Public sector reform and the dead archive

The history of archives since 1975 is characterized by improvisation, rescue operations, and attempts at organization on the hoof. During the first fifteen years of independence, the Mozambican government did not have any specific policy on the management of archives in the public administration. With minor modifications, the new government continued to follow the same system of classification of files used by the Portuguese colonial administration after it established the Arquivo Histórico de Moçambique (Mozambique Historical Archive or AHM) in 1934. In the first two years of independence, the government continued to use official papers and files with the seal of the colonial state. The colonial signature ‘A Bem da Nação’ (For the Good of the Nation) was replaced with ‘Unidade, Trabalho e Vigilância’ (Unity, Work and Vigilance), and then, a bit later, with ‘A Luta Continua’ (The Struggle Continues). But the format, form and style of bureaucratic documents continued almost unaltered. This was a classification system that made it easy for the AHM to accumulate a relatively well-organized collection of colonial records.

After the fall of the colonial regime, the AHM played an important role in rescuing files from the various administrative branches of the government. In the tense and convoluted period of transition, former colonial government agents destroyed files in a deliberate attempt to bury sensitive information. It took a courageous and timely campaign by academics from the University of Lourenço Marques (later Eduardo Mondlane University) to collect files across the country and bring them to the safety of the AHM in the capital city (Costa Reference Costa1984). However, this rescue operation did not take account of the infrastructural and technical capacities of the AHM to keep and organize files (Tembe and Mangue Reference Tembe, Mangue and Froner2018). From reopening its doors in the early 1980s until the present, the national archive has continued to work on the classification and cataloguing of the files from the colonial period collected in the rescue operation. Consequently, the documentary material produced by the socialist government after independence has not received any attention from the AHM. Over the years, the bureaucratic paperwork of the independent state has been accumulating in the basements, attics and verandas of public buildings. There was no policy or state directive on what should be done with this material until the formal end of the socialist experiment in 1990.

Nevertheless, the government had to confront the reality of piles of documents which, on the one hand, took much needed space in public buildings, and, on the other, impaired the normal bureaucratic function as administrative information was increasingly hard to locate. In 1992, as the country was emerging from a devastating civil war and looking forward to its first democratic elections, the Council of Ministers passed the first legislation on archives. This introduced the Sistema Nacional de Arquivos (National Archives System or SNA), in which the AHM was given the responsibility to supervise, coordinate and manage all the documentary archival material of the state. One of the central objectives of the SNA was to ‘preserve the archival legacy’ of the country and ensure that the ‘historical and cultural heritage of the Mozambican nation’ was ‘preserved and valued’.Footnote ¹ The legislation introduced a three-tiered classificatory system for documents: current, intermediary and permanent. Central archives were to be created at the provincial level, where all the files from the various permanent archives of local administration would be brought together. After ten years, the files in the central archives had to be sent to the AHM. Thirty years after the date of their production, the files could be opened for public consultation. The legislation prohibited the destruction of permanent files without due consultation with the AHM, whose director and leading personnel were the presiding officers on the documents’ evaluation commission. Destroying files was liable to criminal prosecution. This was, by all accounts, a very optimistic piece of legislation.

The euphoria of independence had produced a historically conscious society and an urgent need to preserve the past, a task viewed as indispensable for nation building. Although the end of the civil war produced the first legislation on archives in the postcolonial era, it did not lead to the same enthusiasm about preserving the past. The main discourse of the post-civil war years revolved around forgetting the past and looking forward to a better future (Igreja Reference Igreja2008). Although, as Anne Pitcher has pointed out, this was an effort to forget from above, there was little energy from below to replicate the deeds of 1975 as far as preserving the documentary evidence of the past was concerned (Pitcher Reference Pitcher2006). The country was too busy trying to put itself back together after a harrowing conflict to embark on a history rescue campaign like the one envisioned immediately after independence.Footnote ² As a consequence, the well-meaning legislation on archives remained, for the most part, a dead letter.

The gap between the good intentions outlined in the legislation on archives and the material reality of austerity-stricken Mozambique led to the inevitable mushrooming of dead archives. In the late 2000s, for example, the need for additional office space to accommodate the growing number of staff at the provincial directorate of finance in Inhambane resulted in the provincial archives being squeezed into a corner of a dark room where files were piled up from floor to ceiling in no particular order.Footnote ³ Not only did most files remain unorganized in the buildings where they were produced, in some places entire collections were destroyed or ended up in the informal market as wrapping papers. When one of the authors of this article inquired about the archives of the socialist period in the Maputo municipal building in 2015, he was told that the archive did not contain any file dating from before the year 2000. He was told that local employees sold most files to street vendors who used the papers to wrap roasted peanuts.Footnote ⁴ In hard times, civil servants were desperate to make ends meet and to complement their meagre salaries with whatever they could lay their hands on. A famous saying in those days was ‘O cabrito come onde está amarrado’ (The goat eats where it's tethered). Budget cuts in public services and the privatization of state enterprises sent thousands of people into new forms of employment as public sector jobs dwindled or disappeared. To make do, informal markets, which were illegal in the socialist years, sprouted exponentially in every corner. The high demand for wrapping paper was extremely tempting for civil servants holding the keys to basements and attics filled with ‘old’ documents. Distressing anecdotes of how archives ‘disappeared’ can be heard in many government institutions. But austerity does not account for all of these stories. Lack of awareness about the uses of archival documents by government officials has also contributed to the tragic fate of archives. The disturbing end that one particular archive met in the northern province of Niassa illustrates the point. In the early 2000s, the administrator of the newly built building for Niassa's department of finance in Lichinga ordered the burning of entire files because, as one of the authors was told, the administrator did not want ‘old papers in his new building’.Footnote ⁵

In the euphoric years immediately after independence, there were several incidents of local administrators who burned files as an act of ‘decolonization’ and ‘escangalhamento do aparelho de Estado colonial’ (destruction of the colonial state apparatus). But rescue teams from the AHM saved many files from such a fate. No such teams existed in the post-civil war era – not in the AHM and much less in the public administration. If many files had disappeared in the areas directly affected by the civil war – as buildings of local administrations were ransacked and set on fire – a great deal were lost during the economic crisis of the 1990s and early 2000s. As a government report acknowledged in the mid-2000s, the situation of archives in the public sector was dire:

It comes as no surprise that, under the conditions of scarcity that characterized the history of Mozambique since the colonial period and were exacerbated by the adoption of structural adjustment measures in 1987, the ‘area of documentation’ was the last to receive any attention in the distribution of meagre state resources. To address archives’ state of neglect, in 2007 the government introduced new, ambitious legislation creating the Sistema Nacional de Arquivos do Estado (National State Archives System or SNAE).Footnote ⁶ Although the new legislation revoked the 1992 law, it maintained the basic principles outlined in the old law. The major change was the replacement of the AHM by CEDIMO (Centro Nacional de Documentação e Informação de Moçambique, a branch of the Ministry of Public Administration) as the main body responsible for coordinating the organization and management of state archives. The AHM was reduced to the role of guardian of permanent files. Nevertheless, the acting director of the AHM has a prominent seat on the board of the SNAE. The director of the AHM also serves as vice president of the Conselho Nacional de Arquivos (National Archives Council or CNA), a ‘consultation organ of the central directorate of the SNAE’ that defines the ‘national policy on public and private archives and the technical-normative orientation in the management of files’.Footnote ⁷ The legislation also introduced evaluation commissions, the Comissões de Avaliação de Documentos da Administração (CADAPs), which should exist in all government institutions and are tasked with evaluating, selecting, listing and classifying files. These are multi-technical teams comprised of local civil servants subordinated to CNA. In theory, these teams should include a trained archivist.

Unlike the first attempt at organizing the archives, this time around the government seemed determined to make palpable changes through the Ministry of Public Administration. This is discernible in the various short-term courses provided to selected civil servants on file management by the Instituto Médio de Ciências Documentais, a school of documentation science, in coordination with CEDIMO. CEDIMO began a campaign to promote the new rules on file management in the public sector. They first compiled and published a normative manual with detailed guides for file management (CEDIMO 2006). Based on Brazilian models of file management, this guide was expanded into a sixty-page booklet with step-by-step guidance and state-of-the-art procedures on the management of bureaucratic paperwork and archives (CEDIMO 2009). Among many other things, the booklet contains charts for proper book listing and photographs contrasting a well-organized archive with a messy one. CEDIMO's own Boletim dos Arquivos Nacionais (National Archives Bulletin, a quarterly bulletin launched in 2007 specifically to raise awareness of good practice in file management) provided civil servants with summaries of the booklet's major points and key concepts. The cover of the first number of the bulletin reads: ‘Para memória institucional, salvemos os nossos documentos’ (For institutional memory, let's save our documents) (CEDIMO 2007). Civil servants were called on to familiarize themselves with the concepts of archive, file, preservation, and the three-tier lifecycle of documents.

The new SNAE was indeed an assertive policy that pointed in the right direction. It seemed to follow the general mood as Mozambique was experiencing relative economic stability with double-digit growth and a sound reputation among international financial agencies and donors. In fact, the effort to organize the archives was part of a larger, donor-funded project to reform the public sector according to the dictates of good governance, transparency and accountability.Footnote ⁸ Sadly, it has been more than a decade since its implementation but few fruits have been harvested from the effort to ‘save our documents’ and ‘preserve our institutional memory’. Although in a few institutions the management of files improved considerably – the Maputo city council and the Mozambique Reserve Bank being the best examples – in most government bodies dead archives not only persisted but expanded in dramatic ways.

Despite the detailed guidelines of CEDIMO's booklets and the very progressive legislation of 2007, critical gaps in the legislation and much larger structural challenges have rendered the SNAE no more than a well-intentioned but ultimately failing project. First, the new law is not precise about who has actual responsibility for sending files marked as ‘permanent’ to the AHM – CEDIMO, the central archives in the provinces, or the AHM itself. Consequently, no permanent files from the post-independence era have been deposited at the AHM. Second, despite an initial effort to hire archivists or train civil servants in file management, the Ministry of Public Administration fell far short of its goal to provide all institutions with capable personnel to handle files. The CADAPs exist only on paper and, in most institutions, no one is employed full time in the local archive. As a new cycle of economic decline began in 2015 and the very donors who had supported the project began pushing for further cuts in public expenditure – which include shrinking the size of the government – that goal has become even more elusive.

The consequences of the SNAE's failure and the government's interrupted efforts to recruit and train civil servants in file management are all too visible in most government institutions. In those few offices where reforms had been initiated, the organization of archives did not go past the stocking of files in proper boxes and on bookshelves. A researcher visiting the basement of the Maputo province administration building in Matola city will be happy to see that the files are all on shelves and in bookcases but those files follow no classificatory order and no catalogue exists. This is where reform under the SNAE stopped. In most government institutions where reform never took place – as is the case with most local administrations, far away from the capital city – dead archives have continued to grow.

The tense relationship between the government and communication, especially in recent years, is an additional barrier to the successful implementation of the SNAE. While this relationship may not be the weightiest factor in the SNAE's failure – other factors are more important – it plays a significant role in shaping the nature of the dead archive and its content. It is here that the contrasting character of governments in the socialist and post-socialist period becomes more apparent.

Reverse transformations and the politics of secrecy

On close examination, the dead archive reveals an uneasy relationship between the ideals of good governance, which put a high premium on free access to information, and the bureaucratic culture of Mozambique's public institutions. The socio-political transformations that the country has undergone over the last three decades have had an impact on the government's relationship with archival records, its own memory, and the ways in which it communicates with its public (Coelho Reference Coelho2004).Footnote ⁹ While in many ways the proliferation of dead archives is a product of conditions of austerity and the failure of the SNAE, it also reflects the transition from a socialist-oriented government that was transparent about its activities to a multiparty government that paradoxically is highly secretive and has a difficult relationship with communication and institutional memory.

Although the socialist regime that lasted from 1975 to 1990 had no policy or legislation on archives, it was nevertheless an archive-minded state. It referred to its archive whenever new policies were elaborated and announced to the public. Whether in announcing policies or in punishing deviants, the socialist state was generally transparent. This becomes evident in the quagmire of documents that have survived and that populate the dead archive. For example, when the Minister of Agriculture, Joaquim de Carvalho, was demoted from his post and then expelled from the party in 1978, the state published a detailed communiqué to explain the reasons for his demotion.Footnote ¹⁰ When the Secretary of the Mozambique Women Organization's Department of Social Affairs, Esperança Muthemba, was found guilty of ‘má conduta’ (bad behaviour) and expelled from the organization and from the party, newspapers published detailed accounts about how she fell from grace.Footnote ¹¹ Provincial governors, such as Niassa's Aurélio Manave (1975–83) and Maputo's José Moiane (1977–88), kept annotated files on demoted or sacked high-raking party/state officials.Footnote ¹² Carefully documented, such cases were often objects of debate in meetings within the local administration. Governor Manave often deployed them to remind his subordinates to be alert to observing the rules of appropriate behaviour, dutiful responsibility in the workplace, and citizenship. Governors had an archive of high-profile state functionaries whose disciplinary affairs were public matters.

The communicative transparency of the socialist government invited citizens to participate energetically in public affairs and in the day-to-day business of state administration. This does not mean that there was political tolerance for views contrary to the party line, however; challenges to the party's ideology and policies were subject to punishment (Machava Reference Machava2011). But what transparency meant and allowed was a circumscribed space for public participation along the lines of vigilant citizenship.Footnote ¹³ Citizens wrote letters directly to the president, to the party secretary and to provincial governors to discuss matters of public administration or to complain about the inefficiency of a given branch of the government. Newspapers and their readers’ pages were dynamic spaces of public participation, even though at independence the levels of illiteracy were very high.Footnote ¹⁴ Ministers, provincial governors and rank-and-file officers kept scores of newspapers and magazines with annotations and public commentary about their areas of responsibility. When an article in the main daily newspaper Notícias painted a very negative picture of a cotton-processing factory in the Niassa city of Cuamba in March 1981, Governor Manave had the entire staff of the administration in Cuamba busy finding out who had provided the damaging information to the reporter and fixing the problem.Footnote ¹⁵

A reader named Dinho Neto who complained in the readers’ page of the weekly magazine Tempo about the unprofessional treatment he received from a nurse in Quelimane in 1983 was informed by the director of the local district health department that the unprofessional nurse was polygamous and had been sacked from his post.Footnote ¹⁶ Ana Paula Coelho Perdigão, chief medic of the city of Xai-Xai, was obliged to reply to a letter from a group of five citizens of that city who reported to Tempo that a woman had given birth on the street.Footnote ¹⁷ Alberto Massangaia, administrator of the district of Chiúre in Cabo Delgado, wrote a lengthy letter in response to a denunciation by one Chomanga Bororo about a local functionary who was courting a student and drawing her into ‘sexual corruption’. The administrator, having been informed that a criminal case had been lodged against the denounced ‘agent of corruption’, lamented that such a ‘discrediting’ case had been raised publicly rather than directed to the administration.Footnote ¹⁸ Vigilant citizenship kept party/state officers on their toes. No one wanted to be found short in their revolutionary duties. The public gaze seemed ubiquitous and state agents felt obliged to address people's concerns in print media. Governance was, in that sense, a public act.

The transparency and participatory nature of the socialist government compelled state agents to be more conscious of institutional memory. They archived every document that passed through their hands – from the most pressing matters relating to state security, food production or economic projects to the most mundane aspects of everyday life. The frontier between private and public, between high politics and the domestic, was very porous. The archive of the socialist government is populated with stories of domesticity, of intimate relationships – marriage, sex, adultery, love affairs and gossip – which are often associated with or put alongside files about the security of the state and the socio-economic (in)stability of the nation. This is in great part the result of a highly interventionist government for whom every aspect of social life was its prerogative. The archive of the Mozambican socialist government confirms Ann Stoler's argument that ‘matters of intimacy were matters of state’ (Stoler Reference Stoler2001), or, as Jean Allman put it, ‘matters of state are also matters of intimacy’ (Allman Reference Allman2013).

The post-socialist government, by contrast, is characterized by high levels of secrecy and minimal public engagement. Political pluralism – which the ruling party embraced almost by force in 1990 – resulted in a less accountable form of governance (Coelho Reference Coelho2004). Competition with other political forces pushed the ruling party to be more secretive about its business, even though an independent media has grown over the years since 1990. Unlike its predecessor, the government gives hardly any explanation when a minister or a governor is demoted – even if issues of inappropriate conduct are widely known to the public. Policies are only partially announced and the participation of the public has been truncated. Communication is inadequate: for example, the webpages of ministries and public institutions are generally poorly maintained and are often offline.

While advocating and legislating on archival organization appear to follow the rules of good governance and transparency, the neglect and, in some cases, the deliberate acts of destruction of archival material go hand in hand with efforts to legislate and improve archives. In fact, the archive carries a threat to those in power in a multiparty context. One fascinating example of this paradox is the persistent effort by the ruling party to have the saga of the liberation struggle documented in history books. This desire to write the history of the struggle is even more acute among the veterans of the liberation, who have produced more than forty biographical memoirs over the last decade in which the struggle is the central topic (Souto Reference Souto2013). Yet, that same party has not allowed its archive to be opened for public consultation, granting only partial access even for those scholars commissioned to write the official history of the liberation saga.Footnote ¹⁹ As historian João Paulo Borges Coelho has convincingly argued, this ironic situation results from the problematic ‘neighborliness relations established after independence between politics and history’, in which politics held the ‘monopoly of explanations of the past’ (Coelho Reference Coelho2013: 21). The ruling party, writes Borges Coelho, has established itself as the ‘single source of authority’ in the production and dissemination of knowledge about the country's past. The final product of this marriage between politics and history was the predominance of a single nationalist narrative, ‘codified as a script’, which has ‘become an instrument to legitimize’ the party's hegemonic ‘authority and render it unquestionable’ (ibid.). Therefore, opening the party's archive would destabilize the script by giving space to myriad interpretations that could undermine the established narrative on which Frelimo's legitimacy rests. Although the proliferation of biographical memoirs has unearthed details about the struggle, some of which have led to polemical debates, the memoirs have not undermined the script; they have simply added new layers to it.Footnote ²⁰

The deterioration of the political climate in the last decade and the low-intensity armed conflict that engulfed the central region of the country (2013–15) reinforced the secretive character of the post-socialist government. During the sixteen-year civil war (1976–92), general aspects of the war effort were openly debated in the single-party parliament, including budgets and armament acquisitions. Excerpts of proceedings of these debates can be found in the archive of the socialist period. Now, the ruling party must contend with the fact that the same opposition party waging war in the bush also has seats in parliament. Democratic institutions and constitutional procedures along the lines of good governance and accountability have collided in dramatic ways with the odd reality of Mozambique's political environment. The most recent scandal around hidden loans that the government obtained from international banks without the due approval of parliament is one of the outcomes of the politics of secrecy that has dominated the post-socialist period.Footnote ²¹ As the means of communication within the state administration are moving from traditional printed paper to electronic files, new cases of missing archives associated with faulty or mishandled digital archives emerge.

The culture of bureaucratic secrecy was less insidious in the governance of the socialist state in Mozambique, precisely because of the general sense of righteousness of the socialist project and the absence of a formal political opposition. While punitive measures were put in place to deter sensitive state secrets from leaking out to the public – especially military intelligence, which could fall into the hands of the rebel movement and its allies in apartheid South Africa – the socialist government did not have any legislation or administrative rules pertaining to the concealment of information. Law 2/79 of 1979, which established and defined the punishment of crimes against the ‘People's Security and the Security of the People's State’, dedicated one section to ‘agitação’ (agitation) and another to ‘boatos’ (rumours). In the section relating to agitation, Article 35 established that it was a serious crime ‘to publicly, by any means, interpret in bad faith the orientations and the laws outlined by the Frelimo party and state or the objectives which those orientations and laws aim to achieve’.Footnote ²² Similarly, Article 36 of the same law determined that any Mozambican citizen ‘who makes or publicly reproduces, or by any means publicizes or tries to publicize statements or news he/she knows are false … is punished with a sentence of two to eight years in prison’.Footnote ²³ Clearly, it was not the access to information per se that concerned the socialist government, but the instrumentalization of that information, the spread of rumours and ill-intentioned interpretations of the party line, which could undermine the socialist project.

As the ruling party was moving towards the adoption of a multiparty political system and a market economy, it became aware of the need to set out procedures in information management for civil servants. The first of these procedures came in the form of a decree in 1989 (Decree 36/89 of 27 November), which defined – though in ambiguous terms – when public functionaries could provide information to the public. A hierarchy of how and what kind of information civil servants could provide was established, including the penalization of non-observance of these new norms. In Article 4, point 2, the decree states: ‘all the information which for its complexity needs major consideration should be submitted to the functionary higher in the hierarchy by the person to whom the information was requested’.Footnote ²⁴ Point 3 of the same article states: ‘Incorrect information which leads the requestor to error will result in civil or criminal responsibility for the functionary who has provided such information, regardless of the disciplinary measures that might be taken.’Footnote ²⁵ While imposing a rigid, hierarchical structure in information management, the decree left ample room for interpretation, especially on how one evaluates the ‘complexity’ of the information that needs further consideration. Was it a matter of the details of the information requested? Or of the subject of the inquiry? How could one evaluate the degree of ‘complexity’ of requested information? And who should do so? This ambiguity speaks to the government's sudden concern with what type of information the public could be given access to and how this should be done. Ill-intentioned interpretations of the party's orientations were no longer the government's preoccupation, but the kind of information that could be open to public scrutiny.

This new orientation of the government had two important implications for file management. First, the decrees stressed the category of ‘state secrets’ – that is, classified files that government officials consider to be associated with the security of the state. But because the decrees did not provide a blueprint for which files could be classified as ‘state secrets’, functionaries had the tendency to classify most of the files as secret. The most recent pieces of legislation on file management and the statutory role of state functionaries (Decree 30/2001 and Law 14/2009 respectively) introduced the language of ‘technical complexity’, which gave functionaries the discretion to classify documents as well as to remain silent about their content, even after the end of their term as civil servants.Footnote ²⁶ Law 14/2009 reiterated that, should a state functionary or agent ‘disclose or allow the disclosure of classified information which he knows as a function of his service’, he or she would be dismissed.Footnote ²⁷ Second, this emphasis on penalization created an environment of weariness among civil servants tasked with managing government files, making them uneasy about giving access even to files that were not classified.

The tightening of the legislative control of information is only one side of the coin. The other side is the role that functionaries play in concealing information, not necessarily for fear of breaking the law, but out of their sense of political loyalty in a new environment of multiple political affiliations. There is a great deal that can be said about the changing nature of state functionaries with the adoption of neoliberalism, most of which falls outside the scope of this article. But the factor most connected to the argument we advance here relates to the reorientation of functionaries’ vocation from self-denying ‘servidores do povo’ (people's servants) to partisan representatives of the ruling party (Coelho Reference Coelho2004: 10). Although the ideal of a self-denying public servant during the socialist period was for the most part only that, an ideal, it still held sway in the ways in which state functionaries positioned themselves vis-à-vis the public. There was a sense of duty bred by the ethos and discourse of the people's revolution, even if by the mid-1980s it had already started to erode.

Finally, secrecy has reinforced the role of orality as the privileged medium of communication within the state. Orality in itself is not new, but in the socialist regime oral messages were often recorded, transcribed and archived (Gonçalves Reference Gonçalves2013). Whenever the president, a minister, a governor or a district administrator held a private meeting with subordinates or visitors during the socialist period, a team of assistants were in the venue to record the entire session. The archive of the socialist government is filled with meeting minutes in which the conversations of the interlocutors were transcribed word for word. Documentation was a state priority, a branch and a form of governance in and of itself. The post-socialist government dispensed with this practice. Meetings may be recorded, but the transcripts hardly ever appear in the archival record. Despite its hard-line discourse and obsession with internal enemies planning to undermine the revolution, the socialist government was much more open, much more ready to communicate with its public, and less concerned with screening information before archiving it (Machava Reference Machava2011). The archive of the post-socialist government is an ‘unpalatable’ collection of cold reports about development projects, lists of functionaries, and dry summaries of administrative expenditure. It no longer contains the compelling dramas of everyday life minutely described. The transition to multiparty democracy resulted in much tighter political gatekeeping and a secretive state. Guarding, concealing and hindering access to information became the hallmarks of the post-socialist government. This is reflected in its archive and has profound implications for the ways in which knowledge on independent Mozambique is and can be produced.

Getting around the front desk

Who can access the dead archive? And how can they do so? Mozambique's Law of the Right to Information (34/2014) is among the most progressive pieces of legislation in Africa, if not in the world.Footnote ²⁸ It grants the right of access to information to any Mozambican citizen, provided that the information is not classified. The law is in line with the SNAE, and it establishes the obligation for all public institutions to ‘maintain archives available’ and open for public consultation. The classification of information and its ‘qualification’ as state secrets must be carried out by a professional and must be justified according to the rules outlined in the SNAE legislation. Framed in the language of good governance, public accountability and participatory citizenship, the law advocates for a ‘public administration based on the freedom of access to documents and public archives’. Citizens can request access to the archives ‘without the need for the applicant to demonstrate a legitimate interest in and right to their access, as well as the purpose for which information is intended, except for the restrictions provided for in this Law and other legislation’.Footnote ²⁹ Requests to access information must be in written form. According to the law, the request should be addressed to the ‘functionary responsible for the management of files and archives’ in any given public institution. Oral requests are accepted, but the functionary is required to type them up so that a written record of the request is kept in the administration.

Most of the files in the dead archive are neither classified nor catalogued. By law, any document thirty years old is by default permanent and thus open for public consultation. Any current files that are not classified, regardless of age, are open for public consultation. Therefore, in principle, the dead archive is open for public access without any justification as to the purpose for which the information is requested.

However, the culture of public administration in Mozambique often runs counter to the well-intentioned dictates of the law. On paper, the Mozambique government has complied with the exigencies of good governance and public responsibility – as mandated by international donors and aid agencies – while little changes in day-to-day practices (Diallo Reference Diallo, Rubbers and Jedlowski2020). The front desk of any public institution – with its complex web of bureaucratic procedures and hierarchical layers – stands between the seeker of information and the archive. At the front desk, any person seeking information is expected to produce a ‘credencial’ (credential) – an introduction letter granted by a ‘credible’ Mozambican institution (public or private). The credencial must explain the reason for requesting information and the purpose for which it is intended. Anyone who has conducted research in Mozambique might have experienced the power that the credencial confers not only to get access to written records, but to talk to people as well. In local administrations at the district and village level, no work can be done without a duly signed and stamped credencial.

While the request of a credencial is in stark violation of the law, it is in perfect consonance with a deeply rooted bureaucratic culture dating back to the Portuguese colonial administration. Despite the muscular discourse of escangalhamento, the independent state of Mozambique under Frelimo maintained – and in many ways reinforced – the heavy-handed bureaucracy in which signed and stamped papers had power over all other means of communication (Souto Reference Souto2003; Alpers Reference Alpers and Hyslop1999). During the socialist experiment, the credencial was often used along with another colonial invention, the ‘guia de marcha’, a document of safe passage. With the increasing militarization of society during the civil war and Frelimo's obsession with social reform and control of rural to urban migration, the guia de marcha became the quintessential document of the socialist government, without which mobility from one district to another and access to public institutions were impossible (Machava Reference Machava2011). Although it has lost most of its power, the guia de marcha continues to be used within the administration. At the district level, where the erosion of the socialist, one-party culture has happened at a much slower pace within public institutions, the guia de marcha is often conflated with the credencial. Here, the level of scrutiny of the document is very tight and it must pass through multiple layers of hierarchy.

Understanding these bureaucratic complexities is essential for getting around the front desk and for accessing the dead archive. Rather than opening doors, evocations of the law or any claims to the right to information at the front desk are most likely to result in frustration. One must navigate the very slow corridors of the bureaucracy with immense patience to access the local archive. It often happens that, after the credencial has been presented at the front desk, the functionary receiving it will seek the signature of someone higher in the hierarchy before authorizing the person holding the keys of the archive to grant access. Although the Law of the Right to Information states that the functionary responsible for the management of files and archives has discretionary authority to authorize access to information, he or she will often take several days or weeks, seeking the consent of someone higher in the hierarchy: the head of human resources, the permanent secretary, the secretary of the administrator, or the administrator him- or herself.

Consent to grant access to information often comes in the form of a written note on top of the credencial – ‘Autorizo’ (I authorize) – with a dated signature, and sometimes a stamp. It is not unusual that the seeker of information is summoned to meet the person entitled to sign the consent note in order to provide further explanations on the reasons for requesting information. In these ‘interviews’, it is the tone of the conversation, not necessarily its content, that determines whether the consent note is signed or not. Who one knows in the area – a prominent political figure, a party member, or a local figure of authority – is often an asset in these interviews. Rather than the reasons for requesting the information, in fact, the ‘interviewer’ is more interested in knowing the seeker of information. In a political environment of multiple loyalties, where state functionaries are expected to be the guardians of the political party in government, knowing who people are is of critical importance for civil servants. The consent note to access the archive often comes out of a sense of trust – trust that the information is not being given to a member of the opposition or anyone unsympathetic to the government.

Given the shortage of personnel in file management, in most local administrations this task is often carried out by a low-ranked functionary with no authority to decide on such matters without authorization from above. The way in which functionaries at the local level interact with information seekers varies across districts and institutions. It also varies if the information seeker is a Mozambican citizen or a foreigner. State functionaries often rank information seekers according to their perceptions of the danger and exposure involved in providing information. Whereas journalists are often viewed with suspicion, due to their profession they are perceived as the least dangerous and most legitimate information seekers. Foreign researchers, who are often Caucasian (including those working for international NGOs), are perceived to be more dangerous than journalists but less dangerous than their Mozambican counterparts. Foreign researchers from renowned institutions and academic circles are regarded as neutral and of little consequence. By contrast, Mozambican researchers are often perceived as the most dangerous information seekers, especially when they fail to establish their political loyalties in favourable terms to local bureaucrats. For example, a Mozambican student who conducted research for his doctoral thesis in two districts of Nampula province in northern Mozambique expressed his disappointment that foreign researchers could more easily access information than he could. For about five weeks he was not granted access to the district archives while a foreign colleague managed to access the same information in a couple of days. The doors opened for the Mozambican student only after an informal breakfast meeting with a senior party and government member in the province.Footnote ³⁰ The request to access the archives of Inharrime district in Inhambane province by one of the authors of this article was only taken seriously when he began referring to senior government officials whom he could call to expedite the process. In these two cases, the archives were only opened after the researchers had established their political alignment by mentioning well-known political figures with whom local functionaries would not want to be out of favour. The other author, who worked in Niassa province, had to keep the topic of his research as vague as possible to be able to access specific information.

The reluctance to grant access to the archives to national researchers may be explained by the fact that a foreign researcher is likely to write and publish in foreign languages and therefore his or her findings, even if they were to circulate in Mozambique, are thought to have little impact. Whereas a Mozambican researcher is thought to be writing and publishing in Portuguese, and her or his work is more likely to reach the national audience and influence public opinion on matters on which only the government or the ruling party should have a final say. This is the price of a culture of gatekeeping that has reached alarming proportions in the post-socialist era in Mozambique.

Aside from these bureaucratic barriers, once the researcher has passed the front desk and has been given access to the archive, a new set of challenges arises. The most important is how to locate the information of one's interest in an uncatalogued mass of papers. Given the unsavoury state of the archive, the researcher is given the keys to the room (often a basement or cottage) and left alone. As well as the time required to go through all the files to locate those of interest to one's project, the researcher may also need protective gear because the files are often stored on shelves along with spare automobile parts and old typewriters that may fall on one's head at any time.

Working in the dead archive, new questions of deontology and copyrights emerge. What reference system can one use for uncatalogued files? Given the threat of destruction that most files face, how can one verify sources in an archive that no longer exists? If researchers take photographs of the materials – as seems to be the accepted approach among most scholars, either for the expediency of the research or due to the imminent destruction of the files – how can one secure copyright? This question is more pertinent for visual materials such as photographs, map sketches and cards.

The fact that local administrations often recycle archived documents adds another layer of caution when dealing with the dead archive. Running out of paper, local functionaries mine the dead archive for papers that can be used for expedient messages. Ignoring the side of the paper covered with written records, they use the other, blank side and give new life to the document. This document is not returned to the original bundle it belongs to, but tossed into another pile of newly discarded materials. Therefore, it is important for the researcher to pay attention to the multiple lifecycles that documents went through in the dead archive, for one document may contain information from different epochs and on different matters. How one references these kinds of files is also a challenge.

Despite these challenges and complicated pathways, the dead archive offers a wealth of opportunities for knowledge production. Like the professional scholars of African history and anthropology who, in the 1960s, worked at the exhilarating time when archives were being opened and new ways of studying African societies were being introduced, the student of contemporary Mozambique has the opportunity to explore a new scholarly frontier. This is as true for Mozambique as it is for much of postcolonial Africa (Ellis Reference Ellis2002). A great deal of what has been written on Mozambique after independence is based on print media, published official documents and interviews. For the most part, the dead archive has remained unexplored, and with it the everyday affairs of public administration, state–society relations, and the mechanics of state making and citizenship. If we want to know how ordinary Mozambicans made meaning out of the independence era and the socialist revolution; the role they played in shaping the new regime; and how, as state functionaries, factory workers or urban residents, they found a vocation as heralds of the revolution, we will have to bury our hands in the dirty entrails of the dead archive. Here, going to waste is the institutional and public memory out of which the fabric of Mozambique's recent past can be dissected. New actors and silenced historical processes can be reinserted in the main narrative about Mozambique's past. A good example of this is a recent publication on the Mozambican civil war by Morier-Genoud et al. (Reference Morier-Genoud, Cahen and Rosário2018).