Masters Portfolio §F:
Selection, Organization, & Preservation

Management and preservation of information items for current and future users represents the core of our work as archivists. By selecting an appropriate sliver of materials to steward, arranging and describing them to efficiently provide access through archival methods, and preserving them for the future, we fulfill our professional mission of providing access to the historical record while working within our limited resources.

Selection and Appraisal

As much as some people would like us to, archivists cannot preserve everything; we simply don’t have the resources to accomplish anything approaching that scale.  In fact, given our limited resources, we need to be judicious and strategic about what we take on for long-term preservation. As the total cost of stewardship movement has pointed out, acquiring material means shouldering the costs of processing it, describing it, conserving it, and storing it indefinitely in a stable physical or digital environment, among other potential tasks (Weber et al. 2021, 6). Every acquisition takes up valuable space, whether on an institution’s climate-controlled shelves or its preservation storage server.1 Thus, it is crucial to ensure that the items preserved are of enduring value.2

Archivists have two main sets of tools to help them decide what to accession, collecting policies and assessment rubrics. Collecting policies define what a repository is interested in acquiring and what it will not accept (Ham 1993, chap. 3). Such policies should be grounded above all in a consideration of the repository’s realistic capacity for care; in the words of Gerald Ham, “It is ethically irresponsible for the repository to accession what cannot be preserved under proper conditions of storage and made available for use” (1993, 16). They should also consider the prevalence of the sorts of records under discussion, the strengths and weaknesses of the repository’s existing collections, and the collecting priorities of other repositories (Ham 1993, 16). Collecting policies may be guided by one or more selection philosophy, such as institutional documentation or historiographical needs (Ham 1993, chap. 2).

While collecting policies are articulated ahead of time, assessment rubrics help archivists evaluate the material they find in front of them. These tools allow archivists rate on a quantitative scale various attributes of an aggregation of materials, such as quality of physical housing, level of existing intellectual access, and research value, expressed both through both how interesting the subjects of the material are and the quality of the documentation it provides (Philadelphia Area Consortium of Special Collections Libraries 2006; 2008; Columbia University Libraries, n.d.; Lyons 2011). These tools can be used not only to appraise incoming material, but also to evaluate existing collections as part of a collections assessment survey (Lyons 2011; Proffitt 2011; Velte 2025).

Of course, archivists appraise not just whole collections, but also the items within them, with varying degrees of granularity. Many of the decisions in this arena will be based on research value of the sort assessed by appraisal rubrics, though complicated by both the principle of respect des fonds and the impracticality of filling out a rubric for every document or folder (Cresci-Callahan and Staniunas 2025). However, there are tendencies in what value inheres (or does not) in certain sorts of material, and these can form the basis of useful rules of thumb (Cresci-Callahan and Staniunas 2025).

Organization: Arrangement and Description

Archival information organization practices are somewhat different from those used by most other information professionals. Whereas librarians employ bibliographic description methods that assign the same metadata fields to each item with more-or-less the same level of detail, archivists deal in aggregates, describing the hierarchical arrangement of materials in finding aids and relying on this arrangement to lead users to individual items (Wiedeman 2019; 2023; Miller 1990; Shepherd and Yeo 2003). Under this system, each item in the hierarchical tree bears metadata, with each element’s value inherited from the node’s parent if not explicitly specified by the archivist (Wiedeman 2023).

This practice is grounded both in principle and in expedience. On the one hand, this system allows archivists to preserve the context that gives an individual record meaning; the principles of provenance and respect des fonds protects this archival bond by directing archivists to maintain original order (Bearman and Lytle 1985). On the other hand, as Greg Wiedeman has shown ( 2019; 2023), these aggregate methods evolved out of archivists’ need to find some solution, however imperfect, to organize the overwhelming crush of material being handed to them. This was necessary not only because records are created in much greater volume than books are published, but also because most archival holdings are unique. Whereas a library buying a book can usually rely on cataloguing already done by another institution that holds a copy, archivists must describe all their records themselves, from scratch. Thus, archivists are far less concerned with precisely following exact cataloguing standards than with making sure their materials are reasonably well-described to enable access.3 In fact, over the last two decades, many voices in the field (the “MPLP” or “minimal processing” movement) have actively begun to advocate against working at the item-level when processing collections (Greene and Meissner 2005; Prom 2010; Mengel and Smerz 2013; Dundon et al. 2020).

At the turn of the century, there was some optimism that the affordances of digital records might allow archivists to organize digital records at the item level (Bearman 1996; Shepherd and Yeo 2003). However, the intervening quarter-century has shown this hope to be a mirage. The acceleration in the creation of born-digital records (Owens 2018, 190; Conway 2010, 62), combined with the unwillingness of most working people to pause in their main tasks and generate high-quality metadata when they create a record (Doctorow 2001), means that aggregate description is just as important to digital archives as physical ones. Indeed, digital preservation theorist Trevor Owens calls out archival methods (2018, 74), especially the MPLP movement (2018, 129–130), as uniquely suited to managing digital material; he advises that digital preservation work is best done “with a shovel rather than tweezers” (2018, 8). Moreover, some archivists are moving away from the bibliographic description of digitized material used by many DAMS in favor of integrating digitized surrogates directly into finding aids; this trend is epitomized in the draft specification of DadoCM released earlier this year (Foster et al. 2025). Thus, archival methods of organizing information are just as useful applied to digital objects as to physical ones.

Preservation

Preservation refers to “the steps taken to ensure the long-term accessibility and usability of content, including (but not restricted to) activities that prevent content from deteriorating” (Skinner 2022, 180). However, depending on how one defines “content,” one can distinguish multiple kinds of preservation, and I follow Trevor Owens in differentiating between artifactual preservation, which is based on historical contiguity of a specific object, and informational preservation, based on the information an object carries (Owens 2018, chap. 2).4 While the selection and intellectual organization of information objects is relatively format-independent, the shift from physical to digital has a much larger effect on preservation practices, so I will treat each separately.

When dealing with unintermediated physical items, artifactual preservation should be the goal in most cases. The materiality of such information objects is often rich in secondary value and can yield valuable insights about the circumstances of the object’s history when subjected to appropriate examinations. Preservationists should prioritize storing these items in appropriate and stable environments, especially with regard to temperature and relative humidity, to most efficiently promote their longevity (Willer 2022; O’Toole 1989, 23; Environmental Specification for the Storage of Library & Archival Material, n.d.; Nyberg 1987). Repositories should develop emergency response plans so that they are ready to salvage their materials in case of disaster (DePew 1991; Ogden 1999; Albright 2020; Lindblom Patkus 2021). Reformatting materials may be necessary in some cases (e.g., moldy papers or flammable nitrate filmstock that could harm the rest of the repository’s holdings) but should not be a solution of first resort for preservation issues (Baker 2001; Gertz, n.d.; Northeast Document Conservation Center 2007). Likewise, while microform or digital facsimiles can increase access and serve as a backstop in case of catastrophic loss, these surrogates should not be mistaken for substitutes and do not remove the need to preserve the original artifacts (Smith 1999; Baker 2001; Werner 2013; 2017; 2018; Pratt 2018; Fafinski 2022). The exception is audiovisual analog electronic records stored on magnetic media (i.e., audio and video tape), whose machine-intermediation and inherent vice put them in imminent danger of “degralescence” and which must be digitized to maintain future access (Casey 2015).

Digital preservation is a much younger and more complicated field, but I will do my best to condense my profession into a few paragraphs. Digital preservation focuses on informational preservation, but with an artifactual twist. On the one hand, the “computer museum” approach of maintaining the original physical carriers of digital information is hardly ever viable (Owens 2018, 54–58; Digital Preservation Coalition 2015b). This is not necessarily a great loss; lossless copying of information is fundamental to digital systems (Owens 2018, 54–55, 59–61). On the other hand, however, as Owens (2018, 54–55, 61–67) convincingly argues, the  “platform nature” (Owens 2018, chap. 2) of digital systems means that informational objects exhibit significant artifactual qualities, so there is some continuity between the practices of digital and physical preservationists.

From this perspective, digital preservation comes down to two main imperatives: preserve the digital object’s original bitstream, with its artifactual qualities, and ensure that a reasonably faithful rendition of that bitstream’s original rendering remains accessible. The former is accomplished by creating and maintaining multiple copies of the bitstream to provide redundancy and using checksums monitor these copies’ fixity and ensure their authenticity (Digital Preservation Coalition 2015c; Owens 2018, chap. 6; Kussmann et al. 2019; Faulder et al. 2018; Kirschenbaum et al. 2010). The latter is accomplished through choosing or migrating to sustainable file formats (while maintaining the original file), or through emulation of the original computing environment for the purposes of rendering (Digital Preservation Coalition 2015a; 2015b; Owens 2018, chap. 6; Halvarsson 2021a; 2021b; Mellor et al. 2002; Rosenthal 2015; Kaltman et al. 2025). 

I should acknowledge that there are several much-discussed models and frameworks for digital preservation that the foregoing discussion does not countenance, including the OAIS model (Consultative Committee for Space Data Systems 2012), the various standards for trustworthy digital repositories (Research Library Group and OCLC 2002; Jantz and Giarlo 2005; Center for Research Libraries and OCLC 2007; Consultative Committee for Space Data Systems 2011), and the InterPARES project (Duranti 2010). While I am conversant in these standards, I do not think they are ultimately useful to most practitioners. In the words of Trevor Owens (2018, 7–8, 80), “highly technical definitions of digital preservation are complicit in silencing the past. . . . Complex sets of requirements . . . obfuscate many of the practical things anyone and any organization can do to increase the likelihood of access to content in the future.” This “‘over-diagram-ification’ of digital preservation” has led to “many smaller and less resource rich institutions feeling like they just can’t do digital preservation.” In the context of preservation especially, we cannot allow the perfect to become the enemy of the good (Kopin et al. 2025).

References

  1. Despite what some may think, digital items take up storage space as well. Digital items that are to be preserved long-term need to be placed in a high-quality digital preservation storage environment (Schaefer et al. 2018), which is costlier to maintain than typical digital storage. Cloud storage, while potentially helpful, does not obviate this expense; it merely means that institutions rent storage space from vendors instead of maintaining it themselves (Rimkus and Schmitt 2024). Moreover, while the availability of storage has increased exponentially for several decades, the creation of born-digital content has kept pace, and it seems that this trend may (at least for the moment) be stalling (Owens 2018, 190–191). In other words, most concerns about cost of stewarding physical items for the long term apply to digital ones as well. See also the work of Pendergrass et al. (2019) on the environmental sustainability of digital preservation activities. ↩

  2. For a definition of this term, see the Dictionary of Archives Terminology, s.v. “enduring value”. For more on changing ideas about what “enduring” means to archivists, see O’Toole (1989). In keeping with O’Toole’s prescriptions, there is an increasing focus in the archives literature on reappraisal and deaccessioning (Greene 2006; Huggard and Uglean Jackson 2019; Buehn 2020; Martzahl 2020), which is also reflected to some extent in professional practices (Mummey and McDonnell 2025; Mummey et al. 2023). ↩

  3. This can cause problems for data aggregation for discovery. The EAD specification helps to mitigate this issue, though even that standard is so flexible as to cause technical burdens for those working with it (Wiedeman 2019; 2026). ↩

  4. Owens borrows these frames from Matthew Kirschenbaum’s concepts of the autographic and the allographic. For a similar distinction, see the discussion of “Permanence of information” versus “Permanence of original documents” in O’Toole (1989). ↩

© 2026 James Truitt. Powered by Jekyll