DHQ: Digital Humanities Quarterly
Preview
2024
Volume 18 Number 3
Preview  |  XML |  PDF |  Print

Sustainability and Swedish Women's History: Digitizing Photographs from the KvinnSam Archives

Rachel Pierce  <rachel_dot_pierce_at_ub_dot_gu_dot_se>, KvinnSam, National Library for Gender Studies; Humanities Library, Gothenburg University ORCID logo https://orcid.org/0000-0003-3480-1474

Abstract

Digitization has traditionally been viewed as a method of preservation and accessibility facilitation. This definition has resulted in a sense that objectives of sustainability are somehow built into the process of digitization, allowing institutions to not thoroughly or systematically consider what digitization and digitized collections are, how to ensure access over time, and how these groupings of digital files relate to larger collection management. This paper redefines digitization as part of collections development, a broader set of processes with a long history within archival institutions. This definition allows for the use of Michel-Rolph Trouillot's framework for understanding historical production, placing digitization within an ongoing process that shapes access to and thus alters the historical record. The paper calls for a queering of the understanding of archival sustainability to enable the use of paradata in the development of digital archival description practices. This approach demands an understanding of archive collections as shifting and overlapping processes that interact with one another to form the malleable boundaries of what a cultural heritage institution calls its archives.

Introduction: Digitization and the Preservation of Women’s and Gender Archives?

1
A popular history decrying libraries' destruction of physical historical newspapers in favor of technified copies, Nicholson Baker's Double Fold felt like a lone digitization sceptic's jeremiad when it was published in 2002. Then, digitization was seen as a solution for the physical archive's vulnerability, a way of tearing down the institutional silos that circumscribe research, and a method of saving individual institutions loads of money. There were socio-political and economic reasons for this initial positivism. Those working at feminist cultural heritage institutions were perhaps particularly prone to see digitization as a low-cost opportunity to integrate women and gender issues into the broad sweep of historical memory [Withers 2015, 136]. However, as a raft of newer studies have demonstrated, the fates of the physical archive and the digital archive are interwoven [Pierce 2021], digital materials are interpretations of complex physical materials, not simple copies [Björk 2015], and digital documents and systems are far from invulnerable to deterioration and obsolescence [Decker 2020].
2
There is still little research on digital collection lifecycle(s) that often include rounds of digitization, database construction and upkeep, metadata application and restructuring, and material and database decay and retirement. Cultural heritage institutions must understand and plan for these lifecycles if they want to construct sustainable digital and physical collections, especially because the research grants that often fund the construction of databases rarely budget for digital archive maintenance and thus do not require a full description of how this maintenance will occur [Edmond and Morselli 2020]. Further, it is difficult to plan for emerging requirements. Examples abound, from institutions that must retroactively assess their digitized collections to align them with the FAIR principles, to current drives to translate controlled vocabularies into linked data. What if digitization is a multiplication rather than a mitigation of the archive's or archivists' habitual problems with sustainability?
3
This article will address the early 2000s planning for and construction of the image database at KvinnSam, the National Resource Library for Gender Research in Sweden, as well as the subsequent reassessments and potential futures of the database. The main question to be answered is: How can sustainability for digitized archival collections be defined if we understand archival collections as ongoing processes rather than static assemblages of documents? In order to answer this question, the article utilizes Michel-Rolph Trouillot's framework for the construction of history to periodize and analyze KvinnSam's photographic database as process:

Silences enter the process of historical production at four crucial moments: the moment of fact creation (the making of sources); the moment of fact assembly (the making of archives); the moment of fact retrieval (the making of narratives); and the moment of retrospective significance (the making of history in the final instance). [Trouillot 2015, 26]

KvinnSam initially digitized around 1,000 photographs, a fraction of its photographic archival material. Since then, academic research, digitization techniques, and KvinnSam itself have developed in different directions, rendering the selection, description, and technical infrastructure underlying the image database less sustainable. Understanding turning points in this process and planning for material collection and migration practices is critical to the development of long-term sustainable practices.

Literature Review: “Recovery Projects”, Photography, and Sustainability

4
Research on the digitization of photographs is particularly useful in uncovering and thinking critically about issues with digitization and sustainability. As recent scholarship has established, visual representations played an outsized role on the history and trajectory of many historical developments, including the campaign for women's suffrage [Lange 2020], a campaign that has laid the basis for many of KvinnSam's most used archival collections. Visual materials are also very popular on the web, increasing the chances that a wider public learns about topics like suffrage and becomes interested in the physical archive from whence these images come [Lesy 2020] [Vestberg 2013]. Further, institutions far more well-resourced than KvinnSam have focused on photographs as long neglected, chronically under-documented, often disorganized or subject to unique organization logics, and particularly vulnerable to physical deterioration — all issues that digitization can alleviate [Arnold et al. 2017].
5
Given all of these digitization positives, it is not strange that many institutions including KvinnSam initially prioritized digitization of photographic material. Here, it is worth thinking about the history of digitization and its role in the preservation of materials. Photographs are sensitive to light, to moisture, to fingertips, to use. They are constantly changing within the archives, losing definition and color, interacting chemically with the materials around them if they have not been stored individually in acid-free folders. Photographs embody the continuous change that characterizes archives, especially feminist and queer archives, even as “the archives” have become a metaphor for and symbol of stable, stagnant documentation [McKinney 2020, 191–193].
6
Digitization is not necessarily a tool that photographic scholars would include in sustainable photo preservation practices. The materiality of photography is considered central to some scholars, including photography archive theorist Joanna Sassoon; for these scholars, digitization strips these materials of their functional meaning [Sassoon 2004, 210]. Photographs also present unique problems. Katheryn Earle's analysis of a fashion photo digitization project highlights disorganization, the lack of labels, and the need to re-group photographs by theme and create contextual information that facilitates interpretation of these materials [Earle 2017, 51]. In other words, digitization is potentially reductive and most certainly highly resource intensive.
7
This is an area in need of a more nuanced understanding of sustainable preservation, which can provide ground for best practices development for local institutions with varying resources and collections. There is little consensus about how to define “sustainability” within the context of digital humanities projects [Edmond and Morselli 2020]. This is not surprising; there is little consensus about how to define sustainability more generally, though the term is used regularly within a variety of institutional contexts. Further, what counts as “digital humanities” infrastructure is constantly evolving, resulting in the baking in of digital competencies into cultural heritage jobs [Smiley 2019]. As a result, papers like Alex Poole's examine the “conceptual ecology of digital humanities” and use the term “sustainability” extensively but in a multitude of disparate and potentially conflicting ways.
8
Scholarly discussions of digitization and material sustainability individualize cultural heritage, treating documents as singular, remixable items whose use in digital environments can replace physical use. There is little discussion of sustainability and collections, though digitization of select materials implies the creation of entirely new digital contexts — archives, perhaps? — for documents. In general, scholars like Wolfgang Ernst contrast digital collections with physical collections, depicting the former as dynamic and the latter as static [Ernst 2013]. Yet some scholarship has demonstrated that digitization shifts and changes the physical archive, sometimes resulting in better description or different organization [Arnold et al. 2017], or, conversely, in the destruction of physical materials [Baker 2002]. Further, digitization can potentially produce vast quantities of metadata that are more reusable and sustainable than the digital files themselves [Scheltjens 2023].
9
Financial sustainability is often discussed separately from other kinds of project sustainability. However, these questions are particularly critical for institutions looking to ensure access to material illuminating the lives of understudied groups like women. Numerous scholars have demonstrated that digital projects about and by women and other historically disenfranchised groups have shorter lifespans than other similar projects, in part because the institutions and groups devoted to women's and gender issues have less well-secured budgets.[1] This situation has particular consequences for visual materials, an underused, under-described, and yet plentiful kind of source for archival researchers interested in the interwoven nature of women's public and private lives [Katz 2019].
10
In addition, sustainability is rarely connected to issues of representation that are so key to the theorization of historical materials. In the context of archives, discussions of sustainability have centered on democratizing access and providing opportunities for participation in the construction and content of archives. Yet the work of the Lesbian Herstory Archives illustrates the ways in which histories of description constitute representation and are imperiled as digitization demands “clean” data [McKinney 2020, 195–202]. Literature on archives and intersectional, care-conscious labor are imperative here, key to reframing sustainability as a multipronged process requiring the long-term preservation of materials, their histories of description, and their housing institutions, composed of both infrastructures and personnel [Caswell 2021] [Harris 2022]. This literature, in its emphasis on “good enough” approaches, also reframes sustainability as an insufficient end goal in and of itself. The real goal is sustainable access for communities whose need for information is imperative for their political and social visibility and viability [McKinney 2020, 202–203].
11
Digitalization and its sustainability have repercussions for the “assembly work” of history-making detailed in Michel-Rolph Trouillot's seminal Silencing the Past (1995), which notes:

Archives assemble. Their assembly is not limited to a more or less passive act of collecting. Rather, it is an active act of production that prepares facts for historical intelligibility […] They are the institutionalized sites of mediation between the sociohistorical process and the narrative about that process. [Trouillot 2015, 52]

This approach requires that we see archives as non-static, in constant dialogue with researchers, archivists, and other archives, digital and physical alike. Sustainability must be defined within this context, where the continuous production of historical knowledge is done by researchers toggling between past scholarship and the current state of archives, when access points are now almost exclusively digital.

Research Questions

12
The main research question for this article is as follows:
13
How can sustainability for digitized archival collections be defined if we understand archival collections as ongoing processes rather than static assemblages of documents?
14
An answer to the main research question will be built on answers to the following sub-questions, which involve a historical reconstruction based on internal documentation from the initial construction and continued maintenance of KvinnSam's image database.
  1. How can KvinnSam's digitization of photographic material be periodized, using Trouillot's framework for history-making?
  2. How does this process affect how an archival collection is defined? Are digitized documents new collections or copies of old collections?

Materials and Methods

15
The history of KvinnSam's image database will be told using internal documentation from KvinnSam, including meeting notes, preparatory studies, conference notes, and reports alongside an examination of the image database itself. The author's position as a research coordinator and archivist at KvinnSam should be noted, as this position provides access to internal discussions and problems that may direct attention to issues that are not immediately self-evident in the textual documentation analyzed here. It is important to note, however, that my employment began in September 2022, long after the majority of the history covered in this article.
16
The kind of documentation studied here is often referred to as paradata. This concept has been used in a variety of contexts, but within digitization, Dahlström and Hansson define paradata as documentation that covers “the process of digitizing and curating artefacts and documents, and the decisions made during a digitization project” [Dahlström and Hansson 2019, 6]. Though their work is largely interested in library digitization, their reasoning mirrors archival studies scholarship on digital and digitized collections in its focus on materials documenting collection-level decisions, from appraisal [Anderson, Eaton, and Schwartz 2015] to description [Yeo 2015] to platforming [Zhang 2012]. Such materials are important because they could provide long-term context for digitized documents like KvinnSam's photographs as they are moved from one digital environment to another, are transferred to new formats, and/or become obsolete.
17
Periodization of the database using these materials is a historical method that focuses on chronologizing and identifying turning points when work on or approaches to the database shift. This is a qualitative approach that focuses on textual material as a reflection of contemporary decision-making, with incompleteness of that documentation itself a sign of the rhythms of work with the database. Though Samantha Thrift has characterized this emphasis on rhythms and turning points as fundamental to research on women's movements [Thrift 2012], it is useful in analyses like this one, where the identification of thematic rhythms can be used to problematize the assumption that databases exist in three clearly delineated phases: construction, existence/maintenance, and retirement/archiving, an existence that only overlaps with the lives of physical archives in the first phase, during digitization of the documents that will populate the database.
18
Documentation will be used to retrace the history of the image database with a focus on identifying (1) circumstances in which the physical and digital archives were considered together and (2) circumstances in which sustainability was a central question for KvinnSam personnel. These emphases are choices; the construction of a historical periodization involves centering some conversations and issues and silencing other aspects of the past, as Trouillot's seminal work demonstrates. Following Trouillot's deconstruction of history's ever-evolving production and the creation of historical silences that history entails [Trouillot 2015], the image database and questions of sustainability will then be placed in conversation with questions of history as a construction with roots in the building and organization of archival materials that become (always incomplete) historical fact through their archival status and subsequent availability and use in the writing of history.

KvinnSam's Image Database: A History and a Periodization

19
KvinnSam is Sweden's National Resource Library for Gender Research. The organization, founded in 1958 by three women, consists of both a library and archives, and throughout its history KvinnSam has sought to make women's and gender research and history more findable and accessible to researchers and the broader public [Losman 1984]. The library has traditionally functioned as a major player in defining what counts as women's and gender research within Sweden through the creation and application of classification systems and the creation of archival collections that center women and gender as researchable subjects, all the while reacting to the use of these categories within contemporary scholarship.

Turning Point 1: Digitization

20
KvinnSam began digitizing materials in the early 2000s — the first identified turning point — as a response to both user demand and general pressures within the cultural heritage sector that related digitization to an institution's status as up-to-date. The earliest project focused on digitization of historic women's newspapers, due to popular demand and the declining physical quality of these materials. Images were chosen as the second digitization project, for these same reasons. Digitization would also solve issues of findability for photographic materials. Librarians would no longer have to search for images within the physical KvinnSam collections, where photographs were (and still are) under-described and thus difficult to find [KvinnSam 2001a]. In this sense, the library and archives followed the “mass digitization” imperative described by Thylstrup, an imperative that created a sense that KvinnSam needed to catch up with other institutions [Thylstrup 2018]. KvinnSam librarians thus chose materials that had received heavy use or were related to regularly researched topics, questions, and people [Backlund and Sjödahl 2001].
21
Despite the sense of catch-up embedded in this process, the stage of early digitization can be read as both the creation of sources and fact assembly according to Trouillot's model. These sets of early choices made some materials very easily accessible, thereby rendering less accessible materials almost invisible. KvinnSam's digitized periodicals are far from complete or even representative of its physical collection of periodicals. Further, this initial stage of digitization established a set of themes and material types that would continue to be prioritized in the following decades. These unwritten rules for what was important enough to digitize border on the third stage of Trouillot's framework, the creation of narratives. And crucially, the moment of fact creation was built upon past use, or historiography — a looping of the process of historical creation that Troillot's model does not address. This looping, the consistent impact of historiography on digitization practices, is poorly accounted for in studies of digitization, representation, and access.

Turning Point 2: Platform Choice

22
A small working group at KvinnSam was formed, and the group's two members conducted a series of study visits to digitizing institutions, largely within Sweden [KvinnSam 2001b] [KvinnSam 2001c]. This working group attempted to think strategically about the future needs of KvinnSam and future technological advancements that might render obsolete the collection under digitization or change the ways that patrons search for photographic materials [KvinnSam 2001d]. Planning was, however, tightly restricted by both technological siloing and resource limitations. Database technologies available to KvinnSam in the early 2000s were often constructed with other kinds of materials in mind. Five different database infrastructures were under discussion during this period, and several were discarded as options because they were only appropriate for museum materials or bibliographic information [KvinnSam 2001e].
23
After an initial investigation, the working group decided to prioritize the construction of a bespoke database that could function as a part of the institution's own digital infrastructure, largely because other external options seemed too expensive, not suited to photographic documents, not suited to KvinnSam's subject word indexing system, or difficult to upgrade [KvinnSam 2001f]. Further, the database was constructed both beside and as an integrated part of themed historical portals designed to present basic women's history grounded in the history contained within KvinnSam's archival collections. The interwoven nature of these digital projects would then influence the trajectory of the image database.
24
This decision marked a second turning point, narrowing some options while expanding others and, in effect, creating the guidelines for new archives in digital environments. The choice positioned these digitized documents within the digital infrastructure of a chronically underfunded institution. However, housing the database also meant that KvinnSam clearly owned their digitized materials, cementing them as located within Swedish women's history, as defined by the broader KvinnSam website. As noted, extensive portals with key themes from women's history were constructed alongside the image database, and this co-design facilitates searches based in predetermined themes emerging from women's history: suffrage, peace, education, work, and the modern women's movement. A separate database also presents the photographs as a searchable, coherent collection [KvinnSam 2001a] [Backlund and Sjödahl 2001].

Turning Point 3: Selection for Inclusion

25
Selection can also be viewed as a turning point that narrowed the digitized collection, the online image of KvinnSam's archival holdings, and KvinnSam's definition of Swedish women's history. These decisions were not determined in a vacuum; rather, selection choices were affected by KvinnSam's preexisting understanding and periodization of history combined with concerns about legality [KvinnSam 2001a]. The initial choice to focus on materials from the suffrage movement in Sweden — bridging the turn of the 20th century — stemmed from the presence of many known photographs in the archives (meaning there was enough metadata to describe individual items), researcher and journalist demand for these items, and a belief that they were less likely to suffer from confusing rights issues [Backlund and Sjödahl 2001].
26
The group settled on a set of prioritized metadata that reinforced these foci: names of persons, locations, chronological information, activities, and clothing. These choices were built out of the thesaurus that constitutes one of the bases of KvinnSam's work, a woman-centered subject word list that supplements the Swedish subject word thesaurus [KvinnSam 2001a] [Backlund and Sjödahl 2001]. As previous research has shown, the keywords focus on personal and organizational names, years, places, professional positions, and events, alongside clothing descriptions and other user demands that reflect documented non-researcher interests in historical fashion [Pierce 2021]. These keywords can provide a comparatively representative blend of public and private information, approximating some of the messiness of documents contained by the physical archives [Pierce 2023].
27
This process fits with Trouillot's depiction of history-making, which includes researcher use (and re-use) of collections. The approach taken by KvinnSam followed the trend of digitizing “canon” materials that has characterized digitization since its inception, with definitions of the cannon based on past women's history research. Indeed, researcher use characterizes the whole of KvinnSam's current website, formalized in the institution's provision of historical portals that centered a research generation's focus on the “waves” of feminist activism — a structuring device that is currently viewed as problematic and limiting as well as generative [Reger 2017]. While boxing in digitization decisions, a ready-made portal also meant that these materials could be linked to a preexisting historical context, in a sense taking the place of a finding aid in terms of situating the materials within historical processes. What is missing from Trouillot's story is perhaps the larger national superstructures that box in the behavior of cultural heritage institutions. The initial count of 2,000 digitizable images shrank to around 1,000 by the time the database was fully realized, largely due to a questions of ownership and copyright. The construction of digital materials and a database occurred outside of KvinnSam, however. A digitization team at the library performed the digitization, and KvinnSam staff added metadata to these items. This division of labor meant that staff did not directly control their materials, which put them in a unique position. Nor was KvinnSam directly involved in database construction [Backlund and Sjödahl 2001].
28
The photograph database became fully functional in the early 2000s.[2] The interface for the database is simple, with a keyword area and a drop-down list of all the possible subject words used within the collection. Results are presented as a list underneath this search area, with thumbnails of the digitized photographs on the left, clickable titles for the photographs on the right, and the first lines of their descriptions presented below. Users can then view one image at a time, clicking on the linked titles to arrive at a page with a JPEG version of the digital document located above a set of metadata, which includes a plain text description of the photograph, a signum for the archive holding the physical version of the document, and a set of hypertext keywords that provide a certain amount of remixability.[3] To access the high-quality TIFF files, the user must click on the image itself. The database is also integrated with each themed portal, with a left sidebar button labelled “Photographies” for accessing the database interface within the structure of each portal.

Turning Point 4: Shifting Databases

29
This database was maintained until the Humanities Library at the university agreed to begin using Alvin[4], a multi-institutional database for digital and digitized cultural heritage materials that was developed and is maintained by software engineers at Uppsala University. Adoption of Alvin, which went live in the mid-2010s, represents a turning point. The platform was seen as more modern, more flexible, and, crucially, less resource intensive for KvinnSam (though the library system pays a large fee as the leading consortium member), leading KvinnSam personnel to upload all digitized materials to Alvin. However, the materials from the image database mean something different in Alvin, where they intermingle with other digitized materials and finding aids for archives from a variety of Swedish cultural heritage institutions. As a result, the digitized photographs from the KvinnSam database can no longer be read together; neither Alvin's search functionality nor KvinnSam's metadata facilitate finding the initial digitized photo collection as a group.
30
The shift to Alvin initiated a period of stagnancy for the image database. For the last several years, KvinnSam has updated materials in Alvin rather than attending to the bespoke database. This shift in attention is the result of a long wait for a new interface, including a new website that will not include the database, largely for two reasons: (1) it is presumed that the database cannot be cleanly integrated into a new website, and (2) staff have not regularly updated materials and metadata in the database, increasing the amount of work that would be needed to move and integrate it into a new website. Materials are occasionally removed from the database or Alvin due to copyright concerns, and new digitized materials are uploaded in Alvin rather than in the photography database. In both cases, subject words have not been updated, even if the subject word list has undergone change. Because KvinnSam has created archives without recognizing that it has created archives, there is no clear contextual information that can hold the archive together and assist with its migration.
31
Migration has also led to some problems with archival representation. As noted, Alvin contains both archival finding aids (i.e., descriptions of entire archival collections that are often dozens of pages long and contain provenance information, chronological data, and descriptions of the items in each physical box) and individual digitized items from these collections. This diversity may eventually allow for interlinking that will provide needed archival context for digitized materials, but this feature is not yet fully developed. Further, many institutions, KvinnSam included, upload their finding aids as attached documents, which are not searchable via the Alvin search function. This hinders future development of platform design that provides users with clear information about where digitized materials come from and how representative they are of much larger physical archives. Lastly, Alvin covers multiple institutions, all of which use the platform differently. This broader diversity means that the similarly diverse descriptions within KvinnSam's own collections, which were produced through multiple generations of description, digitization, and platforming, are not as visible as they might otherwise be.
32
The shift to Alvin is a turning point for another reason: using the platform clarified issues with KvinnSam's use of and responsibility for archival data and metadata, a responsibility that was assumed with library materials but never really implemented for archival collections. Currently, KvinnSam staff are focused on metadata for archives — both finding aids and digitized or born-digital materials — as a central plank in the creation of future archival findability. A sustainable filing and preservation strategy is now seen as a fundamental building block for the creation of a more sustainable photograph database. The compilation of materials will also allow KvinnSam personnel to establish which tags will be more useful in day-to-day image work, which includes planning for future digitization and the creation of digital exhibitions that showcase KvinnSam materials, competencies, and services. The creation of this alternate system of images and metadata is designed to supplement rather than erase or displace the previous classificatory system applied to individual images, which reflects the history of terminological shifts within Swedish women's and gender research (for descriptions of a similar approach at the Lesbian Herstory Archives, see [McKinney 2020]).
33
In addition, new digitization efforts are planned but organized very differently than was this first push. KvinnSam is currently working in a more project-based method that is less concerned with copyright and more concerned with sustainable practices that have been built up since the construction of the initial image database. These practices include a focus on (1) usability in multiple digital environments (for a new webpage, in digital exhibitions, etc.); (2) the creation of new access points for underused collections and materials; and (3) physical material sustainability, with the potential for more attention to damaged, fragile, and/or underdescribed materials. It is likely that the current image database will be retired when KvinnSam's new webpage goes live, with its contents republished in various locations on the new website, embedding the legacy of the database in the digital infrastructure of KvinnSam while at the same time potentially making that past invisible to users.

Discussion: Sustainability for Digitization Projects

34
Cait McKinney has noted that cultural heritage institutions are affected by a “digital imperative”, or the assumption “that digitization is a necessary, inherently progressive process archives must carry out to stay relevant and preserve their holdings” [McKinney 2020, 159]. What KvinnSam's experience with its image database demonstrates is that reliance on digital infrastructures — whether in-house or outsourced — places materials at risk of disappearing or being rendered unusable due to poor organizational practices. Yet this risk was only visible at the turning point when materials were moved between platforms. Platforms multiply the number of organizational and technical skills required by an institution, rather than relieving librarians and archivists of work. Investment in expensive platforms also refocuses work towards digitized materials, and digital access renders the role and work of archivists invisible to researchers.
35
The development of KvinnSam's digital and physical photographic archives complicates Trouillot's framework given that digitization alters the physical and digital archives and their sustainability while (often temporarily) extending accessibility of some documents. Digitization makes it clear that Trouillot's framework is not linear and that the making of narratives and history affects already constructed sources and archives, altering them and their accessibility and use over time. In KvinnSam's case, sustainability is more profitably linked to the construction and use of paradata in the development of digital archival description practices. Taking the role of history-making into account, sustainability should be linked more to transparency and a willingness to see digitized materials not as copies but as extensions of physical collections, affected by conditions of provenance and collection development.
36
A comprehensive and forward-looking preservation strategy that encompasses all materials — not just digitized items — is crucial for cultural heritage institutions and archival researchers. Good, structured organizational practices are dependent on resources, broadly defined. The multiple definitions of sustainability within digital humanities literature obscure the fact that financial, infrastructural, and intellectual aspects of institutional health are intertwined. Institutions and centers with unstable or chronically limited funding are most at risk for landing in this kind of disorganization. During its many years with a smaller staff, KvinnSam did not work particularly actively with its archival holdings. Maintenance of digital materials is built on the maintenance of physical archival resources; one cannot happen without the other [Arnold et al. 2017]. Arguments that digitization will result in the preservation of physical materials are simplistic at best and misleading at worst. Certainly, digitization can be integrated into a larger preservation strategy, but digitization is not de facto preservation. Sustainability requires viewing digital and physical collections alongside practices as a whole.
37
An issue is defining the “whole” — in order to generate sustainable practices, an institution must define precisely what it is sustaining while also remaining attuned to the fact that all archives have lifecycles. Digitization is a process that creates collections somewhere in between new and old, physical and digital; these are collections characterized by their contingency [Pierce 2021]. What the periodization of KvinnSam's image database demonstrates is the centrality of platforms in determining what materials are digitized, how they are digitized, and what is considered appropriate description. As Johanna Drucker has argued, “Most information visualizations are acts of interpretation masquerading as presentation” [Drucker 2014, 10]. Institutions need to acknowledge this structuring when they start using new platforms. Each new platform or database means new collection development, because archival materials are defined by their interrelationships and larger contexts [Bak 2016]. Each new platform also means shared responsibility and a reliance on collaboration. This sounds good, and such collective work can be highly beneficial. But the “whole” becomes very difficult to define in these instances. What is a collection in digital environments? When and how are the stages of digitization affecting the production of history? How can the ongoing process of digital collection development be made visible to researchers?
38
With digitization, archivists, researchers, and “traditional” archives run the risk of losing a sense of what an archive actually is and where the boundaries of their own collections are. The rhetoric and logics of digitization and digital tools push archivists towards a conflation of “information management” with archiving [Thylstrup 2018]. This is not necessarily a problem, as D.M. Withers notes; how we understand and define an archive must and should shift as the technologies of its organization and description change [Withers 2015]. But change is difficult and it is often viewed as a disruption of rather than a development in archival practices. The splitting of digitization and collection management work between archival and IT departments means that collection custody is divided, resulting in a porous archival boundary. The multiplication of platforms and the transfer of digitized materials from across digital spaces upsets long-held practices concerning the multiplicity of documents within collections (as institutions often define a maximum number of copies to be kept) and collection-based description practices focused on provenance and original order.
39
As a result, digitization is often viewed as an adjacent service or a time-bound development project, rather than a central part of collections development. This perspective places digitized collections outside the process of archiving and history-making that defines physical collections management. The sense of archival abundance held by many members of the public, academics, and cultural heritage institutions obscures the fact that most archival documents are not and will never be online, as well as the issue of representation and continuous maintenance that characterizes all collections, including digital ones. Platform multiplication contributes to the multiplication of a few digitized documents, making them seem more multitudinous than they are and veiling them with an aura of representativeness [Odumosu 2020]. Digitization choices matter and need to be visible; these choices are integral to the sustainability of collections housed within databases, which in turn has repercussions for the sustainability of the larger project of history-making. Each new platform is a “moment of fact assembly”, in the words of Trouillot, creating new silences in the historical record [Trouillot 2015].
40
In order to assure the sustainability of digitized collections from these institutions and thereby women's and gender history, digital collections must be considered as discrete but interconnected archival collections within a system of physical and digital collections both at individual institutions and within a broader Archive. The KvinnSam image database demonstrates that the making of sources and archives, those first two history-making moments, are not singular but can recur over time as sources are made and unmade, collections constructed and then deconstructed. Digital materials support arguments that archives are never stable, that archives might best be viewed as processes and sets of ever-shifting relationships [Lee 2017] [Pierce 2021]. Archivists shuffle and relabel and add materials, documents deteriorate or are rendered unusable, and new accessions reframe pre-existing collections.
41
Digitization adds, in this reading, another layer to this processual archive — digital collections must be integrated into a broader system of collections management that includes collection description to ensure archival and historical sustainability. The migration of collections from platform to platform makes this need clear. In the case of KvinnSam, the initial collection context did not travel with the collection of photographs. When the new website for KvinnSam is built, the suffrage portal that provided historical context for the collection will disappear along with the original database, leaving a set of items with historical archival connections — materials from the only women's and gender library and archives institution in Scandinavia — dispersed in Alvin. Further, the summary of and selection information for the collection will disappear, leaving materials unmoored from the provenance information that gives these documents their archival weight, ensures the archival transparency central to feminist archivy, and offers entryways into Swedish women's and gender history.
42
There are opportunities to rethink archival description by reframing archives as “the archival body”, a queering of the Archive suggested by Jamie Ann Lee. Accepting instability as a natural part of collections with shifting identities — identities that now exist across the physical/digital divide — requires new methods that allow for descriptive multiplicity. Lee phrases this as a need to “understand and represent the dynamic (un)becomings” embedded in archival collections [Lee 2017, 1]. Lee's “Queer/ed Archival Methodology” centers archives as multitudinous processes [Lee 2017]. For KvinnSam, an archiving of both the database and the portals that provided context for database materials must be considered together, as processes that are part of both the history of Swedish women's history and the history of KvinnSam. Description of these collection processes requires understanding collections as constantly overlapping and shifting, in need of levels of description that ensure transparency and make visible collection boundaries that influence interpretation and the writing of history.
43
Here, existing paradata detailed in the results can help in the creation of finding aids for digital/digitized collections that move away from chronic debates about the “authenticity” of digitized documents (for this debate, see [Mak 2012]). Archival finding aids are defined by the Society of American Archivists as “a description that typically consists of contextual and structural information about an archival resource” that establishes the material's provenance and historical authenticity [SAA n.d.]. Scholarship on finding aids for digital collections has focused on born-digital collections [Ruest, Fritz, and Milligan 2022], but digitized groupings of documents are also in need of this level of description. Documentation of broken links and removal of documents due to rights issues, item additions, and the capture and archiving of digital environments is important; if digitized documents are to be used as historical evidence, their creation, management, and use over time must be considered [Owens and Padilla 2021].
44
Description at the collection level provides flexibility. Information about collection creation and management would help users to understand digital archival silences — materials lacking description that have simply been labelled “photo, unknown woman” and the absence of materials from certain collections that have rights holds — and integrate these silences into their understanding of historical evidence and process. At the same time, these collections would also offer opportunities to digitize underdescribed documents and describe them as a group, making it easier to digitize the less well-researched materials that exist, hidden, in the corners of archives. Further, collections-level description provides opportunities for integrating archival and researcher interpretations of materials, rectifying problems that scholars have identified with, for example, queering the archives [Freeman 2023].
45
For users, digitization and the lack of visible archival structure for digitized collections can result in a failure to see these the boundaries of archives and archival collections as important. This is dangerous; as Troillot notes, collection boundaries and the silences that are created in the process of building archives are central to history-making. Digitized collections actively resist silences, which leads researchers to assume that online databases offer a kind of completeness that they most often do not. These materials are instead the result of decisions based in available technical options, a digitization imperative, limited resources, previous researcher use, and readily available metadata/access within the physical archives themselves. Current research points to the usefulness of paradata for “upcycling” historical data, a more sustainable approach to historical research in digital archives that also centers the labor of historical construction [Scheltjens 2023].
46
An answer to the question of how to define sustainability for digitized collections focuses just as much on “collections” as “digitized”. This approach fits with scholarship on archiving photographs and the preservation of photographic meaning as multitudinous and contextual. In this vein, Joan Schwartz asserts that if a photograph's meaning is, as many have argued, grounded in context, “each use must be understood as distinct, though sometimes interrelated” [Schwartz 1995, 52]. Thus, a photograph that simultaneously exists across multiple institutions should not be interpreted as “duplication” or “waste” but “as the logical outcome of the appropriation and re-appropriation of a photograph with fixed content and physical configuration into different functional contexts with the attendant transformation of a single image into multiple documents” [Schwartz 1995, 52]. Questions of sustainability in this context must understand collections development as non-neutral and constitutive of histories via processes of selection and collection construction, maintenance, and even decay. Critical to this model of digitized archival sustainability is a continual questioning of digitization and digital collection-building, given the resource intensiveness and historical world-making that such practices involve.

Conclusion: Sustainability for Women's and Gender History Materials

47
The particular multiplicities and systematic fragility, or unsustainability, of collections of digitized gender and women's archival materials, as observed by [Boyles 2018], affects historical production at each and every one of Trouillot's four moments. The (un)sustainability of digital collections development is closely tied to the resources allocated for collections development and description over time; institutions that have trouble finding the resources to process and describe their physical collections will have trouble dealing with digital collections development in a systematic way. Digitized collections often emerge when a sudden influx of funding appears, frequently because of ongoing research in a particular area combined with the digital imperative. These circumstances demonstrate the non-linear, iterative nature of history-making, as archives evolve alongside research, each affecting the other. Making this relationship visible to archivists and researchers alike is critical for the sustainability of digitized collections.
48
Digital and digitized collections are not simply an addition to an older set of collections development. They require a careful system of curation and collection description that most institutions and researchers have not considered at length. Institutions need to consider the fact that database creation is in itself archival collections development, a process that requires the creation of description that elaborates upon the relationships between collection items and establishes the historical provenance of the collection as an ever-shifting whole. Researchers must interrogate the boundaries and absences created in the process of digitizing some things and not others at particular points in time, with particular resources at hand, and with particular goals in mind. Here, supplying paradata that makes visible the boundaries of and silences within digitized and born-digital collections is critical. This data can hold materials together when they inevitably must be migrated, and it offers researchers a historical dataset that has clearer usage value because of its more concrete provenance.
49
There is space for learning from the foibles of earlier digitization projects, as KvinnSam develops an approach to the eventual acquisition and management of born-digital archives. Sweden is behind in this arena, and its archives, libraries, and museums are preparing with some trepidation for the eventual influx of born-digital archives [Johansson 2023]. Some of the same pressures on specialized institutions remain. It is no accident that the few experiments in e-archives have centered on well-known men; women's and gender archives are constantly placed in a position where they feel they must catch up to international trends (for an overview of trends, see [Ries and Palkó 2019]). Yet these institutions are sitting on digital archives where experimentation can take place. By going through KvinnSam's own born-digital archives and learning from its management of digitized materials, staff can develop a policy and set of sustainable practices for newer archives that will inevitably combine physical and digital documents. The treatment of born-digital archives as requiring fundamentally different practices that will need independent development contributes to the sense that the physical archive is stable. Digitized archival collections offer an in-betweenness that may be highly generative.
50
Discussions of sustainability in archives have centered on participatory and democratic processes that integrate publics into the process of description [Jochumsen, Johnston, and Vårheim]. But there is little consideration of what these processes actually mean across different kinds of archival institutions and how technological know-how factors into these methods of opening up archives. Cait McKinney argues that “technologies evolve along with understandings of gender and sexuality” [McKinney 2020, 191]. Through the lens of queer/ed archives, practitioners might understand digitized collections as fluid archival bodies, grounded in particular physical historical contexts but also open to “ongoing worldmaking” that may not yet be visible to archivists [Lee 2017]. Within this framework, it might be easier to see how the shifting of materials and descriptions of these materials from one digital environment to another is part of the process of historical production. This movement constitutes a rhythm that defines archives. Sustainability of digitized collections depends on understanding their unmoored nature and role in history-making and building this indeterminateness into collections development and description from the beginning.

Acknowledgments

51
Thanks to Anna Sjödahl Hayman, Christa Shusko, and my initial two reviewers for reading and commenting on earlier drafts. Thanks as well to the peer reviewers and copyeditor at Digital Humanities Quarterly, who helped me pull the text together and provided additional citation advice. Thanks also to the KvinnSam staff. It is a joy to work with you all. Finally, thanks to Martin and Oona, my friends.

Notes

[1] See [Boyles 2018]. For a history of the underfinancing of “recovery projects” within literary studies, see [Earhart 2015, 62–89].
[2] The database is still currently available at http://www2.ub.gu.se/kvinn/bilddatabas/ (accessed 4 January 2024). It is due to be retired in 2025.
[3] KvinnSam's subject word list is available at http://www2.ub.gu.se/kvinn/kvinnsam/listor/amnesord.html (accessed 4 January 2024).
[4] Alvin is located at https://www.alvin-portal.org/alvin/home.jsf?dswid=1558 (accessed 4 January 2024).

Works Cited

Anderson, Eaton, and Schwartz 2015 Anderson, B., Eaton, F., and Schwartz, S.W. (2015) “Archival appraisal and the digital record: Applying past tradition for future practice”, New Review of Information Networking, 20(1-2). https://doi.org/10.1080/13614576.2015.1114823.
Arnold et al. 2017 Arnold, T. et al. (2017) “Uncovering latent metadata in the FSA-OWI Photographic Archive”, Digital Humanities Quarterly, 11(2). https://www.digitalhumanities.org/dhq/vol/11/2/000299/000299.html (Accessed: 4 January 2024).
Backlund and Sjödahl 2001 Backlund, B. and Sjödahl, A. (2001) Fotodatabas: En förstudie [Photography database: A introductory study]. Availablet at: http://www2.ub.gu.se/kvinn/ansvar/fotobas.pdf.
Bak 2016 Bak, G. (2016) “Not meta just data: Redefining content and metadata in archival theory and practice”, Journal of Archival Organization, 13(1-2), pp. 2-18.
Baker 2002 Baker, N. (2002) Double fold: Libraries and the assault on paper. New York: Vintage Books.
Björk 2015 Björk, L. (2015) How reproductive is a reproduction? Digital transmission of text-based documents. PhD thesis. The Swedish School of Library and Information Science. Availablet at: http://www.diva-portal.org/smash/get/diva2:860844/INSIDE01.pdf.
Boyles 2018 Boyles, C. (2018). “Counting the costs: Funding feminism in the digital humanities”, in Losh, E. and Wernimont, J. (eds.) Bodies of information: Intersectional feminism and digital humanities. Minneapolis, MN: University of Minnesota Press, pp. 93-197.
Caswell 2021 Caswell, M. (2021) “Dusting for fingerprints: Introducing feminist standpoint appraisal”, Journal of Critical Library and Information Studies, 3(2). https://doi.org/10.24242/jclis.v3i2.113.
Dahlström and Hansson 2019 Dahlström, M. and Hansson, J. (2019) “Documentary provenance and digitized collections: Concepts and problems”, Proceedings from the Document Academy, 6(1). https://doi.org/10.35492/docam/6/1/8.
Decker 2020 Decker, M.J. (2020) “The finger of God is here! Past, present, and future of digital history”, The Historian, 82(1), pp. 7–21. https://doi.org/10.1080/00182370.2020.1725720.
Drucker 2014 Drucker, J. (2014) Graphesis: Visual forms of knowledge production. Cambridge, MA: Harvard University Press.
Earhart 2015 Earhart, A.E. (2015) Traces of the old, uses of the new: The emergence of digital literary studies. Ann Arbor, MI: University of Michigan Press.
Earle 2017 Earle, K. (2017) “From runway to platform: The creation of a digital fashion photography archive”, Art Libraries Journal, 42(1), pp. 47-53.
Edmond and Morselli 2020 Edmond, J. and Morselli, F. (2020) “Sustainability of digital humanities projects as a publication and documentation challenge”, Journal of Documentation, 76(5), pp. 1019-1031. https://doi.org/10.1108/JD-12-2019-0232.
Ernst 2013 Ernst, W. (2013) Digital memory and the archive. Minneapolis, MN: University of Minnesota Press.
Freeman 2023 Freeman, E. (2023) “Defying description: Searching for queer history in institutional archives”, Archival Science, 23, pp. 447-470. http://doi.org/10.1007/s10502-023-09415-9.
Harris 2022 Harris, V. (2022) Ghosts of archive: Deconstructive intersectionality and praxis. New York: Routledge.
Jochumsen, Johnston, and Vårheim Jochumsen, H., Johnston, J., and Vårheim, A. (2023) “Pursuing sustainable futures through LAMs”, in Rasmussen, C.H., Rydbeck, K., and Larsen, H. (eds.) Libraries, archives, and museums in transition: Changes, challenges, and convergences in a Scandinavian persepctive. London: Routledge, pp. 215-228.
Johansson 2023 Johansson, M. (2023) Att upphandla och driva e-arkiv: Marknad, omvärld och behov. Stockholm, Sweden: Vulkan.
Katz 2019 Katz, S. (2019) “Excavating radical women in Progressive-era California”, in Chaudhuri, N., Katz, S., and Perry , M.E. (eds.) Contesting archives: Finding women in the sources. Urbana, IL: University of Illinois Press, pp. 89-106.
KvinnSam 2001a KvinnSam (2001a) Digitaliseringsseminariet [Digitization seminar] notes.
KvinnSam 2001b KvinnSam (2001b) Fotodatabasprojekt: Mötesanteckningar studiebesök på etnografiska museet, 010503 [Photo database project: Meeting notes from study visit to the ethnografic museum, March 5, 2001)..
KvinnSam 2001c KvinnSam (2001c) Anteckningar från möte med Anna, Berith, Inger ang. fotodatabasen, 010828 [Notes from meeting with Anna, Berith, Inger concerning photo database, August 28, 2001]..
KvinnSam 2001d KvinnSam (2001d) Kravspecifikation: Fotodatabas [Technical requirements: Photo database].
KvinnSam 2001e KvinnSam (2001e) Alternative för fotodatabasen [Alternatives for the photo database]..
KvinnSam 2001f KvinnSam (2001f) För-, och nackdelar med kvinnsam resp. egenproducerad databas för digit [Pros and cons with KvinnSam versus a bespoke database from digit's view]..
Lange 2020 Lange, A. (2020) Picturing political power: Images in the women's suffrage movement. Chicago, IL: The University of Chicago Press.
Lee 2017 Lee, J.A. (2017) “A queer/ed archival methodology: Archival bodies as nomadic subjects”, Critical Archival Studies, 1(2). https://doi.org/10.24242/jclis.v1i2.26.
Lesy 2020 Lesy, M. (2020) “Visual literacy”, The Journal of American History, 94(1), pp. 143-156.
Losman 1984 Losman, B. (1984) “Kvinnohistoriskt Arkiv och de kvinnohistoriska samlingarna i Göteborg (The women's history archives and the women's history collections in Gothenburg)”, in Losman, B. and Lönnroth, L. (eds.) Bibliotheca feminarum, Forum för kvinnliga forskare och kvinnoforskning (Forum for women researchers and women's research). Gothenburg, Sweden: Servicekontorets Grafiska, pp. 3–16.
Mak 2012 Mak, B. (2012) “On the uses of authenticity”, Archivaria, 73, pp. 1–17.
McKinney 2020 McKinney, C. (2020) Information activism: A queer history of lesbian media technologies. Durham, NC: Duke University Press.
Odumosu 2020 Odumosu, T. (2020) “The crying child: On colonial archives, digitization, and ethics of care in the colonial archives”, Current Anthropology, 61(22), pp. 289-302.
Owens and Padilla 2021 Owens, T. and Padilla, T. (2021) “Digital sources and digital archives: Historical evidence in the digital age”, International Journal of Digital Humanities, 1(4), pp. 325–341. https://doi.org/10.1007/s42803-020-00028-7.
Pierce 2021 Pierce, R. (2021) “Documents in the Dynarchive: Questioning the total revolution of the digital archive”, Proceedings from the Document Academy, 8(2). https://doi.org/10.35492/docam/8/2/4.
Pierce 2023 Pierce, R. (2023) “Representing women across the public/private divide: Kerstin Hesselgren's image in the Swedish digital archives”, Women's History Review, 33(1), pp. 95-116. https://doi.org/10.1080/09612025.2023.2208412 (Accessed: 4 January 2024).
Poole 2016 Poole, A.H. (2016) “The conceptual ecology of digital humanities”, Journal of Documentation, 73(1), pp. 91-122. https://doi.org/10.1108/JD-05-2016-0065.
Reger 2017 Reger, J. (2017) “Finding a place in history: The discursive legacy of the wave metaphor and contemporary feminism”, Feminist Studies, 43(1), pp. 193–221. https://doi-org.ezproxy.neu.edu/10.15767/feministstudies.43.1.0193.
Ries and Palkó 2019 Ries, T. and Palkó, G. (2019) “Born-digital archives”, International Journal of Digital Humanities, 1(1), pp. 1-11. http://doi.org/10.1007/s42803-019-00011-x.
Ruest, Fritz, and Milligan 2022 Ruest, N., Fritz, S., and Milligan, I. (2022) “Creating order from the mess: Web archive derivative datasets and notebooks”, Archives and Records, 43(3), pp. 316–331. https://doi.org/10.1080/23257962.2022.2100336.
SAA n.d. Society of American Archivists (n.d.) Describing archives: A content standard. Available at: https://saa-ts-dacs.github.io/dacs/06_part_I/02_chapter_01.html (Accessed: 4 January 2024).
Sassoon 2004 Sassoon, Joanna. (2004). “Photographic materiality in the age of digital representation”, in Edwards, E. (ed.) Photographs, objects, histories. London: Routledge.
Scheltjens 2023 Scheltjens, W. (2023) “Upcycling historical data collections: A paradigm for digital history?”, Journal of Documentation, 79(6). https://doi.org/10.1108/JD-12-2022-0271.
Schwartz 1995 Schwartz, J. (1995) “‘We make our tools and our tools make us’: Lessons from photographs for the practice, politics, and poetics of diplomatics”, Archivaria, 40(Fall 1995), pp. 40-74. Available at: https://archivaria.ca/index.php/archivaria/article/view/12096.
Smiley 2019 Smiley, B.L. (2019) “From humanities to scholarship”, in Gold, M.K. and Klein, L.F. (eds.) Debates in the digital humanities. Minneapolis, MN: University of Minnesota Press.
Thrift 2012 Thrift, S. (2012) “A feminist eventfulness: Boredom and the 1984 Canadian leadership debate on women's issues”, Feminist Media Studies, 12(3), pp. 406–421. https://doi.org/10.1080/14680777.2011.615634.
Thylstrup 2018 Thylstrup, N.B. (2018) The politics of mass digitization. Cambridge, MA: The MIT Press.
Trouillot 2015 Trouillot, M-R. (2015) Silencing the past: Power and the production of history. Boston, MA: Beacon Press.
Vestberg 2013 Vestberg, N.L. (2013) “The photographic image in digital archives”, in Lister, M. (ed.) The photographic image in digital culture. London: Routledge, pp. 113-130.
Withers 2015 Withers, D. (2015) Feminism, digital culture and the politics of transmission: Theory, practice and cultural heritage. London: Rowman & Littlefield.
Yeo 2015 Yeo, G. (2015) “Contexts, original orders, and item-level orientation: Responding creatively to users' needs and technological change”, Journal of Archival Organization, 12(3-4), pp. 170–185. https://doi.org/10.1080/15332748.2015.1048626.
Zhang 2012 Zhang, J. (2012) “Archival representation in the digital age”, Journal of Archival Organization, 10(1), pp. 45–68. https://doi.org/10.1080/15332748.2012.677671.
Preview  |  XML |  PDF |  Print
Creative Commons License
This work is licensed under a Creative Commons Attribution-NoDerivatives 4.0 International License.