Co-curation: Archival interventions and voluntary sector

There is a growing trend across the social sciences to engage with archives. Within human geography, this has stimulated a debate about the nature of archives, including moving from considering "archive as source" to "archive as subject." We build on and extend this thinking, suggesting that an even more active appreciation of the dynamic nature of relationships between researchers, owners of records, and archival material is needed. This paper draws on an interdisciplinary study of voluntary action and welfare provision in England in the 1940s and 2010s to highlight how the different iterative processes involved in collaborative archival research are part of what we call co-curation. Co-curation involves the negotiated identification, selection, preparation, and interpretation of archival materials. This has implications for both research processes and outcomes.


| INTRODUCTION
The archive has recently emerged as a subject of methodological interest in a range of disciplines beyond history or archival science, including human geography (Hyacinth, 2019;Mills, 2013a).This "archival turn" is partly indebted to a Foucauldian analysis of the archive as an artefact of knowledge production (Foucault, 1969).Rather than being seen simply as a system of files, the archive is defined as the practice that determines what is filed (Basu & De Jong, 2016, pp. 5-6).This involves a move from considering "archive as source" to "archive as subject" that examines "the practices of collecting, classifying, ordering, display and reuse" (Ashmore et al., 2012, p. 82; see also Stoler, 2002).While archival work has traditionally been perceived to be a solitary process, Ashmore et al. (2012, p. 81) have reflected on their experiences of working with the owners of archival collections as a "collaborative practice, communal knowledge formation" (2012, p. 82).In this paper we extend this thinking to suggest that an even more active appreciation of the dynamic nature of relationships is needed, particularly for private archives that are not mediated by professional archivists.We build too on the growing literature on "archival interventions" made by scholars, which has been variously conceptualised as participatory historical-geographical research, archival activism, or historian-activism (Bressey, 2014;DeLyser, 2014;Flinn, 2011;Mills, 2013b;Oppenheimer, 2020).In doing so, we develop the idea of co-curation: the identification, selection, preparation, and interpretation of archival materials as it is negotiated between researchers and owners of records.
This paper offers fresh insights into the iterative processes involved in collaborative archival research with voluntary sector partners.It draws on an interdisciplinary study of voluntary action and welfare provision in England in the 1940s and 2010s, to explore the process of co-curation between the research team and institutional owners of records.The archives and records of voluntary organisations are strategic assets with huge importance for research, but, like other private archives, should be considered "at risk" because they lack the long-term legal protection afforded to records produced by government.Many such collections retained in-house are subject to the vagaries of waxing and waning organisational interest, staff turnover, office relocation, and mergers (McMurray, 2014).
This paper is our first attempt to define and explore the concept of co-curation.Co-curation, we suggest, is an ongoing process through which the owners of private records work alongside researchers at every stage of a study.It enables access to previously little-used sources as well as generating insights not available in a more conventional research project.Co-curation has benefits for other scholars through interventions that improve the long-term preservation and accessibility of collections.In actively engaging staff in voluntary organisations in work with institutional archives, it also builds interest, skills, knowledge, and capacity, helping to ensure the research has lasting impact for practice.In what follows we set out our concept of co-curation as it applies to working with organisational partners in the voluntary sector.First, we briefly review existing literature on the use of voluntary sector archives in human geography, before outlining the research study and discussing how we engaged with such sources.Subsequently we discuss co-curation as a "process" and then as an "outcome."We touch on important, yet more mundane, aspects of the process that are rarely discussed in methodological literature (Ashmore et al., 2012).We conclude by reflecting on the wider implications of co-curation for human geography and beyond.

| USING VOLUNTARY SECTOR ARCHIVES IN HUMAN GEOGRAPHY
There is new-found recognition of the value of archives and records of and within the UK voluntary sector.High-profile inquiries into the history of public, corporate, and charitable bodies have highlighted the evidential value of records.In the humanitarian sector, leading aid agencies like CARE, Save the Children, and Oxfam have "begun to recognise that their archives are strategic assets for analysing the evolution of humanitarianism in a changing political landscape" (Götz et al., 2020, p. 308).There is growing understanding too that charity archives may preserve stories of marginalised individuals and communities whose lives are not recorded elsewhere.For example, significant contributions to historical geographies of black women in Britain have been enabled through collections such as the Barnardo's photographic archive (Bressey, 2002).However, charity archives have long been under the radar (Newton, 2004), under-resourced, and consequently at risk.Research has examined the vulnerability of in-house archive services (McMurray, 2014;Newton, 2004), records management in charities (Dawson et al., 2004), questions of cataloguing and user engagement (Mills, 2013b), third-party deposit of charity archives (Oppenheimer, 2020), and research uses (Brewis, 2020).
Ketelaar argues that archivalisation is "the conscious or unconscious choice (determined by social and cultural factors) to consider something worth archiving" (2001, p. 133).Like all collections, voluntary organisations' records have been affected by subjective decisions about what has been, or will be, preserved.For private collections there are additional questions over how and in what ways outside researchers might be able to utilise their holdings (Boyer, 2004;Hyacinth, 2019).Oppenheimer argues that we need to understand not just the agency of archivists and record keepers, but, for voluntary organisations, also "the process by which the organisation itself came to value its records in a particular way" (2020 p. 172).Oppenheimer's description of herself as "historian-activist" is relevant here, in that she had been "agitating" for the Australian Red Cross to deposit its archive into a public repository since her first encounter with the records in the 1980s, a process that in fact took over 25 years to achieve.Boyer (2004, p. 170) reflected on working with historical sources through the lens of feminist geography, which included an awareness of how power places and structures identity, lived experience, and social relations in spaces of the past.She emphasised the importance of finding sources at the boundary of public knowledge, such as the non-public archives of professional organisations and charities, to expand one's base of sources.But Boyer also highlighted the challenges that can ensue from accessing and using private archives, including how organisations can choose to "filter" who can access collections and may "preserve documents selectively" in order to present themselves in a more favourable light (2004, p. 172).
In her work on historical geographies of abortion, Moore (2010, p. 265) discussed the absence of archival materials and silences she faced.Moore also highlighted the potential effects of making personal and delicate information public, and the possible conflicts of interest between researcher and participant, even when the participant is dead (2010, p. 268).Dwyer and Davies (2010) considered the contradictory processes of archiving, of giving form to the identities and capacities of past communities, spaces, and landscapes, while simultaneously erasing or eliding that which cannot easily be captured (2010, p. 260).This has resonance with sustained efforts across the humanities and social sciences to increase the representativeness of archives (Johnston, 2001).
Building on such scholarly insights about the potential significance of private collections for research and those which foreground participatory and inclusive approaches to the archive, this paper argues that the process of co-curating research between records owners in the voluntary sector and academics is a form of archival intervention with the potential to become a mainstream research methodology in historical and human geography.

| THE DISCOURSES OF VOLUNTARY ACTION STUDY
In this paper, we develop the idea of co-curation by drawing on our experience of undertaking an interdisciplinary, collaborative research project exploring voluntary action and welfare provision in the 1940s and 2010s.These two decades can be considered as "transformational moments" in which the boundaries between state, voluntary action, and others were rethought (Brewis et al., 2021).The project analysed narratives about the role, position, and contribution of voluntary organisations that emanated from the voluntary movement, the public, and the state.It focused on four fields of voluntary action: children, youth, older people, and the voluntary movement/sector as a whole.
For each field of activity we identified a key voluntary sector infrastructure or umbrella body that was active in the 1940s and which continued into the 2010s: Age UK, Children England, the National Council for Voluntary Organisations (NCVO), and the National Council of Voluntary Youth Services (NCVYS).Our intention to work with one partner in each field was disrupted by closures and mergers within the voluntary youth sector towards the end of the 2010s, leading us to work with two additional organisations -Ambition and UK Youth (Table 1).We also worked in partnership with the Mass Observation Archive (MOA), 1 with Mass Observation data used to explore public narratives, while government policy documents, speeches, and parliamentary debates were used as the sources for exploring state narratives (see Brewis et al., 2021).It is, however, the process of working with our partner voluntary organisations that forms the focus of this paper.

Voluntary movement
The voluntary movement/sector was explored through the papers of the National Council for Voluntary Organisations (NCVO).
Founded as the National Council for Social Service (NCSS) in 1919, NCVO's archive is deposited at the London Metropolitan Archives.We worked with colleagues at NCVO to identify and select 2010s material.

Children
Children England is the 'children's specialist' membership body for voluntary organisations in England.The Associated Council of Children's Homes was established in 1941 by four of the largest charities then providing residential care for children in the UK, with others soon joining.It became Children England in 2009.We acquired the organisation's surviving archival material, dating back to the 1940s, on temporary deposit at UCL Special Collections, and worked with the staff team to select 2010s source material.

Youth
We used the archive of the National Council for Voluntary Youth Services (NCVYS), established in 1936 by 11 national voluntary youth organisations, and which closed in 2016 just as the research was beginning.After closure, the collection was donated to UCL Special Collections in association with this project.Subsequently, we accessed records of Ambition, which was founded in 1925 as the National Association of Boys' Clubs (NABC).These papers are today privately held by UK Youth, following its 2018 merger with Ambition.UK Youth began life in 1911 as the National Association of Girls' Clubs; its archive is at the University of Birmingham.

Older people
The National Old People's Welfare Committee was established in 1940 as part of NCSS.It gained independence in 1970 and became Age Concern.Age UK was created in 2009 following the merger of Age Concern and Help the Aged.We acquired surviving material, dating back to the 1940s, which was taken on temporary deposit at UCL Special Collections for the duration of the research.We worked with colleagues at Age UK to identify and select records relating to the 2010s.
Apart from NCVO and UK Youth, the organisations we partnered did not have archival records in the public domain.Even for these two organisations, their archives did not include the documents we hoped to include from the 2010s.This epitomises the challenges facing research into subjects such as the roots of the mixed economy of welfare, which is often hampered by a lack of access to such sources.Our project addressed a major gap in knowledge by accessing data from these private collections.Indeed, the viability of the research proposal depended on being able to access and interpret the "archival voice" of these organisations.Conversations were held with the four original partners during the preparation of a funding proposal, building on established relationships between members of the research team and those organisations.Subsequent discussions were held with Ambition and UK Youth after the closure of NCVYS, and the subsequent merger of Ambition into UK Youth.

| CO -CURATION AS PROCESS
The co-curation of the archives involved the negotiated identification, selection, preparation, and interpretation of materials.At each stage decisions were made that shaped and re-shaped the form, content, and understanding of the archive.Each stage raised questions about what records to include and exclude as well as highlighting the varying capacities of our partners to engage with the project.

| Governance
Co-curation depends on forging successful collaborative partnerships.Collaboration is often described as being between organisations, but it is enacted by people in organisations (Hardill & Mills, 2013).In securing access, we drew on a mix of past research connections and team members' long track record working with voluntary organisations, including earlier archival interventions (see, for example, Brewis' British Academy-funded "Archiving the Mixed Economy of Welfare" project and AH/W002353/1 AHRC-Collaborative Doctoral Partnership "Charity and voluntary sector archives at risk: Conceptualising and contextualising a neglected archives sector").
We established a steering group with members drawn from our partners, academics with relevant expertise, and a professional archivist from UCL.It met five times over two years, and provided suggestions and feedback on data gathering, joined in on the analysis process, discussed emerging findings, and co-designed dissemination activities.
The formal arrangement of each partnership was specified in detailed Memoranda of Understanding (MoU), which covered access to source material, ethics, copyright, intellectual property, outputs, knowledge exchange, and depositing of data.For example, we agreed that each partner would have the chance to review every publication in which the organisation was mentioned and secured permission to reproduce copyrighted material.The MoU were signed off by senior staff at each organisation and by legal services at the lead university (for further details, see Brewis, 2020).

| Identification
Identification of source materials was the next step.We were able to access some material in third-party repositories: NCVO had deposited its archive  at the London Metropolitan Archives (LMA) in the 1990s, while the transfer of the NCVYS archive  to UCL was facilitated by the research team during the planning stages of the project.UK Youth had also deposited its archive (1909-c.2015) at the University of Birmingham.For everything else we had to rely on private collections, with the situation reflecting the differing levels of priority accorded to records across different organisations over time.The process of tracking materials down was not always simple and often relied on the knowledge of key long-serving staff members.Finding the "right" staff member(s) was a crucial step.The Children England papers from the 1940s, consisting of two boxes of board minutes and circulars, had been kept safe by a staff member and were transferred to UCL as a temporary deposit.Available records from Age UK for the National Old People's Welfare Committee in the 1940s were fragmentary; the few boxes that were tracked down were transferred to UCL on temporary deposit but lacked board minutes or printed reports.In order to fill some of the gaps, additional material was located by the research team at the British Library and this was supplemented with purchase of other material from eBay (see DeLyser et al., 2004).Access to the papers of Ambition was complicated by the merger with UK Youth part-way through our negotiations.Eventually, the research team visited the organisation to select material to transfer to UCL for temporary deposit.While the Ambition archive was the most comprehensive collection of 1940s material out of those we acquired on loan, some of the material was in poor condition after being stored for decades in a damp basement and posed a contamination threat to other collections.
For materials from the 2010s, we faced the issue of the scale of records, alongside the need to provide reassurance to partners regarding the research team's access to sensitive material.The 2010s material was far more extensive in volume, taking time to locate and organise.We were heavily reliant on staff members to help us identify and retrieve relevant documents.Apart from the NCVYS archive at UCL (which was complete up to 2016), this material did not form part of archive collections but were internal working documents, often a mix of hard copies of a few key documents with the rest available as soft copies, filed in multiple places across internal virtual storage systems.Rarely were they subject to formal records management procedures.
Access to contemporary sets of board papers represented the greatest concern for our partners, particularly as the timing of the project coincided with a period of sensitive restructuring and merger discussions, which were played out in the board minutes of several organisations that were involved in the study.The introduction in 2016 of General Data Protection Regulation (GDPR) 2 had shone a spotlight on sharing personal data, which further affected attitudes towards access.One organisation became particularly anxious about sharing documents in which living individuals were named after concerns regarding a potential breach of GDPR elsewhere within the organisation.Without the trust that had been built between the research team and the organisations, and the additional reassurance provided by the MoU, access to these more sensitive documents would not have been possible.Indeed, we did not get full sets of minutes for all organisations, with concerns about access combining with a lack of capacity within the organisation proving insurmountable within the time available for the research.
These private archives lacked the order and structure taken for granted when using archives deposited in a thirdparty repository.None had a catalogue or box list.This created a challenge, both at the stage of identification and also later when it came to referencing materials without the familiar fall-back of box, file, and item codes.If, for 1940s materials, there was a concern about lack of documents, for the newer materials it was one of having too many.Identification and selection was a process of negotiation between the research team and partner organisations.Decisions were made by both sides that affected what was included.There were also examples of missing documents, which only became apparent when their existence was indicated through other sources.Within the process of co-curation, materials may be forgotten, not thought relevant, not possible to locate, or purposefully retained.Our experiences reflect the broader challenge of archival research on voluntary action, which entails accessing what is often considered "dispensable ephemera" via private archives or tracking down scattered records to reconstruct an organisational archive (Brewis, 2014, p. 10).

| Preparation
After identifying and acquiring the materials, the next task was to prepare the documents for use within the project.We had assembled an enormous amount of material, much of which did not address our research questions directly, although it was of wider value providing background and context.A considerable amount of time was spent collecting, collating, cataloguing, reviewing, prioritising, and preparing documents for analysis.The 1940s material was for the most part administrative material, including minutes, annual reports, newsletters, and, in some cases, correspondence and publications.These sources began life as typescript, printed, or handwritten documents.After professional scanning, selected documents were converted into readable PDF or Word files using Optical Character Recognition software, supplemented by manual data "cleaning."The preparation of such sources was both labour and resource intensive, and could not have been undertaken without funding.The 2010s material was either "born digital," produced as Word or PDF files, or in some cases scanned from print copies to create readable PDF files.A key issue here was the scale of the data, running to thousands of documents: without any pre-existing catalogue, all had to be read in order to select the most relevant, inevitably making choices and compromises.
The original intention was to use computer-assisted qualitative data analysis software, more specifically NVivo10, to facilitating cross-team analysis.After an initial reading, those documents or sections of documents judged to be the most relevant were prepared and imported into NVivo.The range of documents and formats that can be analysed within NVivo has expanded in recent years, including enabling the inclusion of both text-based and image-only PDFs.In order to code and query the documents, however, the PDFs needed to be text-based, meaning that even some of the materials from the 2010s had to be converted.In practice, timescales and differing levels of familiarity with the software across the team meant that we did not use it as consistently as we had initially intended, but even our limited use demonstrated its potential utility, particularly given the scale of the data and the ability to share analysis across the team.

| Interpretation
The practice of co-curation included being in regular discussion with key staff at partner organisations to help contextualise and interpret the source material.This happened throughout the study, beginning with extensive initial conversations with potential partner organisations before the proposal was drafted.These conversations continued throughout the research, and -along with our existing knowledge and reading of wider literature and theoretical framings -helped to shape our interpretation of the materials, drawing our attention to certain documents or particular lines of argument, for example.
Steering group meetings were used more directly to inform our interpretation.At two of the meetings, for example, we shared selected documents with members who then worked collectively to identify key emerging themes.This fed directly into the development of a coding frame.
Towards the end of our analysis, we organised workshops with each partner organisation.These events presented emergent findings to groups of staff, trustees, or member bodies, and helped us test out and extend our interpretation of the data from individual organisational perspectives.They pointed us towards some important new avenues for our analysis, in some cases highlighting developments that our own analysis of the documents had not.In addition, we ran several wider events in which our partners joined us in sharing the emerging findings and reflecting on their relevance with other voluntary sector organisations and academics: the ensuing discussions helped refine our analysis, while also raising awareness of the value of charity archives.

| CO -CURATION AS OUTCOME
Co-curation has benefits, we suggest, for all concerned.The quality of our data collection and analysis was enhanced by the co-curation process, with implications for the publication of the project research findings (Brewis et al., 2021).As was noted earlier, research into the roots of the mixed economy of welfare has been hampered by a lack of access to privately held archives of voluntary organisations.The approach that we adopted enabled access, helped with identification, strengthened our interpretation/analysis, and refined our outputs.The evidence that has been generated is of a higher quality, and more robust, as a result.It was also a personally rewarding experience for the researchers involved.
For our project partners, one important outcome from co-curating their archives was the rediscovery of previously little-known or lost documents, images, or objects, which offers potential for new interpretations of the earlier work of an organisation.One example was the identification for Age UK of the forget-me-knot pin badge produced in the 1940s for members of the local National Old People's Welfare Committee lunch clubs.We shared with the organisations the digitised versions of documents we produced, making them more readily accessible and useable.The uncovering of written sources and visual images was welcomed by our partners.However, while rediscovery is exciting, it can also be disruptive of an organisation's own interpretation of its history, which may draw heavily on a foundation narrative (Hilton et al., 2013) or have been reworked to shape current agendas.Sensitivity to this is needed.
A third outcome relates to long-term preservation of the organisational archives we worked with.The research team is continuing to work with our partners to secure a sustainable future for all these collections.The acquisition of the NCVYS archive by UCL was a serendipitous outcome that coincided with discussions about involvement in the research.At the time of writing, plans are underway to retain the Children England archive at UCL Institute of Education, and negotiations are ongoing about Age UK's and Ambition's archives.This will ensure that these important records are preserved and available for others to research in the future.
While our experience of co-curation was positive, it was not without its challenges.The timescales of academic research can often feel at odds with those of voluntary organisations.While our partners were deeply supportive of the research, it was rarely a central priority.At times this meant that we were asking more of the organisations than they had the capacity for.It likely also meant that partners were frustrated by the relatively long timescale of the study, particularly when this appeared at odds with our occasional requests to turn things around quickly.Taking the time to nurture the relationships and build trust was key.There is a risk that getting too close to research partners makes it difficult to keep a critical distance.We suggest that the ways in which our project brought voluntary sector archives into conversation with each other, and with state narratives and public narratives, enabled triangulation and allowed more critical and challenging questions to be debated.Finally, while co-curation can help secure the future of valuable archives, it will likely shape the collections in ways that reflect not just the agendas of the archive holders themselves but also those of the researchers.This is likely to be particularly marked for our project, through its focus on the 1940s and 2010s, which may have skewed the archival records towards these two time periods.

| CONCLUSIONS
In this paper we have provided an account of co-curation through a focus on our engagement with the private archival collections of voluntary organisations.Securing this access enabled us to address a lacuna in research into the roots of the mixed economy of welfare.Methodologically we have moved beyond conceiving of an archive as a system of files, to thinking of an archive as a practice, to a third stage of co-curation.Co-curation might be seen as part of a broader iterative approach, one that is not neatly staged, is actively negotiated and shaped, and that involves choices being made -by all concerned and at all stages in the process -about what records are included or excluded.We have focused on the different iterative processes involved in collaborative archival research, which we argue leads to the production of co-curated collections.
Co-curating private archives demands the allocation of time and resources by the owners of records.Co-curation includes the identification of questions, partners, and materials and is built on trust and sustained through negotiation.Staff can lack the time to search for records and are unlikely to have professional archiving or records management skills.Co-curation also involves academics actively intervening in discussions about archive records.It can offer opportunities to researchers for reciprocity, to give back to organisations, the owners of private archives, and for the (re)discovery of the past by organisations.We were fortunate to have both the support from the organisations and the funding from a research grant to support our endeavours.
Co-curation has implications for human geography, and for other allied disciplines.Importantly, co-curation could open sources at the boundary of public knowledge, such as the private archives of voluntary organisations (Boyer, 2004), which to date have remained an under-utilised resource.It may enhance research quality through promoting methodological innovation that will lead to new, substantive insights.It may also enhance research engagement and impact through the relationships that are at its core.While co-curation should not be approached uncritically, it should, we suggest, be seen as a useful addition to the human geographer's methodological toolbox.