The tide is shifting in research data management and preservation. There is a move away from innovation-only funding opportunities, but as yet, few funding streams are available to maintain the digital scholarship whose development has been grant funded. My conversation, entitled Data Preparation: Striving to Balance Sustainability and Innovation outlines the process of data preparation in the Lesbian and Gay Liberation in Canada (LGLC) project starting in 2013, in the face of funding guidelines that favoured, at the time, novelty over long-term preservation. Much has changed in research data management in the intervening years, with the balance shifting toward data management plans and the adoption of standards. This talk will invite a discussion of best practices for meeting funder requirements while planning for a project whose long-term maintenance may not have financial or institutional support.

This conversation takes as its departure a small aside in the responses to Ed Folsom’s Database as Genre: The Epic Transformation of Archives published in the PMLA in 2007, in which he suggested that databases and narrative are intrinsically opposed.

In response to Folsom’s polemic, Jerome McGann was left wondering where TEI-XML stood in the PMLA debate on about databases, especially considering that Folsom’s largest Digital Humanities project, the Walt Whitman Archive, is a TEI-based project. His response to Folsom, Database, Interface, and Archival Fever, outlines his two difficulties with Folsom’s articulation. The first difficulty, according to McGann is what he calls a loose way of thinking about our paper-based inheritance as well as about these new digital technologies (158). Folsom and Price’s Walt Whitman Archive is not, McGann notes, a database (which Folsom protests, praising Hayles for showing how vital metaphors are, and describing the term database as a metaphor, a base onto which we put things that are given (data). (1608)); more accurately, it is an online archive driven by an inline markup structure (XML) and an XSL-generated interface (1588). Parts of the Walt Whitman Archive do run off a MySQL database, but McGann’s point holds – the project is a TEI-based one, not a database-based one. McGann also undermines the idea of database as the ultimate tool of the twenty-first century by proposing that for scholars interested in migrating our cultural inheritance to digital environments… TEI and XML … are better [than database models] because they model some of the key forms of order that are already embedded in textual works like Whitman’s … [and] because they understand that works like poems and novels are already marked data — while acknowledging that TEI and XML are not wholly sufficient to represent data for scholarly purposes (1589).

Fifteen years after this debate in the PMLA, with the new cultural engagement with sustainability and research data management, McGann’s arguments are perhaps more prescient than ever. TEI-backed linked data may prove to be realisation of his vision of the ultimate tool of the twenty-first century Humanities scholarship specifically because, unlike databases or dynamic visualisations, it is easily archived. TEI’s readily achievability, exchangeability, and its potential to contribute to the semantic web mean that it stands a good chance of contributing to future knowledge. This conversation will conclude with suggestions for encouraging the use of sustainable formats and reflections on how the LGLC project members hope the LGLC data might be preserved, but also how it might contribute to future scholarship.

Works Cited

Freedman, J., Hayles, N.K., McGann, J., McGill, M.L., Stallybrass, P., Folsom, E., 2007. Responses to Ed Folsom’s 'Database as Genre: The Epic Transformation of Archives.' PMLA 122, 1580–1612. https://doi.org/10.1632/pmla.2007.122.5.1580

To avoid fragility, online digital projects require stable hosting, regular maintenance, and backup; to avoid disappearing, they need proper archiving. This session will consider some of the problems of Hosting, Maintenance, Backup, and Long-term Storage. Having had experiences from the perspectives of the academic, the developer, and the project manager on digital projects, James Cummings will present some examples of problems that can – and have – occurred in these four key areas.

(1) Hosting problems include a wide range of possible stumbling blocks. From even getting sufficient access to appropriate institutional server space (once described by Miriam Posner as the hardest computing problem in the world) to the problems caused by the inevitable march of technology. Some issues encountered include the unexpected change of platforms and software; the disappearance of servers; inadvisable personal hosting by the researcher, without institutional support or access; and the challenges for an institution hosting a project whose leader has left in one way or another.

(2) Maintenance problems seem inevitable as all projects, however minimal, need some maintenance to at very least their infrastructure. While the guidelines of The Endings Project are an important and laudable goal, historically most projects aren’t as proactive in the planning stages of lessening the maintenance burden which will arise as a result after the project has concluded. Sometimes this happens when a project relies on individual rather than institutional infrastructure, or when a project is orphaned for whatever reason within an institution, or when basic software and hardware updates and security patches are not done.

(3) Backup problems one would think would be a thing of the past in a day when digital backup can be seamless and transparent. Although there are still many instances where materials cannot be backed up for legal reasons, in most cases it is not that backup can’t be done but that it hasn’t been done or has been done poorly. There may be no physical, archival copies of data alongside the digital copies, or the location of backups is insecure or unknown. Individuals may feel backing up to Dropbox on a single project member’s account is sufficient, or that placing their materials in a close repository which limits access is acceptable.

(4) Long-term storage continues to be a problem and there are some choices a project can make which makes their long-term preservation for later re-use more of a possibility. Problems tend to occur when inadvisable technical choices (such as file-format selection) have been made along the way, or when a project’s focus on interface or functionality (e.g., search functions) has delayed or complicated the storage of essential data. The most common limitation is when after the project is done, the data is only preserved in, and accessible from, a single repository.

Cummings will give examples of a small number of projects in which some of these problems have been evident, and what these projects did – or should have done – to mitigate the effects.

What happens—or can happen—when, after creating online resources, institutional leadership changes, archival web sites are taken down through redesign or rebranding? This may occur with non-traditional collectors or large institutions. What are the rights of the institution versus those who created, recorded, and prepared the material for display? What are the rights of funders, and the general public? What special challenges can develop if vulnerable populations are involved, for example, Indigenous groups concerned with cultural preservation, or students whose career prospects are potentially damaged when access to material they helped create disappears? What are the best practices to ensure that the rights of all involved are protected? And if even best practices fail to guarantee rights, what strategies might be put in place to mitigate damage done?

A significant case study, the Banff New Media Institute will be provided with two other case studies, the Daniel Langlois Foundation archives and the Women’s Labour History Archives.

The Banff New Media Institute (1995 - 2010), located at The Banff Centre in Alberta, Canada, was an international think tank on the present and future practices of new media, a training hub, co-production centre, business incubator and accelerator, and a comprehensive research infrastructure and residency program, and exhibition curator. Over the course of its first ten years it held over 150 summits, symposia, workshops, conferences and events which were recorded and relevant records such as release forms, agendas, analysis of events and other memorabilia kept. These events brought together leaders from around the globe who were engaged in developing the science, creative practices, jurisprudence, business models, design practices and theories in the emerging new media world. The BNMI ensured strong representation of women and diverse presenters. A feature of BNMI was its deep commitment to Indigenous creators and technical leaders through inclusion at its events and within its research and co-production, and in collaboration with local Indigenous communities. Co-productions were described and catalogued, and the documentation made available on-line, as were exhibition descriptions and documentation through the Walter Philips Gallery and in physical form. The BNMI also created Horizonzero.ca an online gallery, commissioning environment, and site for analysis of the emerging new media world with a focus on Canadian practices within a global context. It provided a site of reflection for the BNMI activities (for example nano technology creative research) but stretched well beyond its boundaries.

I raised the resources and rallied the personnel in collaboration with Susan Kennard (BNMI Director from 2005 – 2010) and Manager before that time and researcher and curator Sarah Cook, to create The Banff New Media Institute archives in 2004-5 as I was leaving my job as Founding Director of the Banff New Media Institute (BNMI). We received significant support from Telefilm, the Canada Council for the Arts, Heritage Canada, and Library and Archives Canada. At the completion of the project all audio files from summits and workshops were available on-line as well as all event agendas, biographies of presenters, many papers that were provided, and event analysis. The on-line archive was easy to navigate and search. A physical archive with finding aides was created which held the corresponding catalogues, BNMI brochures, and all manner of planning documents.

Sarah Cook and I created a book, Euphoria & Dystopia: the Banff New Media Institute Dialogues (2011. Banff Centre Press & Riverside Architectural Press) which chronicled and analyzed the global era of the Banff New Media Institute, the Institute itself through commissioned essays, and provided analysis of six themes that ran through the ten years, excerpts from the summits and workshops as well as detail in a comprehensive appendix of events and attendees. It also included a DVD of Horizonzero.ca. An e-publication followed two years later, with hotlinks throughout to the BNMI archives.

Five years ago, during a change of leadership, new institutional strategy and web site redesign, The Banff Centre took the entire archives down, all the co-production information and any mention that the BNMI had existed at The Banff Centre. In 2019 The Banff Centre chose not to renew Horizonzero.ca. The Banff Centre is now reconsidering the value of the BNMI and its archive as it builds its digital renewal.

The Daniel Langlois Foundation archives was funded by Daniel Langlois the co-founder of Softimage. It became an important repository for the many new media art works funded by the foundation and had an excellent physical and web presence. After a decade Langlois’s interests moved on and the archive was closed. It took many years for it to find a new home, to the disappointment of artists and scholars.

The talk will address questions of responsibilities and rights in these two instances. The strategy for the current creation of the Women’s Labour History Archive with VIVO, an artist-run centre and Simon Fraser University, reflects lessons from these two examples.

When funded scientific or creative projects end, records and work remain. Today these records are often ‘born digital’ or include some work converted from analogue to digitized formats. Records preserved in digital media are not only of interest for the legacy of projects that have ended, but can constitute a continuing resource for future scientific or creative activity. In this session, Dr. Marontate explores how diverse (and sometimes incommensurate) value systems embedded in technologies and in values of practitioners raise challenges for digital preservation.

Designers of projects for digital preservation with resilience for future uses face challenges in anticipating and understanding changing forms of digital media. Technologies used in creation and preservation of digital media shape both creative practices and conservation activities. When technologies change, records of past work and established practices, even very recent ones, may be lost or altered or become unviable. Even recent work (such as statistical data sets or videos) may become unusable and inaccessible when software programs are upgraded or new hardware replaces old. Computer upgrades and transformations in equipment have become commonplace events. Most computer users have experienced loss or change in data when they adopt new systems. Modifications to the original may have serious consequences, rendering valuable work and data sets meaningless or indecipherable. How often do we discard old work or data sets because they are difficult or impossible to use with new software or equipment? When technical practices change, new assumptions and values may emerge about the meaning of the collections.

Yet, information about the values of people involved in creation and preservation activities is often fragmentary and incomplete. Even within a single project creators, researchers, archivists and conservators may prioritize different qualities. These values have an impact on what is created and how it is preserved. It is not always possible to gather such information later. Creators of images, texts and sonic materials eventually disappear. Even when they are still alive their recollections and priorities may change. Researchers, technicians and research assistants move on. Documentation of the work of creators and collaborators (including people providing technical support) may not include the information needed for maintenance of the full range of features of materials in the preservation of digital collections.

How can digital preservation acknowledge the multiple meanings of works for creators, and the activities of various types of team members involved in both the creation of digital materials and the preservation of archives? How are decisions made about what to preserve and how to document it? What are the implications of such decisions for long-term maintenance of digital media? Some types of digital works, such as audio-visual materials (like sound recordings) may present specific features for preservation activities, for example, as new technologies are developed to allow greater flexibility, larger data sets and different playback options.

The polysemy (that is, multiple meanings) of digital works presents challenges. The works preserved themselves may have different meanings for originators or creators, and for people involved in preservation. Pragmatic approaches in French cultural sociology suggest ways of taking diverse (and sometimes competing) value systems into account in digital preservation initiatives. These approaches do not establish normative principles. The goal of studying meanings in the generation and conservation of digital media is not to determine what to preserve or how to conserve digital media. Rather the goal is to identify how diverse value systems and multiple meanings of works influence digital media preservation.

Dr. Marontate begins with a brief reflection on ways of studying the interplay of symbolic and material dimensions in digital documentation practices. These include analysis of many types of activities, among them: changing technologies and their impact on features of digital collections, new facilities for embedding documentation in metadata, and ways that varying values shape both the creation of works and their preservation. Examples of preservation activities illustrate how the 'meanings' or socio-cultural significance of activities are preserved in digital formats. Digital archives that include historic materials may raise specific questions too, since multiple processes may have transformed materials over time. Next, she considers innovative theoretical and methodological frameworks for analysing value systems proposed by Nathalie Heinich, based on studies of conflicting value systems by Luc Boltanski and Laurent Thévenot. Debates in archival and conservation networks also illustrate the multiple challenges in crafting memories of past for the future. She concludes with a critical analysis of the potential uses of studies of changing technologies and values for understanding and designing digital media preservation.

While some digital humanities (DH) projects can be undertaken with free software and whatever time a scholar already has to hand, larger and more collaborative projects generally need additional resources. In the traditional book- and article-based humanistic workflows, universities, colleges, archives, special collections, publishers, and libraries have established labor and funding models that aim to support text-based research, writing, publishing, distribution, and preservation pipelines. But DH projects—collaborative or otherwise—rarely fit into these pipelines and subsequently cannot readily avail themselves of these preexisting labor and funding models. The path towards a collaborative DH project instead requires scholars to invent funding, often at every stage of their project's lifecycle. Like other humanists whose institutions fail to provide them with sufficient funding for their scholarship, many digital humanists turn towards self-funding to cover part or even all of their project costs. But when self-funding becomes infeasible, as it often does with larger and more collaborative projects, digital humanists typically turn towards grant-writing.

Grants might seem like the optimal solution to digital humanists' funding problems, as they entail receiving funding for a specific length of time to generate a specific project, much as other humanists are funded to visit a specific archive or come under contract to write a specific book. A variety of funders often grants for the creation of digital projects, which makes it relatively easy to align a scholarly agenda with the agenda of those funders in order to create something mutually beneficial. Furthermore, grants come with numerous additional benefits, including a peer review process that helps to establish the digital project's scholarly value prior to the investment of significant time and money; built-in project publicity from the grant funding announcement; and institutionally internal and external prestige for having won a grant. University and college administrators are particularly pleased with humanists' growing ability to join scientists and social scientists in bringing in outside funds to support both scholarly research and the overall operating costs of the institution. That these grants are not always sufficient to create a project—requiring additional rounds of grant funding—simply requires thinking of a project in stages and the occasionally creative re-envisioning of a project.

However, while short-term funding can support the creation of digital projects, it fails to support the publishing and preservation stages of a project's lifecycle. It's not uncommon for DH projects to vanish after five or ten years, as opposed to traditional books and articles that could remain available for decades or even centuries. A scholar’s responsibility to a book can end with initial publication and outreach, but there is no digital equivalent to the legions of librarians who place books on climate-controlled and secure shelves for access, repair the physical item when it is damaged, and eventually determine if the item is worth long-term preservation in a special collection or if it can be deaccessioned for lack of use. Sustaining or preserving each individual digital project requires additional labor and funding, which can last an indeterminate period of time. These long-term costs do not fit neatly into a single grant cycle, prompting humanists to resort to increasingly creative grant applications and fundraising to reinvent old projects, bury their heads in the sand in denial, and/or accept the ephemerality of many (though not all) of the products of their hard work.

This essay will explore the role of soft and hard funding in the digital humanities, with particular attention paid to the ways funding intersects with sustainability and preservation. It will discuss a variety of types of soft funding available from public, private, and institutional sources, and the ways the needs of funders can align with or compete with the desires of scholars. The availability and sustainability of funding as project needs change over time, along with funders' agendas, can influence project teams' decisions about the direction a project will take, who leads it, who can contribute to it, what technology supports it, on what severs it is hosted, if and when it can be updated, as well as if and where it will be preserved. Not considering the implications of financial support received can be detrimental to both the project and the researchers involved. Seeking funding should become part of wider strategies that enable the formation of mutually beneficial relationships, promising avenues of research, and sustainable revenue streams, rather than a time sink that explodes a project's scope in ways that undermine the project.

The collaborative dimensions of many digital humanities projects are often complicated and constrained by ideas of value and conditions of labor in higher education. Increasingly, collaborative digital projects have made efforts to assemble core project teams that grant agency and authority to the various people and professionals whose contributions are essential to the success of these initiatives: students, community stakeholders, librarians, archivists, programmers, designers, developers. But these forms of collaboration are often varied, contentious, and messy, given that they span a range of different professional contexts while operating within environments that traditionally privilege and prioritize the roles and contributions of faculty members and their attendant scholarly outputs like monographs and journal articles. It may be tempting to adopt best practices and collaborative methodologies that prioritize cleaning up these particular messes instead of addressing their implications. This talk hopefully steers us towards other futures and possibilities for digital collaboration. For example, instead of designing project roles and workflows that acknowledge the precarity of student and postdoctoral labor in ways that privilege the longevity and endurance of a digital project, what would it mean to center collaborative work on the interests, needs, and aspirations of those particular contributors? Instead of prioritizing digital projects initiated by tenure-track or tenured faculty, what would collaboration and support look like if we deployed institutional and administrative resources to support projects initiated by students, librarians, archivists, artists, community partners? Instead of remediating academic investments in definitive, comprehensive, authoritative, unified forms of publication in our ideas of digital projects with clear beginnings, middles, and endings, what would it look like to create conditions of collaboration that emphasized a wider range of possibilities?

In this talk I will highlight the ways that the physical and textual records left behind by a project — web sites, publications, citations, institutional forms of documentation, digital assets secured for long-term preservation — can at times offer an incomplete record and sanitized view of collaboration on digital initiatives, a perspective that often reveals how particular collaborators are valued and treated by institutions and professional networks beyond their immediate departments and campuses. I will reflect on my experiences with collaboration in various project roles — graduate student, postdoctoral fellow, instructor, consultant — and how these particular contexts have informed my collaborative contributions and their reception. And I will discuss the ways that arguments for best practices in project development, collaboration, and project endings are inevitably challenged and complicated by the institutional and professional realities that shape the roles of specific collaborators and their working conditions.

My perspectives on collaboration draw on previous and ongoing efforts as a director of digital projects, as a consultant on digital initiatives, as a frequent collaborator with community partners, librarians, and archivists, and as an instructor specializing in digital public humanities. These perspectives are also informed by the work of Moya Bailey, Michelle Caswell, Marika Cifor, Julia Flanders, Paige Morgan, and Jentery Sayers, among others. I will discuss the role of graduate labor on Our Marathon: The Boston Bombing Digital Archive and the ways that project’s ideas of labor, forms of institutional support, ideas of audience, and investments in crowdsourcing variously shaped its development, accomplishments, and ending. I will highlight aspects of my ongoing work on Mapping Violence, a digital initiative that relies heavily on undergraduate labor and collaboration, that is informed by relationships with community stakeholders, that complements non-digital forms of public history and advocacy, and that aspires to self-reflective, iterative, and slower forms of collaboration and project development. And I will talk about my role as a postdoctoral fellow in an American Studies department and Public Humanities graduate program: my approach to designing courses that create opportunities for collaboration and project prototyping, my roles on digital initiatives created in collaboration with community partners, and the ways that my status as a postdoc has shaped (and, at times, constrained) my abilities to develop, support, and collaborate on digital projects.

We are at a turning point in the preservation of analog media. Experts predict that analog tapes will be unplayable by the year 2025. Much of the primary research data of linguistics, musicology, and anthropology is still located on this media and, in many places, the task of preservation and repatriation of the content remains to be taken up. Our research group, based at three Australian universities, addressed this challenge in the early 2000s, building a repository and digitisation workflow that now includes 115 terabytes of material, representing 1,267 languages in 14,000 hours of audio recordings along with manuscript and video materials. In the course of doing this, we train new researchers in appropriate methods for creating reusable primary records, and built a platform for citation of data, allowing verifiability and, more importantly, reuse of that data. We have done this with several 1-year infrastructure grants over the 18-year life of the project. While there was an Australian National Data Service, it paradoxically did not store data, and the current version, the Australian Research Data Commons, also has no national service for curating research data ( and so, no ‘Commons’|). So, despite large funding sources apparently available for research data, it is typically not available for the kinds of humanities records we are concerned with. More frustrating is the refrain that humanities researchers receive from national research funders that we are unable to articulate our infrastructure needs.

In the absence of institutional or national work to preserve the outputs of research in our disciplines, our project, the Pacific and Regional Archive for Digital Sources in Endangered Cultures (PARADISEC), has developed exemplary methods and services based on accepted standards for metadata and data formats. We have trained a new generation of researchers, especially in our work as the repository for the ARC Centre of Excellence for the Dynamics of Language. We act as a repository that curates and preserves research at all stages of its creation: recordings deposited during fieldwork, or at the end of a research project or a researcher’s career (or life). To support longevity we have always written an XML file of the complete catalog entry to the item (the files), so each item is self-describing. We are now building an OCFL (Oxford Common File Layout) version which reduces dependence on monolithic databases to interpret the items and their metadata. This allows us to build microservices for viewing items, and to develop more meaningful ways of returning files to source communities, together with contextual metadata.

And, perhaps more important than any of this, digitisation has allowed us to engage with the communities in the Pacific, PNG and South-East Asia, that are the source of the recordings we can now return, half a century or more after they were made.

I will show that a repository for primary research data has four major benefits:

• Citability of research data for verifiability/replicability of research claims
• Provision of a curated corpus of data for reuse
• Establishing a home for end-of-life research materials
• Licensed access to records for the people recorded, their communities, and their descendants, thus mitigating the earlier perception of academics ‘stealing the data’.

References

Nick Thieberger. 2016. What remains to be done – Exposing invisible collections in the other 7000 languages and why it is a DH enterprise. In Digital Scholarship in the Humanities. 32(2), 1:423–434. http://dx.doi.org/10.1093/llc/fqw006

Nick Thieberger & Linda Barwick. 2012. Keeping records of language diversity in Melanesia, the Pacific and Regional Archive for Digital Sources in Endangered Cultures (PARADISEC), pp. 239-253 in Nicholas Evans & Marian Klamer (Eds). Melanesian languages on the edge of Asia: Challenges for the 21st Century. LD&C Special Publication No. 5. Honolulu: University of Hawai'i Press. http://scholarspace.manoa.hawaii.edu/handle/10125/4567

Digital humanities projects often begin with specific scholarly goals in mind, whether these goals are editorial, theoretical, archival, or methodological. When projects begin, all is promise, potential, and excitement. The grant funding structure on which many projects are founded also demands that projects begin with an optimistic view and a future-oriented perspective. A team dreams up a vision and creates a plan for executing it, and it’s difficult both practically and epistemologically to think about an ending while constructing a beginning. Often these projects begin in heightened moments of collegiality, collaboration, and shared vision among scholars; they’re premised on new and exciting relationships as much as they are on research materials and scholarly goals. They’re also often labours of love. This discussion will focus on digital preservation as it concerns researchers: their own careers, their scholarly aspirations and their collegial relationships with one another. I will draw in my remarks from my own experience and will contrast two projects with two very different approaches to endings: one ongoing collaborative digital endeavour, the Modernist Archives Publishing Project (MAPP), and one grant-funded, constrained exhibition project, Make Believe, which had a clear and limited trajectory and timeline and a definite end-point. I will focus here on the narrative and affective dimensions of digital project life cycles. What does it feel like to talk about endings, especially among close-knit scholarly teams who have collaborated for years?

A digital project is born, and then its life proceeds and gets complicated. Of course, as project teams devise digital methods for exploring humanistic questions, they are often confronted with challenges and opportunities they would not have encountered were they writing a single-author monograph or creating a print edition. One such challenge, as the Project Endings teams acknowledges, is defining the endpoint of the project in a digital environment that is by nature iterative and surrounded by a mythology of ongoingness and perpetuity. The print publishing process, conversely, naturally produces a moment of completion: a book is published and then it is out in the world and often that’s the end of the active day-to-day work on that particular project. The developmental arc of the monograph is so well-established at this point that we’re used to it; it doesn’t possess the complexity as discussions of digital project endings. Much as the author might wish to continue to revise and augment the book after it’s done, the fixity of the printed object at least imposes an artificial endpoint on the active work of the project. The existence of a physical object can also seem more definitive than any ending a digital project might devise or arrange for itself. The milestones in the life of a digital project — the launch of a website, say — are often possessed of a more malleable quality. One often launches a website with the idea that it will change and grow over time.

However, the discourses of ongoingness and iteration in digital project work, while they do point to a truth about the type of media in question, have their limits. As digital project team members move through their careers – from PhD to postdoc to demanding jobs both within and outside the academy — finding space for an ever-expanding and ever-continuing digital project within the team members’ individual careers can be a further challenge. Grant cycles come to an end, program centres open and close, and resources run out. Thinking about the long life of a project is twinned with thinking about the long future of an academic career, and even, these days, of the discipline. Thinking about these matters requires a multi-dimensional approach: we need to think beyond institutional repositories and mirror sites and consider the lived experience of project making and about the structure of the stories we tell about digital work.