanding to include a much wider context surrounding the publication of a scholarly outcome. (Lavoie, et al, 2014, pp. 8–9)

A number of University of Minnesota Press publications on Manifold are drafts or “in-process” work from their Forerunners series. Conventions around the scholarly record would indicate that drafts are not part of the scholarly record and do not require long-term preservation for the sake of that record; however, insofar as these publications are used, cited, and themselves participate in scholarly discourses, they may still represent important pieces of discourse that would leave lacunae in the scholarly record.

The second relevant tension for scholars and publishers is around innovation. For some digital humanists and other scholars who make use of digital tools, innovative technologies do not simply enhance scholarship but rather constitute an essential element of their work. For these scholars, some technologies that can hinder preservability also allow for new kinds of scholarly communication. Networking between one publication and other online content is sometimes an essential component of digital scholarship, whether that content is data, software, social media content, audiovisual content, or text-based work on the web. The authors and publishers of Cut/Copy/Paste and The Chinese Deathscape may regard the interactive maps in these works to be essential; in cases like these, more preservation-friendly alternatives are suitable only to the extent that authors and publishers consider them suitable within the vision of work.

With these tensions in mind, there are a few activities in which scholars, publishers, platforms, and communities can engage to mitigate the challenges of preserving new forms of digital scholarship. The first is to make preservation a formal and intentional part of workflows, training, and professional development. When scholars, production staff, and platform developers think about even the most basic tenets of digital preservation, they can make decisions and ask questions that will prevent preservation roadblocks when their work reaches preservation services. If authors and publishers are in the habit of procuring copies of remote resources, for instance, preservation services can generally be confident that they will be preserved, despite many potential uncertainties in web archiving and emulation.

Second, all parties involved should use and continue to develop standards that help preservation services to scale the preservation of digital publications. This includes using non-proprietary, broadly supported, and widely adopted open file formats; building websites according to web standards for development and accessibility; and following (and making sure vendors follow) the recommended standards for EPUB and PDF. It also means advocating for new standards, such as a standard for archival EPUBs, that would help both scale the preservation of EPUBs and further guide publishers toward preservability even when the full standard cannot be met.

The companion document to this report, “Guidelines for Preserving New Forms of Scholarship,” was written in response to the work of this project. The guidelines suggest ways to make new forms of digital scholarship—from ebooks in conventional formats enhanced with interactive resources to digital publications custom-built using web technologies—in ways allow for more reliable and scalable preservation. While they build on decades of experience and scholarship on digital preservation and digital publishing, these guidelines represent the beginning of an intervention with publishers and platform developers, who, we assert, must now take part in a preservation process that in the past did not require their participation. As such, we expect to build on these guidelines as they are put into practice, and as we learn more from publishers, developers, and preservation service providers.