Editors’ Note: Today’s post is by Gali Halevi, MLS, PhD, Collection Development Director at CLOCKSS. Reviewer credit to Chefs Haseeb Irfanullah and Dianndra Roberts.
For many years, Digital preservation has often been framed primarily as a technical problem, with considerable attention devoted to file formats, storage infrastructure, redundancy, and preservation platforms. While these are important concerns, they do not fully capture what preservation actually does, which is determining what remains part of the scholarly record, and by extension, what future researchers will be able to see, cite, and build upon. Once viewed in this light, preservation stops being purely technical and becomes inherently selective. Within this context, decisions about what is ingested, how it is described, and which workflows are prioritized shape the record over time.
In cases where diversity, equity, and inclusion considerations are not explicitly a part of those decisions, existing imbalances in scholarly communication can quietly be carried forward. For example, well-resourced institutions, dominant publication formats, and English-language outputs are more likely to enter preservation pipelines while regional scholarship, community-produced materials, and non-traditional outputs are often less visible and therefore more vulnerable.
It is important to note that this imbalance is, in most cases, not deliberate. Much of it tends to emerge from how preservation systems are designed. Most preservation workflows rely on formal publishing channels, structured metadata, and standardized delivery mechanisms. Therefore, these approaches work efficiently for journals and books produced by established publishers. They are less suited to policy briefs, community reports, multilingual publications, datasets without persistent identifiers, or locally hosted platforms, and so, over time, preservation coverage begins to reflect the mainstream of the publishing ecosystem rather than the full breadth of scholarly activity.
How Injustice is Codified in the Scholarly Record
Metadata practices and requirements further shape what is preserved and how it is understood. While standardization is necessary and improves interoperability, it also introduces constraints. For example, contributor roles may not reflect collaborative or community authorship, affiliation fields may not accommodate independent researchers, and subject classifications may struggle to represent interdisciplinary or locally grounded knowledge. These limitations do not necessarily prevent preservation, but they can reduce discoverability and influence which materials are prioritized.
Selection policies also play a role. Preservation initiatives often prioritize content based on perceived impact, inclusion in recognized platforms, or alignment with existing ingest agreements, resulting in a tendency to favor established journals and publishers. Materials from underrepresented regions or emerging forms of scholarship may fall outside these pathways, even when inclusivity is an explicit goal, which means that limited resources for preservation follow the path of least resistance.
Geography adds another layer. Institutions in well-funded regions often benefit from mature repository infrastructure and participation in collaborative preservation networks. Others rely on smaller-scale or locally hosted solutions that may lack long-term guarantees. Without deliberate coordination, this creates uneven preservation coverage, where scholarship from some regions is structurally more vulnerable to loss.
A related challenge is the uneven awareness of preservation itself. In some regions, digital preservation is not widely understood as a distinct activity, and institutions may assume that hosting content online is sufficient. This gap becomes particularly visible during periods of conflict, political instability, or natural disaster. When infrastructure is disrupted, locally hosted repositories, journals, and research websites can disappear quickly, sometimes without backups or external preservation arrangements. In these moments, the absence of preservation planning is no longer abstract and is directly translated into the loss of scholarly records, institutional memory, and regionally important research. The risk is greatest precisely where resources and awareness are already limited.
It is also important to understand that differences in collection development practices across libraries, museums, and archives also shape these outcomes. Libraries typically build collections to support access and breadth and are guided by user needs and publishing output. Museums tend to curate selectively, focusing on culturally significant objects within defined collecting scopes. Archives, particularly digital preservation archives, operate differently again. They frequently preserve content at scale, often through agreements with publishers or platforms, rather than through item-level selection. For large-scale dark archives, such as CLOCKSS, this creates a particular challenge. Content flows into the archive through established ingest relationships, and expanding representation requires proactive outreach, flexible ingest models, and a willingness to accommodate content that does not fit traditional pipelines.
Community-led archives highlight both the risks and the opportunities since they prioritize local knowledge, multilingual materials, and culturally specific content that larger systems overlook. At the same time, they frequently operate with limited technical support and uncertain funding. The result is a paradox. The materials that broaden representation of the scholarly record are often those with the least secure preservation pathways.
Some efforts are beginning to address these gaps. For example, metadata schemas are being expanded to support diverse contributor roles and multilingual description, collaborative preservation networks are exploring ways to include smaller repositories, and there is growing recognition that non-traditional research outputs should be treated as part of the scholarly record rather than exceptions. These developments suggest a gradual shift from preservation as infrastructure toward preservation as stewardship.
Establishing Solutions and New Practices
Rethinking selection criteria is also part of this shift. Instead of focusing only on highly cited or formally published outputs, institutions can consider regional representation, community relevance, and disciplinary diversity. This does not eliminate the need for prioritization, but it broadens the definition of value. In addition, we should reconsider the workflows themselves as they can create unintended barriers. Automated harvesting tends to favor platforms with standardized APIs or deposit requirements tied to specific identifiers, which can exclude researchers without access to those systems. These constraints are often invisible until examined closely but, with small adjustments, can significantly expand coverage.
Taken together, these issues show that digital preservation is not just about maintaining access. It is also about shaping the future historical record of scholarship. If preservation strategies prioritize efficiency and scale alone, they risk reproducing existing imbalances. Integrating equity considerations encourages a more representative approach without compromising technical rigor. The goal is not to replace technical priorities, but to recognize that durability and representation are connected. While infrastructure ensures survival, inclusive practices ensure that what survives reflects the breadth of scholarly activity. Without both, preservation may succeed technically while remaining incomplete in substance.
Addressing these challenges will require collective action since no single archive, repository, or institution can ensure inclusive preservation on its own. Greater collaboration across regions, shared infrastructure, and coordinated outreach to low-capacity institutions will be essential. In addition, supporting organizations with limited technical resources, providing guidance on preservation readiness, and lowering barriers to participation can help broaden representation in the preserved record. A more inclusive preservation ecosystem depends on community effort, global partnerships, and sustained commitment to ensuring that scholarship from all regions has a secure place in the long-term scholarly record.
