Digital Repository Legacies: A Case Study in Assessing Organizational Trustworthiness

Abstract Academic libraries rarely discuss cases of digital repositories that do not meet the standards expected of trusted digital repositories. Implications from inconsistent adherence to technical and professional criteria often surface during migration projects. In 2020, Stony Brook University Libraries began migrating assets to a mono-repository environment. Persistent historical factors presented challenges to repository trustworthiness. This case study discusses a survey project to evaluate legacy repository statuses in the contexts of infrastructure, documentation, and staff capacity. It considers a paradigm of organizational accountability in digital asset stewardship and offers insights for reconciling inherited legacies with aspirations to be a trusted repository.


Introduction
A digital repository is a platform to preserve, collect, manage, and retrieve digital assets. Its aims and objectives present a case in duality: for a user, it provides a means of online access to content of high research value and for a library, it additionally creates a space for stewarding its digital assets. Repository management requires thoughtful planning that integrates the traditional tenets of librarianship including appraisal, acquisitions, collection development, and cataloging with the technological processes of ingest, digitization, and migration. Institutional repositories (Gonzales, 2018;Rieh et al., 2007) in academic libraries often contain the research output of the institution's community and may include research journal articles, electronic theses and dissertations (ETDs), and materials produced during university-wide operations (New World Encyclopedia, n.d.). In this case study, an institutional repository ("IR") refers to "a set of services that a university offers to the members of its community for the management and dissemination of digital materials created by the institution and its community members" (Lynch, 2003b, p. 328). A digital repository ("DR") includes both the IR and the digital content management applications and systems utilized for curating digital assets. paradigm of organizational accountability in digital asset stewardship and offer insights for reconciling inherited legacies with traits characteristic of a trusted repository.

About the university and libraries
Founded in 1957, Stony Brook University (SBU) is one of four research-intensive campuses within the State University of New York (SUNY) system. Located approximately 60 miles east of Manhattan, New York, SBU is ranked among the top 1% of universities in the world with an enrollment of nearly 27,000 students (Stony Brook University, n.d.). The curriculum encompasses 200 majors and minors ranging from medicine to music. SBUL's collections and services are central to the university's research mission and intellectual endeavors. Collectively, SBUL is the largest academic research library on Long Island, New York (Nassau and Suffolk counties) and comprises eight distinct facilities. Of particular note is its membership in the Association of Research Libraries (ARL) and the division of Special Collections which stewards rare books, manuscripts, archives, and historical maps dating from the 11th century.
Beginning in the late 1990s, SBUL embarked on digitization projects including a regional effort to expand access to local history collections and participation in SUNY's DSpace IR, used primarily for university-created content. Initially, the information technology (IT) infrastructure supporting these projects was co-managed by SBUL and external departments that serve the larger university. With this arrangement, the libraries benefited from campus-wide expertise and consultation. In recent years, administration of SBUL's IT activities shifted to departmental staff primarily tasked with managing an integrated library system (ILS). Investment has been made in developing multiple digital repositories, but yet to be realized is a program for digital asset management. This status has created issues for moving collections to a single, up-to-date local instance of DSpace. Separate from the libraries' online catalog, the DRs at SBUL have historically been used as short-term storage and access systems. Be Press, an IR software, is currently used exclusively to publish faculty scholarly output. Local and consortial versions of popular repository platforms including OCLC's CONTENTdm and Atmire's DSpace store and disseminate SBUL's digital files of cultural heritage and university history. Since 2016, SBUL also uses locally-hosted versions of Omeka Classic and Omeka S as digital exhibition platforms. At this time, no existing repository system is used for preservation purposes.

Research questions
To aid the in-process repository migration project and inform future technology planning, this study considered the following research questions to assess SBUL's organizational infrastructure and capacities for managing DRs and digital assets. The normative metrics were taken verbatim from "Appendix 3: Minimum Required Documents" of Trustworthy Repositories Audit & Certification: Criteria and Checklist (TRAC) and the rating scale is part of the Digital Preservation Coalition's "Rapid Assessment Model" (DPC RAM).

Literature review
Impetus for developing digital repositories Rationales for investing in DRs are diverse and principally intertwined with a library's efforts to further its institutional academic and research missions. Repositories can serve as forums for disseminating and showcasing scholarly output, function as open access alternatives to traditional publishing models, and be utilitarian virtual storage spaces for digital assets. Dale and Gore (2010) discuss the lineage and transition of process models in the context of preservation and their relationship to trusted digital repositories through enumeration of the Open Archival Information System (OAIS), InterPARES (the International Research on Permanent Authentic Records in Electronic Systems), and DCC (Digital Curation Center) Curation Lifecycle Model which segues to Trustworthy Digital Repositories (TDRs). Writings exploring the intended use and users of IRs describe fluctuating dichotomies among them. According to Novak and Day (2018), a dialectic exists in defining their purpose as either a channel for scholarly communication or as a tool for supporting research administration, with the institution as the prime user rather than a faculty scholar. Jones (2018)

draws connections between
IRs and the open access movement, suggesting IRs should be receptive to acquiring unpublished and published research, and facilitate unfettered access to this material. Plutchak and Moore (2017) systematically outline the aims of IRs by juxtaposing Crow's (2002) emphasis on IRs as viable open access publishing models with Lynch's (2003a) articulation of the "distinction between scholarly publishing and scholarly communication" (Plutchak & Moore, 2017, p. 28). Lynch (2003a) asserts that IRs foster scholarly communication dialogs that exploit the unparalleled access inherent to the digital medium and foster new scholarship practices for tenure-track faculty. While stressing the strategic importance of an IR as a possible "engine of change" (Lynch, 2003a, p. 7), Lynch cautions sustainability relies on "a serious and long-lasting commitment to the campus community (and to the scholarly world, and the public at large)" (p. 6).

Standards and best practices
Adherence to standards combined with flexibility is a favored practice as new approaches and benchmarks emerge, evolve, and may shift over time. Several technical reports and documents provide guidance and benchmarks for establishing, implementing, and sustaining digital repositories and the assets they contain. The OAIS (Open Archival Information System) reference model is informed by ISO 14721:2012, a technical standard for managing digital assets (ISO, 2012). This standard outlines the practical applications and investments required of organizations to ensure the accessibility and preservation of digital assets. Audit and Certifications of Trustworthy Digital Repositories defines best practices and identifies evaluative means for curating digital assets (Consultative Committee for Space Data Systems, 2011). This essential document articulates the components of a "trustworthy digital repository" (TDR) and includes auditable criteria for measuring organizational commitments to infrastructure, digital object management, and security risk management. Crucial to planning is formulating and expressing a mission, as this aim anchors the activities that ensue for building digital repositories. "At the very basic level, the definition of a trustworthy digital repository must start with 'a mission to provide reliable, long-term access to managed digital resources to its Designated Community, now and in the future'" (Consultative Committee for Space Data Systems, 2011, pp. 2-1). Seeking to delineate the nuanced difference between evidence of trust and perceptions of trust for digital repositories, Bak (2016) unpacks the word "trust" and the historical underpinnings that have manifested in distrust. He astutely points out that Trusted Digital Repositories: Attributes and Responsibilities (Research Libraries Group, 2002) includes an auspicious void, noting the word "trust" is never defined nor does it appear in the glossary of terms. Additionally, the word trust in the context of TDR is different; it is "technocratic because it refers back to the administrative and professional competence of the repository and its staff. This technocratic conceptualization of trust is auditable precisely because it is based upon the mandate, resources, actions and obligations of the institution, as represented by that institution. This avoids the messiness and contingency of trust in the real world, and as defined in the Merriam Webster dictionary, which requires building relationships with communities and with individuals" (p. 381). Dryden (2011) succinctly describes the historical development of standards for TDRs from a perspective of coherence and trust. "The vision of a distributed system of trusted repositories of digital collections was (and is) a worthy goal, but achieving it would require considerable agreement about the components and characteristics of such a repository. In other words, a high degree of standardization was needed" (p. 128). Comparisons of outcomes between repositories with and without certification are limited. Husen et al. (2017) studied the commonalities and divergences between recommended and certified repositories. Among their findings were that "less than 6% of recommended repositories obtained some form of certification" (p. 8). They advocate for standardization in practices and incentivizing their adoption to improve the landscape of data curation. While Donaldson (2020) questions if certification status ultimately will correlate to better preservation capabilities and asset maintenance, the author does assert that TDR standards "promise that TDR certification will make repositories' stakeholders more confident that the data they contain will be protected, properly managed, and available for future reuse" (p. 2).

Management and infrastructure
The management of DRs and the preservation of digital assets are inextricably connected. Programs surrounding digital initiatives benefit from planning documents that assert institutional commitments to preservation, articulate policies, and delineate the technical standards and benchmarks that informs the continuum of its projects. A planning framework typically includes goals, objectives, outcomes, and descriptions of evidence that demonstrate the institution's investment of resources for managing its digital initiatives. Collaboration in crafting this type of document connotes transparency to internal stakeholders and fosters a sense of reliability to external stakeholders. Preservation activities particularly need to be grounded within a stable, well-supported infrastructure to ensure long-term requirements are met. McGovern and McKay (2008) provide a methodical overview of the historical linkage between them and suggest strategies for leveraging their common objectives for long-term efficiency and effectiveness.
Planning considerations for responsible management and maintenance of IRs are extensive. The expansive scope of preservation in the digital asset lifecycle can often be overlooked when building DRs. As noted by a leading conservation center, "A welldesigned digitization program includes not just digital capture but also appropriate care and repair of original materials and long-term management of the digital files it produces" (Northeast Document Conservation Center, 2021). Ross (2012) addresses the importance of preservation planning in the methodological foundations of digital libraries. "It is about maintaining the semantic meaning of the digital object and its content, about maintaining its provenance and authenticity, about retaining its 'interrelatedness' and about securing information about the context of its creation and use" (p. 45).
The value and costs of IRs are topics minimally explored in professional literature. Stein and Thompson (2015) conducted a survey of institutional motivations behind migrating digital asset management platforms and found that many institutions wished for more local control over their systems and often chose to migrate to open source software. Burns et al. (2013) discuss the challenges of trying to study, calculate, and articulate findings in these areas citing a lack of standardized use of IRs across institutions. According to the authors, "A surprising find was the near equal median expenses between annual operating costs for institutions that use open source software and institutions that use proprietary solutions" (Burns, Lana, & Budd, n.p.). Arlitsch and Grant (2018) expound on this theme, suggesting a more efficient, national IR has been unachievable as libraries are reluctant to "abandon local control" (p. 264). The University of Houston Libraries formed a "Digital Asset Management Systems Implementation Task Force" to perform an environmental scan of repository system options that involved conducting a needs assessment and system testing (Wu et al., 2016). Hardesty and Homenda (2019) describe a migration to Fedora and argue that updating digital repository software encompasses an "ecosystem change" that includes considerations for end user interfaces, collection management tasks, and data structures.

Professional development and training considerations
The inherent progressive nature of DRs necessitates staff to obtain and demonstrate proficiencies in digital standards and practices, and to maintain currency as these requirements change over time. Examples of the types of varied education and training areas are: project management; systems implementation, upgrade, and migration; metadata creation; digital curation; copyright; and digital forensics. Locally-managed repository and digital asset management system migrations are complex and entail investment of library staff time. Many libraries used the "build-your-own model" and customized open source repository softwares such as DSpace, and Fedora to fit their local needs (Benchouaf et al., 2016;Gilbert & Mobley, 2013). Morton-Owens, Hanson, and Walls (2011) detail the skills needed to develop open source software solutions in libraries at New York University Health Sciences Libraries (NYUHSL); they include having "solid programming and database experience" and also institutional capacity to run separate library servers versus using institutional servers. Colorado State University (CSU) migrated from DigiTool to DSpace because of a historic lack of major system updates to DigitTool and evolving user needs at CSU. Software developers at CSU used the DigiTool's API and wrote custom scripts to export and convert data to DSpace SAF format for ingest (Benchouaf et al., 2016). Wu et al. (2020) describe the development of the Bridge2Hyku Toolkit, a suite of digital asset migration strategies and open source software to help libraries and cultural heritage institutions migrate digital collections to the turn-key open source digital system, Hyku. Gilbert and Mobley (2013) chronicle how a regional cooperative digital repository system migrated from a licensed, selfhosted installation of CONTENTdm to a custom built Fedora Commons 3 repository.
The consistent use of metadata fields and controlled vocabularies has a direct impact on the functionality of DRs and effectiveness of information retrieval. Training in the fundamentals and in emerging standards for describing digital files may include coverage of the different types of metadata (administrative, descriptive, technical), content standards, and the divergences between metadata and cataloging. These skill sets are needed to accurately describe assets and to create access points to them. This work also impacts the progress of migration work. Neatrour et al. (2017) discuss how metadata remediation and assessment are important considerations to digital asset management migrations. The Samvera repository MODS to RDF Working Group (2019) issued a white paper recommending metadata crosswalks from MODS to RDF predicates in 2019, which Hardesty and Homenda (2019) argue is integral to migrating descriptive metadata with digital assets to the Fedora 4 repository. Tani et al. (2013) survey how libraries mitigated metadata quality issues in repositories and described different proposed frameworks to characterize quality assurance and current remediation approaches. Kenfield (2019) provides an overview of metadata documentation in IRs of ARL (Association of Research Libraries) members, underscoring metadata as a means to communicate transparency and trustworthiness to users. Mering (2019) describes a case of how the use of personal names and Open Researcher and Contributor ID (ORCID) identifiers are utilized in metadata practices to standardize names and mitigate inconsistencies among them. Simons and Richardson (2012) posit from their survey results that gaps in knowledge exist due in part to changeability of professional competencies. Among their conclusions are that the skill sets required for IRs are often attained as the work is performed rather than through formal academic coursework. Consequently, staff would benefit from continuous professional development and training sessions based on specific needs. Staff can also be incentivized by supervisors and managers to enlarge their knowledge of IRs. Attebury (2018) outlines the role of administrators in academic libraries and how they can take an active role in facilitating continuous learning among staff. A key objective for crafting a culture of learning is to create "an organization capable of both responding to and anticipating change" (p. 429).

The problem: Reconciling legacy assets for migration with organizational readiness
In a meeting with SBUL faculty librarians, library administration announced DS6 would serve as the new repository platform for managing all digital assets: master, preservation, and access files. At the conclusion of the migration, SBUL would potentially have three repositories: DS6 for asset management, Omeka S for exhibitions, and Be Press for scholarly works. However, without a strategic plan, timeline, and defined staff roles with procedures, the readiness to execute the migration was unclear. A working group of library faculty and staff was established to participate in this project. Selection was based on organizational reporting structure and not expressly by the skill sets needed to complete the work, as per TDR recommended practices (Consultative Committee for Space Data Systems, 2011).
Leading up to the migration, repository systems had not been receiving major system updates and patches. Concurrent to the repository migration, SBUL carried out a pilot project to reprocess digitized art exhibition catalogs held in DS4. An objective of this project was to assess and evaluate metadata quality and interoperability issues that could arise during the larger data and metadata migration to DS6. The pilot project revealed gaps in training and knowledge that resulted in internal and external expectations not being met. Examples include voids in project management, software updates, server maintenance, metadata standards, and preservation activities. Compounding these exposed inconsistencies were issues of adhering to timelines and mutual priorities.

Methods
For this qualitative and quantitative research, the authors used three sources for evaluation. The criteria was taken from Appendix C, "Minimum Required Documents" of the Trustworthy Repositories Audit & Certification (TRAC) Criteria and Checklist (Center for Research Libraries (U.S.) & OCLC, 2007, p. 81). The rating scale from 0 to 4 is the Digital Preservation Coalition's maturity modeling tool, the "Rapid Assessment Model" (DPC RAM) rating scale: 0 represents "minimal awareness," 1 represents "awareness," 2 represents "basic," 3 represents "managed," and 4 represents "optimized" (Digital Preservation Coalition, n.d., pp. 7-13). The assigned ratings were informed by the findings of two SBUL rapid assessment digital preservation benchmark reports that also used the DPC RAM to measure the state of SBUL's DRs (Reijerkerk, 2020(Reijerkerk, , 2021. SBUL's digital assets librarian evaluated and reassessed the digital preservation capabilities of the libraries between 2020 and 2021. The authors created a table to organize the assessment process and to record findings. Columns A, B, C, and D represent the eighteen minimal required documents and examples of evidence specified in TRAC. Column E correlates to the criteria of the DPC RAM rubric for "Organizational Capabilities." Examples from the DPC RAM rubric were mapped to the SBUL evidence. Column F records the evidence and accounts for documentation, transparency, adequacy, and measurability. Evidence for the benchmarks included: statements made in meeting discussions with repository managers; availability of or lack of documentation; feedback on policies and procedures from content curators; review of the web presence for SBUL working groups (e.g., Digital Initiatives Working Group); an inventory of digital assets; a review of repository documentation from vendors; study of other R1 libraries digital asset processes; consideration of best practices and standards for DRs; and analysis of previously conducted preservation assessments. Column G is the rating scale. Column H has notation of suggested mitigating actions (Table 1).

Results
The total number of points for the rating scale was 72, if all 18 elements were rated at level 4 ("optimized"). SBUL attained a score of 25 out of 72. SBUL has achieved a score of 1 ("awareness") or higher in 16 of the 18 categories. The TRAC criteria of "A3.1" and "A3.5" which correlate to procedural accountability and policy framework rated at 0 ("minimal awareness"). "A3.1" can be improved by creating a written definition of designated community(ies) and publishing policies detailing SBUL's preservation strategies. A score of 1 ("awareness") can be attained for "A3.5" by creating a workflow to systematically solicit feedback regarding software and service adequacy. No attribute received a score of 3 ("achieved") or 4 ("optimized"). The highest scores of 2 ("basic") aligned with the DPC RAM category of "IT capability" within the TRAC criteria of "B. Digital Object Management." Discussion SBUL has many opportunities to expand its capacity, reach, and impact in the realm of digital initiatives. While there were many unmet benchmarks, the nuances among the scores exposed key opportunities related to policy and information technology that could be amplified to strengthen its DR infrastructure and to increase trustworthiness. Most essential and foremost is to develop a comprehensive planning document with a lifecycle framework for preserving digital assets. The preservation architecture should be technology agnostic. Fundamental considerations to address include purpose and aims, policies, designated community, procedures, and defined roles of participants. The objectives of the program can be mapped to measurable criteria that aligns with best practices and benchmarks as per TRAC and DPC RAM. These normative metrics would provide qualitative and quantitative data to evaluate viability, to perform internal assessments, and to support third-party audits. The framework developed during this present research can be built upon annually.
Migrations are innate in the lifecycle of digital assets. Regular reviews for currency in technological infrastructure would circumvent the need to reactively mobilize. Steps toward organizational accountability could be made by fostering a culture of continuous learning and with the expectation that staff will continually complete training in digital preservation. While libraries should invest in and emphasize employee development, free online programs do exist that can reduce knowledge gaps such as DPOE-N and Digital Preservation Coalition resources. Limited perspectives of repository development create technological relics, as has been the experience at SBUL with CONTENTdm. It is imperative to prioritize implementation and documentation of critical system updates and regular maintenance over secondary operations. The addition of a digital repository systems manager to administer operational needs, training, and support across platforms would be an impactful addition to library technology and organizational infrastructure. If these measures are not met, partnerships with external stakeholders and providers should be explored and negotiated, as per TDR recommended practices.

Organizational infrastructure
The retrieval and functionality of digital assets is dependent on sound organizational planning and an infrastructure driven by well-formed written policies and procedures, and with an institutional investment in and commitment to its sustainability and growth. Touchstones of this planning scope are articulation of fiscal transparency, feedback channels, and accountability. Evaluation of SBUL's current attributes against the TRAC "minimal required documents" checklist using the DPC RAM ratings suggests the current capabilities and readiness of SBUL in this area need improvement and additional resources to reliably store, migrate, and provide long-term access to its digital collections. A historical lack of a repository mission and goals combined with undocumented procedures, processes, and workflows have contributed to minimal usage by external stakeholders, technical obsolescence, and data loss. The forthcoming DS6 repository needs to accommodate multiple designated communities that presently are loosely and informally defined. The existing preservation benchmark reports and this assessment provide a foundation for comprehensive planning. The development of a strategic plan coupled with the effective communication and implementation of it has potential to bolster many attributes. In the short-term, crafting a transitional document with immediate priorities would serve as a prelude for a more comprehensive planning process for digital asset management. Yet to be formalized are policies within a planning framework for preservation activities, modes of acquisition, and legal considerations. The criterion for soliciting feedback (A3.5) received a score of 0 because there are no processes in place. Currently, SBUL's DRs are not configured to collect user analytics and to date, no assessments of software and digital service adequacy have been conducted. This area could quickly be improved with adding a department-level email address on a digital collection webpage or by creating and linking to a Google form for users to submit comments. Digital asset management includes activities of ingest, preservation strategies, archival storage, and access management. For criterion A5.5, which centers on policies for ingest and associated liabilities, the evaluative evidence shows minimal processes are in place. Challenges to rights are presently managed on an adhoc basis. While some rights-related metadata is recorded in item records, SBUL does not express to creators the parameters of rights and permissions, or how to transfer them. Verbatim language from the consortium-developed policies of RightsStatements.org could be used to develop brief guidelines and to communicate options to contributors about transferring rights. From a preservation standpoint, documentation about rights, particularly copyrights, should be embedded in each digital asset throughout its lifecycle to maintain original context and provenance. SBUL's DR infrastructure is not developed enough to meet current service needs. Finally, fiscal responsibilities and processes are not fully transparent. Criteria A4.3 (financial procedures) could be met by sharing budget sheets that account for anticipated allocations, expenditures, and revenue sources. This action would advance strategic planning particularly in capacity building, e.g., for acquiring software and hardware and for projecting needs in anticipation of future repository growth.

Digital object management
Digital object management would benefit greatly from participation of current staff members who possess technical expertise, credentials in the domain of digital archives, and historical information about legacy collections. In discussions and communications within the library, a persistent issue was conflation of "refreshing" (copying to another storage format or medium) and "migrating" (transferring content to a new platform or system, although the incompatibility issues or (meta)data loss might result). The processes are not the same. This misinterpretation has implications for the DS6 migration and future migrations. The process of moving (meta)data with the notion that once the data is moved, the migration is complete, does not consider the holistic managerial responsibilities of digital assets. Moving data from legacy repositories from one server (e.g. CONTENTdm) to another (e.g. DS6) is refreshing, not a migration. Refreshing, while a preservation strategy is not a "full service" strategy meaning that it is not enough by itself to constitute preservation (Cornell University Library, 2003). If the intent is to migrate the legacy repositories, the project working group needs to be equipped. One approach is to invest in digital preservation training, including the "Digital Preservation Management Workshop and Tutorial" developed by MIT Libraries which defines migration as a process much broader than simply copying data from one technology to another: "migration is a set of organized tasks designed to achieve the periodic transfer of digital materials from one hardware/software configuration to another, or from one generation of computer technology to a subsequent generation" (MIT Libraries, n.d.). While there are no universal definitions to preservation strategies, there should be organizational definitions of strategies that form the basis of repository management going forward.

Technologies, technical infrastructure, and security
Considerations for technologies, technical infrastructure, and security emphasize the critical processes for maintaining compliance of systems, assessing hardware and software, and creating procedures to survey and maintain technology. Processes for technology watches and media change (C1.7) were scored a 0 rating in conformity with guidelines outlined by the TRAC document. SBUL is the least prepared to meet this criteria at current levels out of all eighteen areas. If adaptation to new technologies is to be met, monitoring changes in hardware (C2.1) and software (C2.2) will require a commitment to acquire current hardware and software. A lack of updates required SBUL to migrate systems without a clarity of the potential outcomes and to contend with additional diagnostic work resulting from these lags. An area of heightened concern is risks of security breaches. Going forward, a program of regular systemic upgrades would ensure the servers are not compromised. This implementation would increase the scores of multiple criteria including C1.7 and A3.5. This mitigating step would also support achieving the next status level for A1.2 because it interfaces with a contingency plan and exit strategy. An up-to-date system is more prepared for a migration; it also alerts staff to software and hardware issues, of the imminent need for refreshing, and impedes bit loss.
In meeting sessions for migration work, there was limited understanding of DSpace software and minimal consultation of open source documentation (Lyrasis, 2021). This finding aligns with SBUL preservation benchmark reports which noted the organizational-wide lack of digital asset knowledge; no shared understanding of repository migration processes; and unclear current and future capacity of server space storage. Dodd (2019) asserts those in library management positions would benefit from fully grasping their subordinates' responsibilities and skills and that librarian competencies should "produce measurable results based on employee and patron feedback, and support productivity statistics" (p. 689). To better prepare staff to maintain DRs, a program of professional development focused on the specialized requirements of repository management (e.g., systems, platforms) and asset management (e.g., file management, preservation) needs to be supported.

Conclusion
Institutions facing DR environments similar to SBUL can model this study using a modified version of the TRAC list of minimum required documents and DPC RAM rating scale to assess the state of asset management and to identify areas that may require immediate attention. A fuller design process could then ensue, particularly at organizations where little documentation and historical information exists. Without a substantiated record of programmatic work, planning will require an interrogative approach that incorporates background research and use of survey tools. A judicious course of action that balances aspirations with capacity is recommended, as it demonstrates a commitment to the standards that embody a TDR and to the responsible stewardship of digital assets.
Digital repositories have created opportunities for libraries to lead and actively contribute to programs that ensure the long-term preservation of cultural heritage, research output, and scholarship. The services associated with them are rooted in providing context, content, and access and require "a balance of investment and emphasis among these three elements" (ARL, 2009). The sustainability of digital assets depends on investment in resources and requires reaffirming commitments to maintain them. This article advocates for integrating organizational accountability in the paradigm of digital asset stewardship. Striving to meet and adherence to criteria outlined for TDRs, regardless of whether an institution receives certification, is a proven way libraries can validate and communicate capacities to preserve and provide long-term access to its digital assets. Benchmark guidelines and self-evaluative assessment tools are sources to aid institutions in identifying strengths, weaknesses, challenges, and opportunities for managing digital assets held in repositories.
For both content creators and library staff, there is little merit in collecting digital content if the institution does not have the capacity and resources to preserve it. As Salo (2008) cautioned, "any library that cannot or will not start by developing a repository constituency with real commitment to depositing material on an ongoing basis, however, should not start a repository at all" (p. 119). Further, "An institutional repository is a useless excrescence unless it is part of a systematic, broad-based, well-supported data-stewardship, scholarly-communication, or digital-preservation program" (p. 121). However, by connecting digital asset preservation to mission-driven goals and objectives, libraries can demonstrate their value and engage external stakeholders to enlarge capacity. In this way, digital initiatives become embedded in wider institutional planning and foster development of cultures of continuous assessment and professional development. By envisioning a program for digital repositories grounded in the attributes of trustworthiness, inherited digital legacies can be transformed and as a result, so can the organizations that steward them.