Rebalancing the Historical Narrative or Perpetuating Bias? Digitizing the Archives of the Mineworkers’ Union of Zambia

Abstract This article examines the project to digitize and preserve the archives of the Mineworkers’ Union of Zambia and has two aims. The first aim is to discuss the process of cataloguing and digitizing an archive that has undergone significant deterioration, and the theoretical and practical challenges to achieving this. The second aim is to relate making this archive more accessible to questions of knowledge production. Despite its limitations, the value of this archive is that it is primarily composed of documents produced by Africans about the world as they saw it. These are not the records of external powers, colonial officials, or those studying African peoples.

Introduction 1 It will become increasingly common for Africanist historians to work with digitized archival material, and this makes understanding the process of the creation of such archives, their selection, and the implications for making particular archives more accessible critical. This article describes the efforts to preserve and digitize the Mineworkers' Union of Zambia (MUZ) archive, a project initiated by the author in May 2018, and assesses the lessons this project may have for comparable digitization projects elsewhere on the continent and the implications for knowledge production. This project has already been mentioned briefly in this journal by Hyden Munene in his survey of Zambian archives, but was then at a nascent stage. 2 At the time of writing in mid-2020, this collection had been digitized but is not yet available to researchers, as the final stages of the project have been prevented by the constraints imposed by the coronavirus pandemic.
MUZ still exists as a trade union that organizes workers in Zambia's mining industry and has a long history, tracing its origins back to the trade unions formed by African mineworkers in the late 1940s. The union was studied extensively during the 1950s, 1960s, and 1970s, and, more recently, was the subject of an important work by Miles Larmer concentrating on the period between the country's independence in 1964 and 2005. 3 The union has faced serious difficulties since the early 1990s, however. The prolonged recession in the mining industry and subsequent privatization of the mines created enormous financial problems for the union. This is the salient background for the deterioration and disorganization of the union's historical material, something which is paralleled with many other archival collections in the region. As will be discussed more fully below, when the project began there was no neatly delineated archive as documents were scattered across the union's head office and stored in different ways with varying 1 I am grateful to the Mineworkers' Union of Zambia for allowing unrestricted access to their archive and to the International Institute of Social History for funding the project. Versions of this paper were presented at the Southern Africa History Society conference at Rhodes University in July 2019 and the Ramblers seminar at the University of the Free State in October 2019. I would like to thank the audience on both occasions for their feedback and engagement with the paper. I am also grateful to Andreas Admasie and the journal's anonymous reviewers for their comments.
2 Hyden Munene, "Mining the Past: A Report of Four Archival Repositories in Zambia," History in Africa 47 (2019), 359-373. 3 Miles Larmer, Mineworkers in Zambia: Labour and Political Change in Zambia, 1964-1991(London: I.B. Tauris, 2007. degrees of damage. It was the prospect that the union's historical documents would be irretrievably damaged that motivated this project. This article has two aims. The first is to discuss the practicalities of the digitization and preservation of archives as the experience with MUZ has parallels and perhaps useful lessons for other projects, including persistent power cuts preventing progress with digitization. This includes a discussion on how the archive was constituted due to the lack of certainty around the size and parameters of the collection, and the tensions between access, ethics, and ownership. The inclusion of very recent documents in the archive, interspersed with older material, containing medical and financial information that should be preserved, but must not be made available yet, meant that it was not possible to make the MUZ archive freely accessible online. Instead, two copies of the digital material have been created. One will be hosted at the International Institute of Social History (IISH) in Amsterdam, Netherlandswhich has funded and supported the projectand the other will remain with MUZ in Kitwe, Zambia, where the physical archive will remain.
The second aim is to relate this collection and the implications of making it accessible to researchers to questions of knowledge production in Central Africa. Although at a global level, digitized collections on African labor and social history are few, at a local level, the amount of available information about male mineworkers on Zambia's Copperbelt is considerable, and their lives and experiences have been at the center of Zambia's historiography. It is possible that, by making these sources more accessible, this project will reinforce the bias within Zambian historiography towards male mineworkers. This is not only a question for this project. Existing biases in the historiography are often reflections of bias in the availability of historical source material. Digitization projects may therefore inadvertently perpetuate that bias by making some sources more accessible at the expenses of others.

Context and Provenance of the Archive
MUZ remains an active trade union and organizes and represents all grades of workers in Zambia's mining industry to engage in collective bargaining with employers, who include some the world's largest multinational mining companies. At one time, MUZ was a powerful and influential union, but exists today in a diminished form and the decline of the organization mirrors the deterioration and decay of its historical material.
In its current form, MUZ was formed in 1967 in a government-backed merger of the three trade unions in the mining industry which represented African employees, though its roots go back further than this into the colonial period. Trade unions were formed by African mineworkers in 1948 with the assistance of a British trade unionist sent by the British Government to foster "moderate" unions that would engage in collective bargaining without resorting to strike action or becoming involved in politics. However, the newly established union quickly exceeded these bounds and initiated a series of tumultuous disputes during the 1950s and 1960s over wages, working conditions and racial discrimination, and African mineworkers won significant wage increases.
Strikes and unrest on the mines continued after Zambian independence as MUZ had, at best, an uneasy relationship with the government and many mineworkers clashed with government policy, especially in the early years of independence. In 1972, Zambia became a one-party state and the labor movement on the Copperbelt became a center of resistance to the ruling United National Independence Party (UNIP) in the face of persistent efforts by UNIP to bring MUZ and Zambia's mineworkers under tighter state control. Strikes, protests, and riots on the Copperbelt during the 1980s undermined the oneparty state, and the labor movement played a crucial role in the movement to restore democracy, which led to the elections in 1991 and the defeat of UNIP. 4 Resumption of multi-party democracy heralded serious difficulties for MUZ as Zambia's economic problems worsened. The mining industry was fully nationalized in 1973 and, unhappily for Zambia, the period of state ownership coincided with a protracted slump in the world copper industry. Rising debt levels and economic recession eventually resulted in the wholesale privatization of the mining industry in the late 1990s as financial assistance from the International Monetary Fund and the World Bank was dependent upon economic reforms that included privatization. 5 This lengthy economic recession and the form that privatization took is the salient background for understanding the present state of the MUZ archive.
"Mass unemployment" was "perhaps the greatest impact of privatization" in the assessment of Alistair Fraser and John Lungu. Mining employment fell from 56,582 in 1991 to 19,145 by the time privatization was completed in 2001. 6 When the copper industry revived with the global commodities boom in the 2000s, the labor structure of the privatized mines was quite different. Comparatively, few workers are today directly employed on permanent contracts and instead mining operations rely heavily on contractors and sub-contractors, who have proved much more difficult to unionize. 7 The sharp contraction in the mining industry hit MUZ hard. boom in copper prices, membership continued has continued to decline and had fallen below 17,000 by 2016. Consequently, the union's financial resources, generated both by membership dues and selling services to members, were severely constrained. 8 In the late 2000s, MUZ almost went bankrupt. To make matters worse, the organizational capacity of MUZ was sapped by splits in the organization in 2003 and 2010 that led to the emergence of two rival trade unions: The National Union of Miners and Allied Workers and the United Mineworkers' Union of Zambia.
All of this has left the organization ill-equipped and ill-resourced to preserve its own historical record. During his doctoral research in the early 2000s, Miles Larmer found that MUZ's head office has "no organized archive of its record," and some records "were accessed from a filing cabinet only with the aid of a crowbar." 9 Ching Kwan Lee found that even basic documents like recent collective bargaining agreements could not be located in the union's archive in the 2010s due the lack of an inventory or organization. 10 Further deterioration raised the possibility that this historical material would be irreparably damaged and become unusable. This was not an unrealistic prospect as, regrettably, in early 2019, while the digitization project was underway, heavy rainfall damaged the roof of the union's offices and water leaked into the main hall where documents awaiting digitization were being stored. 62 folders containing important documents were destroyed as a result, and others partially damaged.
The archive is located at MUZ's head offices in Katilungu House, Obote Avenue, in central Kitwe, the largest urban center on the Zambian Copperbelt. Organized record keeping began in 1973 when MUZ established a Research Section and what is now constituted as the union's archive can be traced back to then as earlier documents from the colonial period were largely destroyed in a fire in the 1970s. 11 The earliest documents date back to 1967, the year MUZ was formed, and the most recent are from 2016. The bulk of the archival material is from the 1990s and early 2000s.
Individual branches of MUZ also kept their own archives but recordkeeping at a branch level has been uneven, and sometimes disrupted by industrial upheaval. Mufulira's branch records were destroyed in 1985 after striking mineworkers burnt down the local union office. Luanshya branch officials assiduously kept records from the 1950s and some of these were after retrenched miners set the union office alight. 12 These branch records became intermingled with records kept at the head office and other branch records were subsequently transferred to the head office. It seems likely that the union has never employed a trained archivist, but there was initially a structure to the collection. Documents produced in the 1970s and 1980s were organized according to which union department produced them, and by year, with a clear referencing system and inventory, which is explained in greater detail in the section below. This organization process continued until the early 1990s when the mining industry, and Zambia more generally, suffered a period of acute economic dislocation and the union's membership slumped. After this, organized filing ceased entirely, and documents were added to folders regardless of subject, year, or type of document, then dispersed throughout the building. In this sense, the fortunes of the union were inscribed upon its archive, making it possible to get a sense of the union's struggles to survive and adapt to rapidly changing circumstances from the changing form of the archive.
This kind of rapid decline in the face of economic difficulties and political upheaval is not unusual. In a report for this journal in 1991, Mwelwa Musambachime explained that the archives of UNIP were open to researchers at the party's headquarters in central Lusaka and were overseen by a Research Bureau staffed by respected scholars, with clear policies regarding access. 13 By the late 2000s, with UNIP long out of power, the archives were stored in a warehouse in Lusaka's light industrial area and the records were decaying rapidly. The archive was preserved only by a timely British Library project run by Giacomo Macola and Marja Hinfelaar, which digitized the documents and revised the catalogue. 14 Sometime in the mid-2010s, the physical archive was removed from the warehouse and its precise location and the status of the documents is currently unclear.

Preserving and Digitizing an Archive: Practical Considerations
The genesis of this project came from enquires by MUZ officials about the possibility of digitizing their union's archive after my then PhD student Hyden Munene carried out some of his doctoral research in this archive and realized that the material had both important historical value and was in danger of deteriorating. I had been digitizing material on population and labor statistics in Zambia and thought, not appreciating the extent of the union's archive, that a similar small-scale digitization project could be replicated at MUZ.
When I realized that this was not possible, I raised the idea of a digitization and preservation project with staff at the IISH, who were enthusiastic about such a project. Although less prominent than the British Library's Endangered Archives Programme, the IISH has in recent years carried out several digitization projects to preserve collections on African labor and social history in collaboration with African trade unions and universities, including the archive of the International Trade Union Confederation-Africa in Togo. 15 The Institute houses one of the largest repositories of labor history in the world and is a suitable host for the MUZ digital archive as it already has material on MUZ among the papers of the Miner's International Federation, which MUZ was affiliated to.
In November 2017, meetings were held between two staff members from the IISH, me, and the MUZ leadership including the then deputy general secretary, and now general secretary, George Mumba. From the outset, MUZ officials were very supportive about the project and eager for it to go forward, and this has been critical to its success. This support is due to the fact that the union is keenly aware of its own history. Photographs of former officials adorn the walls of the MUZ offices and the organization makes efforts to educate its members about the history of the union through training courses and social media. 16 One staff member generously vacated his office to provide space for the cataloguing and digitization to take place, and it was essential to have this space for the work to proceed, while other union staff assisted with administrative work, such as drawing up employment contracts for the research assistants and keeping financial records of salary and tax payments. The union's research officer, Charles Muchimba, who has been working for the union since the mid-1980s, proved to be an invaluable source of information about the archive and the recent history of MUZ. Indeed, many documents in the archive were reports and correspondence authored by him.
Three research assistants were recruited for the project on the advice of Raymond Sikanyika, the electronic resources librarian at the Copperbelt University in Kitwe, through his involvement with the Library and Information Association of Zambia. These research assistants -Enala Kufakula, Shalom Zulu, and Tapiwa Zuluwere all recent graduates from the BA 15 ITUC-Africa, "Memorandum of Understanding IISH and ITUC-Africa," 30 June 2017, https://www.ituc-africa.org/Memorandum-of-Understanding-IISHand-ITUC-Africa.html (accessed 20 April 2020). 16 For instance, see the post on the official MUZ Facebook page "Who Was Lawrence Katilungu?", 8 January 2020, https://www.facebook.com/permalink.php? story_fbid=1044657902593682&id=538081536584657 (accessed 10 April 2020). Lawrence Katilungu was the first general secretary of MUZ's predecessor union, and the union's head office is named after him. Library Science and Information Management at the University of Zambia, and it should be made clear that it is these three persons who have carried out the actual work of digitization in terms of scanning the documents. These research assistants were employed by the union and their salaries were paid from the project funding, which was transferred directly to the union's bank account. In this way, the research assistants could be formally employed and receive social security and pension contributions, as well as a salary. The main problem with this approach was that international bank transfers from the Netherlands to Zambia were consistently time-consuming and the bank charges were exorbitant.
Following initial meetings with the MUZ leadership, the IISH agreed to fund and support the project. The Institute provided equipment in the form of a Futjitsu ScanSnap SV600 desktop scanner, Lenovo Thinkpad laptop, and hard drives. Lower-cost equipment such as breathing masks, gloves, dust jackets, which could be sourced in Kitwe, was also paid for by the IISH. This protective equipment was necessary due to the quantities of dust and mold that had accumulated on some files. Training in using the scanner and associated software was provided and certified by the IISH and carried out by a member of the Institute's Addis Ababa office in October 2018. The scanner and laptop were donated to MUZ once scanning was complete and a staff member completed training in how to operate the scanner, so that MUZ can continue to make electronic copies of documents if they wish. Funding paid for the equipment and training, but most of the funding was absorbed by salary costs for the three research assistants.
The project had two stages. The first stage involved the identification, collection, and cataloguing of historical material, while the second stage involved the digitization and preservation of these documents. Identification and collection in the first stage involved significant difficulties as what I refer to as the Mineworkers' Union archive, and the material that has been digitized, was not previously stored in one place and did not form a single clearly delineated body of documents. It is difficult to convey how dispersed and disorganized these documents were. Most were stored in the union's main hall or on a balcony overlooking the hall, but were also located throughout the building in storerooms, the offices of union staff, a disused toilet and even above the lift mechanism. Many documents were in cardboard folders or leverarch files, while others were bound up with string or stored in large canvass bags. This disorganization created practical and theoretical problems.
We tried to collect all material not currently in use by union staff over a three-week period in June 2018, following guidance from the IISH that archival records were "Non-current records preserved, with or without selection, by those responsible for their creation or by their successors." 17 However, it was not possible to make a systematic search of the building because the organization is still a functioning trade union and many offices were in continual use. MUZ officials were generous in their assistance for the project, but the searching and relocation of archival material had to be balanced against avoiding overly disrupting the day-to-day responsibilities of their jobs. Indeed, more generally, working with the union meant accepting that the project was necessarily low on the organization's priorities, as union officials and staff often had other pressing commitments on their time. For instance, in May 2019, when I was in Kitwe trying to arrange a new Memorandum of Understanding to extend the project, a dispute between the Zambian Government and Konkola Copper Mines threatened to shut the mine, and thousands of union members faced redundancy. Union officials were therefore justifiably preoccupied with this.
I was very conscious of my own role in shaping the MUZ archive. Although I am grateful for the assistance and guidance provided by archivists at the IISH, I should stress that I am not a trained archivist, I am a historian whose familiarity with archives comes from my experience as a user of archives, albeit primarily those archives about Zambia's mining industry. In this project, I feel uncomfortably close to what Fabienne Chamelot, Vincent Hiribarren, and Marie Rodet termed "a form of disciplinary arrogance," whereby historians ignore or diminish the work of archivists. 18 Disorganization of the material meant difficult decisions about what could be included in the archive. Archivists at the IISH provided guidelines and advice on creating an inventory and arranging an archive which began with the description of an archive as "an organically grown entity, which should be kept together, and may not be split up." 19 In practice, I found this impossible to adhere to as alongside folders we found large quantities of loose sheets of paper. Many of these sheets had clearly previously been organized into folders but putting this back together would be a mammoth task and would require physical space not available in the union's offices. Difficult decisions needed to be made about what it was feasible to catalogue and digitize and, regrettably, this meant excluding loose material, even though it doubtless contains much useful and relevant material for historians. Loose material was stacked together and left, so is effectively split from the rest of the archive. I was unable to think of a cost-effective solution to this problem.
Cataloguing began once material had been collected from around the building. The research assistants and myself read through every folder and bundle of documents together to ascertain the subjects of the enclosed documents as well as the dates of the earliest and most recent documents in each folder. This process also revealed that some files were too badly damaged to digitize, as poor storage coupled with Zambia's climatic cycles of dampness followed by dry heat rendered some documents illegible. This information was used to create a draft catalogue in an Excel spreadsheet. Following the IISH's basic principles on the arrangement of archives, each entry in the spreadsheet contains a file name, sequential reference number, brief description of the contents, years of the earliest and most recent documents in the file, physical description of the file (e.g. blue cardboard folder), and the original file reference if present. 20 During this process, the spreadsheet was sent to archivists at the IISH as well other researchers for comments on how it could be revised and improved. 21 A few days into the cataloguing process, we found the original inventory showing the organizing principle for the collection. In this system, folders were assigned a reference according to whichever department or branch of the union created them. MUZ/10, for instance, was the Safety & Health Department and was further sub-divided into categories, so MUZ/10/H were inquest files following accidents, and sometimes further sub-divided by union branch, so MUZ/10/H/2 designated inquest files from the Mufulira branch. I thought it might be possible to recreate the catalogue using this organizing principle, but less than 15 percent of the folders catalogued had a reference from the original inventory and this inventory was based on the union structures that existed in the 1980s and have subsequently changed. Moreover, most folders from the 2000s and 2010s contain a mixture of material, often on very different subjects, which makes it hard to assign them to a particular department or topic. Instead, all files were numbered sequentially more or less in the order that we found them and subsequently digitized in that order. The priority was quick cataloguing of the material at the expense of creating an organized archival catalogue.
The second stage of the project was the digitization itself. This began in October 2018 after the training was completed and continued until June 2020. As noted above, the work of carefully separating out each page from the files, scanning the pages, and then reconstructing the file in the same order was carried out by the three research assistants, although they did not all work at the same time. Since there was only one scanner for the project, only one person could scan documents so the research assistants either worked on their own or in twos, with one scanning documents and the other preparing documents for scanning. The use of a second scanner would likely have sped up the digitization process, but there was not sufficient space for one in the office where digitization took place.
Digitization was hampered by other practical problems alongside the lack of space. These problems may seem obvious or trite, but they were not a factor originally considered with the original project timescale, which envisaged the entire archive could be digitized within 15 months. The most serious problems were persistent and lengthy power shortages during 2019 as low water levels at Kariba Dam following a drought limited hydroelectric generating capacity. The union cannot afford to run a generator, so power cuts meant that most work in the offices ceased, including digitization. Fundamentally, digitization depended upon a reliable supply of power to run a scanner and laptop, and power cuts during June 2019 and then from September 2019 to January 2020 unavoidably delayed progress. These problems are hardly unique to Zambia. Many places across the continent struggle with reliable electricity access or are not connected to a national grid at all. Other practical problems included lack of storage space for files during digitization and deterioration of the building itself, which led to the loss of some documents due to water damage, as described above.

Access and Ownership of Digitized Archives
Although digitized, this collection will not be made freely available and accessible online. Researchers and anyone else wishing to use the collection will still need to visit an archive: the IISH in Amsterdam or MUZ head offices in Kitwe. This might seem like it defeats the point of digitization as it means interested parties will still need to travel to an archive, which may be increasingly impractical at a time when funding for research is likely to become more constrained and international travel less justifiable due to climate change. However, access needs to be balanced against ethics and ownership, and one aspect of the archive will be made accessible as the inventory will be made available on the IISH website. This means that researchers can have some idea about the archive's contents, plan their research in advance and demonstrate the feasibility of their research plans to funding bodies. 22 This was not possible previously as no complete inventory existed and the original partial inventory was misplaced (see above). The few researchers who sought to consult the MUZ archive were therefore making visits that were somewhat speculative.
There is a tension between access, ethics, and privacy. Initially, the aim of the project was to make the digitized archive freely available online. Once cataloguing and examination of the material began, it became clear that there would be serious ethical problems with this. The disorganized nature of the material means that older documents are sometimes contained in the same folders as contemporary material, some of it highly sensitive. Financial informationsuch as recent salary statements and bank details of union officialsand medical records of union members are interspersed with older material. These documents must be preserved and digitized, and one day will be useful for historians, but cannot be made available yet. Here, ethics clearly trump access.
For these reasons, the MUZ leadership were also concerned about making the archive open access and had the final decision on the subject as the physical and digital archive remains the property of MUZ. Ownership and "the very definition of an archive" has been challenged by digitization when an archive can exist in many forms and multiple copies. 23 In this project, ownership of the different forms of the archive is embodied in controlling access to the archive. The agreement reached with MUZ stipulates that those wishing to access the archive will need to contact the union and obtain permission. I think this is an appropriate condition of access and MUZ is open to researchersas evidenced by freely granting permission to digitize their archivebut it is possible that future MUZ leaderships will be less receptive to scholars, especially critical ones.
The physical archive will remain at Katilungu House and part of the project has been to organize and preserve the documents themselves. Following scanning, each file was numbered and stored in acid-free cardboard boxes that were manufactured by a firm in Zambia. The retention of the physical archive is important for several reasons. Twenty years ago, Peter Limb warned about the problem of "document drain," the process of the acquisition of African archival collections by better-resourced Western libraries, a move often justified by preservation. 24 This only adds to a situation whereby "scholars from Africa and other developing countries will find more comprehensive and better-preserved collections" in libraries in Europe and North America "than in their own countries." 25 Information about the Copperbelt's past should be accessible to those who live there.
The other reason is that inequalities between the Global North and South are also manifest in a "digital divide" and therefore "digitization has the potential to exacerbate unequal access to historical sources," by privileging audiences in European and North American universities over local audiences. 26 Many internet users in Zambia, and elsewhere on the continent, connect only using their mobile phones and it is often impractical to read and store large documents on phones, while downloading digitized files can consume large amounts of data. I hope that the archive will be a resource for the tens of thousands of current and former union members as well as researchers. During the digitization process, mineworkers visiting the offices on union business often came to ask what was happening (the cataloguing and scanning of stacks of documents being a visibly different activity to normal union business) and expressed interest in the project. Retention of the physical archive is therefore important for users who do not own devices such as laptops to copy digitized documents or may be less familiar with using computers to access material.
The retention of the archive in Katilungu House is not without problems. The building itself is a large one (see Figure 1) constructed in the early 1970s when the union was still a sizeable and powerful organization. Since the union has declined in size and faced financial problems, more sections of the building have been rented out as office space for other organizations and companies, even a nightclub. 27 At present the physical archive is in an otherwise unused storeroom, but it is a small room without sufficient space even to install metal shelving to house the documents.
It is also possible that if the union faces more serious financial difficulties then more offices in the building will be rented out and available space for union activities will become constricted, and the physical archive will have to be moved again. The possible relocation of the physical archive within Zambia was discussed with union officials but the only plausible place that it could be housed that is relatively nearby is the Zambia Consolidated Copper Mines archive in Ndola, which houses the records of the mining companies. 28 However, this archive is also short of storage space, making relocation there impractical.

Knowledge Production in Central Africa
There is an unresolved tension in this project relating to knowledge production and the implications of making certain kinds of archival sources more accessible to researchers. At a global level, there are few digitized collections on African labor and social history and, in this sense, the digitization of the MUZ archive makes a marginalized subject more accessible and visible. At a local level, however, the amount of available information about male mineworkers on Zambia's Copperbelt and their history is enormous, including work by African mineworkers and trade unionists themselves. 29 As Larmer has observed, "few aspects of African economy and society have been as fully studied as the copper mines of Zambia and their workers." 30 The lives and experiences of these mineworkers have been at the center of Zambia's historiography. Indeed, in her 1989 book on domestic workers in Zambia, Karen Hansen noted that the existing literature "readily leaves the impression that contract migrant labor to farms and mines was the only, or the main, form of which labor which developed in the colonial period," an observation that remains largely accurate. 31 What is possible then is that by making sources on MUZ more accessible this project will reinforce the bias within Zambian historiography towards male mineworkers and inadvertently contribute to the marginalization of other topics and perspectives. The MUZ archive is largely about male mineworkers, though not entirely, as discussed below. The archival record 27 At the time of writing, MUZ occupies only the top two floors of the building. 28 Munene, "Mining the Past," 9-13. 29 Matthew Mwendapole, A History of the Trade Union Movement in Zambia up to 1968 (Lusaka: Institute for African Studies, 1977). Mwendapole was an elected official in branches of a predecessor union to MUZ at Broken Hill (Kabwe) and Nkana. 30 Larmer, Mineworkers in Zambia, 2. 31 Karen Hansen, Distant Companions: Servant and Employer in Zambia, 1900-1985(Ithaca: Cornell University Press, 1989 invariably privileges certain historical actors and the kinds of histories that are written often reflect sources that are most accessible to historians. Identifying sources for the histories of marginalized groups and perspectives often requires more work because the sources are less accessible, and therefore giving male mineworkers greater prominence in the available sources could discourage work on other topics. Samuel Daly makes this point that an unintended consequence of digitizing records is that it "may end up making historians less likely to seek out materials that are scattered or difficult to access." 32 Certainly, one aim of this project is to encourage research about Zambia's Copperbelt and African labor history, but it may do so at the expense of other topics. I think this is a question that all digitization projects must consider carefully: How do we strike a balance between preserving the records and heritage of African societies but avoid reproducing existing gaps and bias? I do not have a ready answer to this question, and it is not an aspect that I considered at the outset of the project. One potentially fruitful way of engaging with this question may be reflecting on the motivations for undertaking digitization projects, and how different archives come to be selected for digitization. Who selects archives for digitization? Mostly those from outside the continent. A 2017 assessment of the British Library's Endangered Archives Programme reported that 80 percent of the grant holders for projects on Africa-focused collections were based at Western universities. 33 Some scholars have accused historians and archivists of developing a "digital savior complex" in embarking on efforts to save "their" archives. 34 Having the priorities of digitization projects primarily shaped or determined by the interests of external researchers may contribute to these projects perpetuating existing historiographical biases.
Digitization projects are often the intersection of personal or professional interests and institutional priorities. The MUZ project came about from the intersection of the wishes of MUZ officials to preserve their own records, the institutional priorities of the IISH in expanding African labor and social history collections, and my own research interests in the mining industry. Certainly, I did not embark on this project solely out of the goodness of my heart, but because I have a longstanding interest in the topic. These kind of collaborations between the Global North and South can be fraught and need to carefully consider in whose interests the projects are being carried out. Many scholars are skeptical about such collaborations. For instance, many Africa-based participants in meetings of the Aluka project, a large-scale effort to create a digitized archive of Southern Africa's liberation struggles funded by the Mellon Foundation, regarded it from the outset as an attempt to "appropriate Africa's patrimony and subvert intellectual property rights and national heritage." 35 The historiographical bias towards Zambia's mineworkers does mean that there is a ready scholarly constituency who will use it. 36 The Copperbelt was the subject of considerable academic interest for several decades from the 1930s until the early 1980s, with a modest revival in interest since the early 2000s. The first generation of scholars were specifically drawn to the region by the impact of industrial mining on African societies, and knowledge production was closely intertwined with African labor on the mines. It was a major strike by African mineworkers in 1935 and the perceived need in the colonial government to understand social changes in African societies that provided the impetus to establish a research institute whose work and legacy has had an enduring influence on African studies: the Rhodes Livingstone Institute (RLI). 37 Underlying much of the academic work in the colonial period was the assumption that Zambia was moving inexorably towards a modern industrial economy, whose archetype was Western Europe, with its attendant social relations. 38 MUZ and its predecessor unions were closely studied in this light. The emergence of a mineworkers' union was regarded by scholars at the time as yet another confirmation that developments in colonial Zambia were following patterns already established Europe. As RLI director Max Gluckman argued, strikes by African mineworkers in the 1950s and the emergence of a militant union leadership "brought into the open the emergence within the African urban population of affiliations based on what we can call 'class principles'," and this was something that could be expected "from the history of Europe." 39 The formation of a large, well-organized trade union for African 35 Allen Isaacman, Premesh Lalu, and Thomas Nygren, "Digitization, History, and the Making of a Postcolonial Archive of Southern African Liberation Struggles: The Aluka Project," Africa Today 52-2 (2005), 59. 36 The number of scholars working on the Copperbelt is relatively large, by the standards of Africanist scholarship. At the African Studies Association of the UK conference in 2016, for instance, there were four panels entirely devoted to Copperbelt research, with 16 papers in total. 37 M.C. Musambachime, "The University of Zambia's Institute for African Studies and Social Science Research in Central Africa, 1938-1988," History in Africa 20 (1993  workers therefore rendered the Copperbelt a familiar place, in the sense that it came more closely to resemble industrial areas in Europe which scholars were familiar with, and in the post-war period trade unions were mass organizations in Western Europe and North America. MUZ continued to be a focus of study in the post-colonial period, with work on the relationship between organized labor, the state and economic development, a debate over whether Zambia's unionized mineworkers constituted a "labor aristocracy." 40 I think that there are ways that the MUZ archive could be used to reflect on and re-visit this earlier work, albeit at the peril of reinforcing the abovementioned historiographical biases, but also to address other topics. The archive has a different chronological scope than other archives in Zambia, as it is focused entirely on the post-colonial period. 41 The National Archives of Zambia is primarily an archive of the colonial period and contains few documents after the late 1970s as, following the declaration of the oneparty state, the ruling party UNIP took over responsibility for documents deemed to be political. 42 Other post-colonial documents at the National Archives were destroyed or sold for scrap paper to market traders in the early 1990s. 43 This is not unusual. Florence Bernault contrasts "rich, organized and biased" colonial repositories with the seeming "meagerness and incoherence" of post-independence archives. 44 Across the continent, as Samuel Daly notes, "the sources of post-colonial African history are increasingly found outside of state archives." 45 The chronological focus of the MUZ archive corresponds with something of a lacuna in studies on the Copperbelt. Much of the existing scholarship focuses on the colonial period and comparatively little has been written on the history of recent decades, with the obvious exception of Larmer's work. 46 Much, I think, remains to be said about the transformation of the Copperbelt from a place of relative prosperity in the 1970s with secure employment and relatively comprehensive social and welfare services to the mass unemployment and poverty of the early 2000s. The classic account of this is James Ferguson' Perhaps one of the most valuable aspects of the MUZ archive is that it is primarily composed of documents by Africans about the world as they saw it during a period of great crisis. Though often seemingly mundane or procedural trade union matters, many of the documents are about efforts by African trade unionists and mineworkers to navigate, both collectively and individually, a period of disorientating and disruptive social and economic dislocation. These are not the records of external powers, colonial officials, or those studying African peoples, though this is not to say that documents in this archive are an unproblematic representation of Copperbelt life or should be read uncritically.
I hope that new avenues of research will be facilitated by this archive as much of the material relates to comparatively understudied topics. There is considerable material on healthcare and disease and how Copperbelt society grappled with the emergence of the HIV/AIDs epidemic, reports on accidents on the mines, details on the international links of MUZ with trade unions across Southern Africa, and beyond. Alongside this is much conventional labor history material and the voices of ordinary mineworkers themselves like the drillers who protested they were "working like slaves in our Country." 50 Many files contain what might be regarded as incidental material: leaflets for restaurants, promotional magazines, CVs from job applicants, details of businesses renting MUZ-owned properties, that will be useful for social histories of everyday life on the Copperbelt.
The MUZ archive also covers the period in which gendered roles and cultural expectations changed as women entered the mining workforce and the ranks of the union, as until recently women were excluded almost entirely from jobs in the mining industry. 51 There has been some work on women's history on the Copperbelt and the role of gender in shaping the realities of 47  life and work on the Copperbelt, but it is limited in comparison to the work on male mineworkers. 52 There is much in the MUZ archive that could support such worksuch as the records of the MUZ Women's Advisory Committeeand this includes fascinating material on the role of gender and women's protests during the decline of the industry. In October 1989, for instance, rumors that fees were to be introduced at the mine hospital provoked demonstrations in Kitwe, Kabwe, and Mufulira by miner's wives. In Kitwe, a crowd of 500 women dressed in work overalls and miners' hard hats "took to the streets and besieged the ZCCM Central Offices." 53 However, despite the presence of such material, the archive is primarily focused on the life and work of male mineworkers. Most union officers and all union General Secretaries and Presidents have been male. In suggesting the potential for the MUZ archive to support relatively neglected research topics, I do not wish to misrepresent the central focus of the archival material.

Conclusion
Digitization is not a panacea for archival preservation and accessibility. The process of digitizing the Mineworkers' Union of Zambia archives raised a variety of practical and theoretical problems that likely have lessons for future digitization projects. The account presented here of deteriorating archival documents, inadequate storage, disorganized material, and persistent power shortages doubtless corresponds with the experience of historians and archivists at many archives on the continent. The circumstances of Zambia's socioeconomic development since the 1970s has shaped the form of the MUZ archive, but similar processes have shaped archival material in other parts of Africa.
Identification and collection of material proved difficult. What is referred to in this article as the MUZ archive was essentially created by this project as the parameters of the archive were not clear and, with material distributed around the union officers, then the widest definition for the archive was all printed material stored within the four walls of Katilungu House. The MUZ archive may therefore expand in the future not only as the union still exists and continues to produce records but also because a more thorough search of the union's offices may yield more documents from the 1970s, or material from an earlier period. I suspect that archival preservation projects seldom encounter neatly delineated collections of historical documents awaiting preservation and digitization, and the choices about what to digitize are difficult decisions.
The ideals of access (making information freely accessibly online), ownership (ensuring that creators of the records retain control), and ethical considerations can be mutually exclusive, and it may not be possible to achieve aims that satisfy each of these criteria. Sensitive and recent material interspersed with older records due to the disorganization of the archive meant it ethically impossible to make the MUZ archive freely accessible. Instead, digitization replicates more conventional forms of research as researchers will still have to physically visit an archive to access the material.
Creating two digital copies of the MUZ archive and retaining the physical archive in the union's offices with the union controlling access to archival material was the solution to competing aims of access and ownership. Providing both a physical and digital copy of the archive in Kitwe will hopefully enable Copperbelt residents to access this material but the continued location of the archive in the union's office in Katilungu House is not without problems. There is a risk that archival material could be damaged due to the state of the building, as happened during the project, and MUZ may face further financial and spatial constraints and be forced to move the archive.
The latter part of this article has considered questions of archival preservation and knowledge production. It is clear that making more historical sources accessible can correct a bias in the historical narrative at one level but potentially perpetuate a bias at another level. Although digitized collections on African labor history are few, and this collection was at risk of becoming unusable due deterioration of the physical documents, Zambia's unionized mineworkers' have been a central focus of the historiography. The archive is being preserved so that it can be used, and while there is material within it to support research into understudied topics, the most obvious use of this material is to study MUZ itself and the recent history of Zambia's mineworkers. Although questions about preserving documentary heritage while avoiding reproducing existing bias do not have easy answers, this article has suggested that such questions can be approached by reflecting on the motivations for engaging in particular digitization projects and who is involved in the selection of these projects.