Archives, Epistemic Injustice and Knowing the Past

ABSTRACT

In this essay, I argue that the destruction or hiding of archives can cause long-lasting epistemic harms and constitute complex ethical challenges. The case of Kenya’s ‘migrated archives’ is argued to be an example of how actions in the past can have long-lasting epistemic consequences and can cause contemporary epistemic injustices and harms related to one’s knowledge of the past. The perpetrators of such harms and injustices are argued to have a backward-looking epistemic responsibility and to be liable to make epistemic amends. The practice of acknowledgement is suggested as one possible way to make effective epistemic amends. I argue that making effective epistemic amends would constitute a step towards addressing epistemic harms and injustices related to our knowledge of the past. However, it is important to remember that this would only constitute one out of many necessary steps in addressing epistemic injustice and that further individual, institutional and ideological changes are necessary.

KEYWORDS:

• Archives
• epistemic injustice
• Kenya
• epistemic amends

Introduction

The focus of this essay is primarily two related phenomena: actions in the past that cause or contribute to contemporary epistemic harms, and epistemic injustices stemming from our knowledge of the past. By examining the historical case of Kenya’s ‘migrated archives’, I demonstrate that actions in the past can cause long-lasting epistemic harms and constitute complex and significant ethical challenges. The essay is situated in relation to two recent developments in the literature on epistemic injustice: the theorising of denial of past wrongdoing as a form of epistemic injustice (Altanian 2020; Oranlı 2020; Song 2020) and the interest in epistemic injustice and archives (Melanson 2020; Wouters 2020). The discussion primarily pertains to archives as a terrain for epistemic harms and injustice, and the issues that arise when archives are censored, hidden, or outright destroyed in efforts to sanitise collective memory. Following Song (2020), I argue that the perpetrators have a backward-looking epistemic responsibility and that some morally responsible agents are liable to make epistemic amends. I conclude by demonstrating that this framework elucidates part of what ethics requires of us in addressing epistemic harms and injustices related to our knowledge of the past. This essay consists of five parts and proceeds in the following manner. In section one an account of what archives are taken to be throughout the essay is given and some of the important roles that archives can fill in our dealings with the past are laid out. Section two is dedicated to the recent scholarly literature on archives and epistemic injustice. In section three my focus turns to instances where archives have been destroyed or hidden in order to mask the truth about the past. By examining the case of Kenya’s ‘migrated archives’, I argue that the actions of the British perpetuated several epistemic injustices and produced significant epistemic harms. However, I would like to caution the reader that the account of the history of the ‘migrated archives’ offered in this paper is limited in scope and primarily focused on the actions of the British. For a fuller and more nuanced account of the history of Kenya’s ‘migrated archives’ I would direct the interested reader towards the extensive historical scholarship on this topic. In section four I turn to the recent literature on denial of past wrongdoing as a form of epistemic injustice to add further nuance to our understanding of the epistemic harms in the case of the ‘migrated archives’. Lastly, in section five I adopt Song’s (2020) account of backward-looking epistemic responsibility to argue that the perpetrators of epistemic harms and injustice in the case of the ‘migrated archives’ have a responsibility to make amends for past epistemic injustices.

Archives

To start off, a clarification of what archives are taken to be throughout this essay is necessary. Following the work of Verne Harris (2002), archives are understood as constructed windows into personal and collective processes rather than a reflection of reality and that archives simultaneously express and are an instrument of prevailing relations of power. The postmodern turn in archival science has shifted the perception of archivists as passive or neutral keepers of records to one where archivists are active participants in the creation, collection, preservation and choice of information provided by their archives (Schwartz and Cook 2002). Archives are thus no longer to be viewed as neutral or impartial. Schwartz and Cook (2002, 7) propose a view of archives as ‘dynamic technologies of rule which create the histories and social realities they ostensibly only describe’, where some events are included, and others are not. In this view archives are not only socially constructed institutions, but arenas for knowledge production about the past and shape our notions of history, memory and identity (Schwartz and Cook 2002; Stoler 2002). Knowledge of the past is thus controlled through archives as certain stories are privileged while other stories are marginalised as archives are set up by the powerful in society, for the purposes of the powerful (Schwartz and Cook 2002). Decisions on inclusions and exclusions of records have significant implications for what will be remembered or forgotten in the future (Wouters 2020). Despite this, archives are often vital for our dealings with the past. Within the paradigms of transitional justice, archives have been argued to be able to fill important roles in the quest for social justice, ranging from documenting human rights violations to assisting in processing accountability, to their role in the establishment of a collective memory (Altanian 2017; Harris 2002; Jimerson 2007; Schwartz and Cook 2002). Scholars such as De Baets (2004) and Nickson (1995) have taken a particular interest in archives of repression. De Baets (2004) suggests that there are moral interests at stake in the archives of repression and argues that former victims or, in their absence, the relatives of victims are entitled to know about the repression they were victims of. Archives of repression may facilitate rehabilitation and mourning processes, and even function as a form of historical and moral justice. Closely related to the moral interests are the legal interests of citizens looking for evidence to substantiate charges against their former oppressors (De Baets 2004). Access to archival material can play a vital role not only to pressing charges but also for exercising individual rights to amnesty, clearing one’s criminal record of offences, compensation, inheritances and the restitution of property (Altanian 2017; De Baets 2004).

Archives are also assigned a central role for the ‘right to know’ in the United Nations Commission on human rights final report on the ‘Question of the impunity of perpetrators of human rights violations’ prepared by Louis Joinet (1997). The right to know is not just a right of any individual victim or their family to know what happened, but also a collective right that draws upon history to prevent human rights violation from recurring in the future. The right to know is combined with a ‘duty to remember’ on the part of the state, to prevent the revision or negation of history and to ensure the remembrance of oppression, as it is part of a people’s national heritage. The preservation of archives detailing human rights violations is one of the measures outlined in the report as central for these purposes (Joinet 1997).

Wilson (2012) argues for the value of archives in supporting indigenous claims of mistreatment, however stresses that archives are not the only factor in addressing past wrongs, but that a strong archival record can be essential in doing so. Hastings (2011) demonstrates that records that once were created for social control or repression can be repurposed for social justice, as such documents can be used to enable recognition of past injustices and used in the movement towards redressing them. Altanian (2017) reminds us that we ought not to forget that human agency is not only necessary when putting the information from archives into narratives but also in the creation of the archives themselves. Archivists can only archive what they receive, and the material they receive generally comes from dominant actors, and in that sense archives mirror society and its institutions. Others, such as Verne Harris (2014) are sceptical of attempts to archive and address human rights abuses, arguing that they often do less good than what one might assume.

Archives and epistemic injustice

The main concern of this essay stems from a recent article by Joe Melanson (2020). He argues that archives have the potential to both transmit and perpetuate epistemic injustices that were committed in the past. Before explicating on Melanson’s arguments in more detail an account of the concept ‘epistemic injustice’ is necessary.

In her seminal book on epistemic injustice (2007), Miranda Fricker shines a light on ethical aspects of two basic epistemic practices: conveying knowledge to others through testimony and making sense of our own social experiences. She distinguishes between two forms of distinctively epistemic injustice: Testimonial injustice and hermeneutical injustice. Testimonial injustice occurs when a hearer gives less credibility to a speaker’s word than they ought to due to an identity prejudice. Hermeneutical injustice occurs prior to testimonial injustice as a result of ‘a gap in collective hermeneutical resources putting someone at an unfair disadvantage when it comes to making sense of or expressing their social experience' (Fricker 2007). Hermeneutical resources are sets of shared concepts that are available for use in understanding oneself and communicating with others. These resources are not socially neutral and often reflect the experiences and interests of privileged people, thus they are often better suited to describe the experiences of the privileged than the experiences of those marginalised (Ivy 2016).¹ Following Fricker’s initial account, a significant body of work has been produced on epistemic injustice and Fricker’s conceptualisation has since been criticised and expanded upon, and new forms of epistemic injustice and oppression have been identified (See for example: Anderson 2012b; Bohman 2012; Coady 2010; Dotson 2012; Hookway 2010; McGlynn 2020; Medina 2017; Pohlhaus 2012; Wanderer 2017).

Melanson’s (2020) argument starts from the notion that archivists face unique ethical challenges when dealing with material that documents human rights abuses and mistreatment of marginalised communities, as these materials both can contain sensitive subject matter and have potential uses that involve archivists in ethically charged situations. Especially as such materials are often used to seek redress for past injustices. In many cases, such archival material is a product of the conceptual frameworks of the perpetrators and the archives thus have the potential of forwarding those narratives, rather than those of the victims. There are plenty of such examples to be found throughout the scholarly literature (Gibbs 2010; Wilson 2012; Hasting 2011; Valderhaug 2011; Halilovich 2014).

In a novel move, Melanson (2020) adopts the notion of epistemic injustice for his exploration of how institutions like archives are involved in social knowledge practices and can unjustly cause epistemic harms, while also highlighting the possibilities for archivists to combat such injustices. He argues that archivists must actively engage with cases of epistemic injustice and, similarly to Harris (2002), Jimerson (2007) and Wouters (2020), that archivists should set aside traditional assumptions about their role and adopt activist approaches. Melanson (2020) argues that archives have the potential to transmit epistemic injustices committed in the past, and that archivists, due to this potential, have a duty to work toward epistemic justice.

Melanson (2020) goes on to demonstrate some of the particular ways in which epistemic injustice can occur in archives, starting with testimonial injustice at the point of records creation. The most detailed examination of a case of testimonial and hermeneutical injustice at the point of record creation can be found in Wouters (2020). Melanson (2020) notes that government documents, especially those created during periods of human rights abuse, often exclude the testimony of the victims, thus largely excluding their voices and suffering from the archival record, while magnifying the government’s interests. Similarly, Hastings (2011) notes that such documents are often produced to justify the government’s actions, to ensure social control, and to reflect the testimony of those who benefit of such policies rather than those marginalised by them. Through Melanson’s (2020) analysis, further examples of how testimonial injustice can manifest in archival material are explicated on. Examples of these include the removal of particular viewpoints from the records or when certain formats of materials are unfairly seen as less credible than others. Archives have traditionally mainly collected written documents, which have been perceived to be more credible than for example oral traditions, which have led to the marginalisation of groups who traditionally rely on oral mediums.

Melanson (2020) identifies a further form of testimonial injustice related to archival work drawing on the work of José Medina (2011). Medina (2011) argues that excess credibility attributed to one party can lead to credibility deficits for other parties. Melanson (2020) argues that this can happen in cases where the evidentiary value of certain forms of documentation, such as official governmental documents, are overvalued. Such an example is found in Hastings (2011) work on Japanese Americans who were sent to internment camps during and in the aftermath of WWII. Scholarly research found that at least $77 million was due in reparations to these Japanese Americans. However, as only taxation documentation was deemed acceptable proof of loss, the actual reparations paid out were much smaller. Finally, Melanson (2020) demonstrates how hermeneutical injustices can be perpetuated through the archives by preserving conceptual frameworks developed under regimes of oppression that are largely unable to articulate the experiences of those oppressed. If these materials are then later unreflexively drawn upon, as in the case analysed by Gibbs (2010), these injustices are transmitted from the past to the present through the archives. Melanson (2020) concludes that there is a significant lack of testimony from marginalised individuals in historical records, and that these records often reflect a society that lacked the conceptual frameworks to recognise and articulate significant injustices experienced by those marginalised. While this already provides a good reason for scepticism when it comes to the value of archives for dealing with past injustices, another significant concern remains, namely the sanitising, hiding or destruction of archives and archival materials.

The ‘migrated archives’

Archives and archival material can be sanitised, hidden, or destroyed for many different reasons. De Baets (2004) identifies a wide range of rationales for why archives detailing repression have been destroyed. He distinguishes those which take place during regimes of repression such as removing the evidence of abuse and their perpetrators and removing evidence of command chains and different mechanisms of repression and surveillance, and those which take place after the end of a repressive regime such as members of a new regime destroying incriminating evidence or to depreciate the remnants of the old regime. Wilson (2012) notes that archives are often targeted when regimes attempt ethnic cleansing in order to facilitate a claim that those who are to be targeted by the cleansing never lived in that particular country or region. Since land records, civil registers and other documents include inconvenient truths that would prove the targeted groups existence in the country or region such documents are destroyed in order to erase the group.

Systematic destruction and hiding of archival material took place as British colonial rule ended in many former British colonies (Anderson 2015). Potentially the foremost of such incidents took place in Kenya as the British destroyed the archives detailing their misconduct in an attempt to edit, sanitise and censor history (Anderson 2015). During their colonial rule in Kenya the British had heavy-handedly put down the Mau Mau rebellion, imprisoning more than 80.000 people without trial, hanging over 1000 ‘terrorists’ and subjecting local people to surveillance and interrogation on an enormous scale (Anderson 2005; Bennett 2011; Elkins 2005). The intention behind the destruction of the documents was to prevent the records of the counter-insurgency falling into the hands of the nationalist government coming into power in Kenya in 1963 (Anderson 2015). Destruction was not the only method adopted to sanitise the archival records as certain documents were to be retained in British possession and secretly returned to the United Kingdom (Anderson 2015). The combination of destroying and hiding away documents detailing British misconduct effectively denied both the victims of repression and other Kenyans access to critical materials related to their history. It also ensured that the post-colonial Kenyan state would never be able to assemble a full record of the actions of the former colonial rulers. This was also hidden from the British public until 2011, when a High Court case in which veterans of the Mau Mau rebellion sought compensation from the British government for their injuries, made national headlines. As part of the investigation, the documents that were retained and hidden by the British, were located and scrutinised for the first time since 1963, revealing abuse and torture (Anderson 2012a). Throughout this process, it became apparent that the British had kept and hidden archives from all over the colonial empire (Badger 2012), sparking a wider debate regarding whether the British had sought to manipulate history by censoring archives, not just in Kenya, but in all of their former colonial territories (Anderson 2015).

The processes for selection, destruction and retention of records had been in place in the British system of government since the 1800s and at present only around 5 per cent of all British government records are selected for retention, and thus around 95 per cent of all records are destroyed or never made accessible (Hampshire 2013). While the processes in the British colonies often varied from this norm, it is still a troublesome realisation that so little of the records are kept and made accessible to the public (Anderson 2015). Specific categories of documents were identified for destruction and retention. Those which were to be destroyed, were often burned, or sunk into the ocean while those which were to be retained were listed, packed and transported to London (Banton 2012). Colonial officials in Kenya went to a great length to keep this process secret, including the exclusion of African staff from offices where documents where categorised and deliberately changing the file system so that the removal of documents could be disguised (Anderson 2015). All documentation relating to the joint intelligence committee, containing political intelligence material and those files graded as ‘Top Secret’ were designated for destruction. The documents to be retained were selected based on four main criteria (Anderson 2015):

1. Those that might embarrass the Government;

2. Those that might embarrass members of the police, military forces, public servants or others such as police informants;

3. Those that might compromise sources of intelligence information

4. Those that might be used unethically by ministers in a successive government

A large number of files were transported to the United Kingdom from Kenya, considerably more than from any other colony. We do not know what was ultimately destroyed in Kenya, as no lists detailing what documents were destroyed appear to have survived (Anderson 2015), which has led the historian Elkins (2005) to call for the creation of new, alternative archives based on indigenous papers and recorded oral testimonies to replace the colonial records that were destroyed.

The ‘migrated archives’ as they are called, were finally revealed and made available to the public in 2011. The process leading up to the reveal was a product of several sets of actors with a shared interest in the recovery of the hidden archives such as the victims of British abuse in Kenya, seeking compensation for their injuries and losses, activists from Kenyan human-rights organisations who campaigned for recognition of the abuses committed during the colonial period, and the historians who uncovered the story of the ‘migrated archives’ (Anderson 2015). Of particular interest among these were a group of Mau Mau veterans seeking compensation for the abuse they had endured as the British cracked down on the Mau Mau insurrection. Despite knowing their own suffering, the authority of their oral testimony was not enough to be able to move forward with a legal case (Elkins 2005). For them, the archive possessed a different kind of power as it represented ‘independent’ evidence supporting their claims (Anderson 2015). The Mau Mau veterans were thus victims of at least one form of epistemic injustice, namely testimonial injustice, as their testimony was not given its due credibility until it could be substantiated with evidence produced by the British themselves. Not only did the British deny, and go out of their way to hide the evidence of the extent of their abuse, but they also long after the fact denied the existence of the ‘migrated archives’. One such example took place as recently as 2006 when legal representatives from the London law firm Leigh Day made a general request for documents relating to the suppression of the Mau Mau rebellion to the Foreign and Commonwealth Office (FCO). The reply they received stated that all information held by the FCO regarding that particular period had been transferred to the National Archives and made public, which was far from the truth (Anderson 2015). In the subsequent five years more requests were made, both by Leigh Day and from historians seeking to use the British Freedom of Information legislation to gain access to the ‘migrated archives’. In response to all of these, the government claimed that searches had been undertaken but that no further documentation existed (Anderson 2015), which was proven demonstrably false with the revelation of the ‘migrated archives’ in 2011. The British government was hiding away its secrets, denying the people of Kenya and other former colonies full knowledge of their history.

Today the ‘migrated archives’ are available to the public at the National Archives. Anderson (2015) suggests that it is very likely that the Kenyan government will request for them to be repatriated in the future. In 2013 the British government reached a settlement with more than 5000 Kenyan citizens who claimed to have been victims of abuse and torture in the 1950s. A decision that was undoubtedly reached in part due to the documents found in the migrated archives (Anderson 2015). However, that will not make up for the lost knowledge of about these past injustices wrought by the destruction of the archives.

Even though decisions of what to retain and what to destroy are common practice when it comes to archives it does not hold up as an ethical justification for what the British did throughout their colonies. The explicit aim of the destruction and stealing of the documentation was to maintain the good reputation of Britain and anything that could reflect poorly on the United Kingdom and its colonial administrations was to be destroyed or to be kept in secrecy (Anderson 2015). This explicit aim to sanitise history and deny past injustices committed during colonial rule was acted upon and then later denied. While the story from Kenya has garnered significant attention and is highly problematic there are instances where recourse is impossible as a result of all records having been destroyed. For example, in the former British colony then known as Aden, the civil servants did not have time to go through their papers to determine what was to be kept and was to be destroyed due to the insurrection in 1967. Instead, every document in the Governor’s Office was burned on a huge bonfire as the Governor and his staff escaped the country (Anderson 2015).

This is a very practical example of what Walter Mignolo (2002) calls the geopolitics of knowledge. It is very much a case of a colonial power making decisions on what is and what is not to be known with dire consequences, particularly for the colonised. The actions of the British officials are also in direct conflict with the right to know as set out by the United Nations commission on human rights. By destroying and secretly hiding documentation of the human rights violation perpetuated in response to the Mau Mau insurrection, the British violated the right to the truth by obscuring the truth about past events. They failed the duty to remember by not preserving documentation necessary to create and preserve a collective memory that accurately reflects the repression that took place. Rather, in direct conflict with the right to know the British themselves engaged in revisionism. As the British destroyed the documentation, they denied victims and their family’s knowledge about the circumstances and extent of the violations that took place.

This situation is rife with epistemic harms. Epistemic harms can be distinguished into two types: (A) Denial of the status of the subject, and (B) negative epistemic and practical consequences (Song 2020). The former is caused by unjust exclusions, such as the exclusion of marginalised individuals and group from knowledge production processes. The exclusion results in a process of objectification where those excluded are viewed as lacking the full capacity to reason, which Fricker (2007) takes to be essential to human value. Those who are unjustly excluded are thus denied part of their humanity. While a closer investigation of epistemic harm and epistemic injustice in the aftermath of the Mau Mau rebellion than possible within the scope of this article would be necessary to provide a clear picture of the extent and gravity of such harms, it is possible to trace the contours of such exclusion in the treatment of the Mau Mau veterans. As suggested earlier, one such example is the treatment of the Mau Mau veterans who sought compensation for the abuse they endured. They were the victims of testimonial injustice as their oral testimonies were not assigned due credibility and thus dismissed, leading to their case not being brought to court until other evidence was revealed. Particularly troubling is the fact that the evidence from the ‘migrated archives’ was considered to be ‘independent’ and thus assigned more credibility than the oral testimonies of the victims, despite the vast evidence of destruction and sanitisation of the archival material.

That the written accounts were assigned more credibility than oral accounts reflect the Eurocentric epistemic hegemony that generally assigns greater credibility and value to written accounts then to oral testimony in the recording of history.² This point has forcefully been argued by, among others, Boaventura De Sousa Santos (2018). He argues that if one considers the whole world as a landscape of written and oral knowledges, one realises that oral knowledges are more common than written knowledges, although seen as less prestigious for the simple reason that the criteria for ascribing prestige are established in contexts where written knowledges are dominant. Similar to the insights of the archival scholars that were discussed earlier in this paper, Santos (2018) argues that written text itself often functions as the official transcript of the elites. During colonisation the written text was often used to silence the oral texts of the colonised, and at the same time the colonised often turned to the oral to resist colonial domination. In his works, Santos (2018) highlights examples of projects where oral history has been employed as a tool of resistance, both against capitalist and colonial domination.

There are plenty of examples of the second form of epistemic harm in the given example. As the British masked the truth, they by necessity produced negative epistemic effects such as false knowledge and collective ignorance, making it harder, if not impossible, to establish a truthful collective memory about the past. While the loss of knowledge epistemically harms all people, it produces more pernicious harms for the victims and their families, which becomes apparent for as the Mau Mau veterans were denied compensation. Furthermore, as argued by Song (2020) those who face denial, for example victims and their kind, bear extra practical costs when they try to express their experiences as they might suffer from undue distrust or would be considered bad informants unjustly. This elucidates the point made in the first section that a strong archival record is essential in addressing past abuses. As the British destroyed, hid and denied the existence of much of the evidence of the repression in Kenya, they undoubtedly made it harder, if not impossible, to address the full extent of the abuse.

This brings us to the second of the two recent developments in the literature on epistemic injustice that this essay engages with, namely the theorisation of denial of past wrongdoing as epistemic injustice. I take this development to primarily consist of three articles published in the second half of 2020, two papers by Altanian (2020) and Oranlı (2020) on the denial of the Armenian genocide and the paper by Song (2020) on Japan’s denial of military sexual slavery. The cases discussed by Altanian (2020), Oranli (2020) and Song (2020) carry many similarities to the situation discussed in this section. Song (2020) offers an account of a backward-looking epistemic responsibility that elucidates the possibilities of making epistemic amends for past epistemic injustices and harms.

Denial of past wrongdoing as epistemic injustice

Altanian (2020) argues that genocide denial is an injustice perpetuated not only against the direct victims and survivors of genocide, but also against future members of the victim group. She argues that in cases of persistent and systematic denial, or denialism, a type of epistemic injustice is perpetuated: testimonial oppression. Altanian (2020, 2) defines testimonial oppression as unjustified epistemic coercion through institutional practices that brings about both epistemic and other harms insofar as the testifier occupies a specific epistemically and ethically relevant status, the status of a knower. Following Fricker (2007) she takes being a knower to be a capacity of essential human value and that testifying is a part of that capacity. Similar to the distinction made in the previous section she takes the primary harm of testimonial oppression to lie in being disrespected in one’s status as a full human subject on the basis of institutional or structural discrimination but adds that one thereby is being stymied in one’s ability to exercise one’s epistemic agency. Furthermore, she adds that in cases of testimonial oppression this harm is perpetuated against a particular social group for the benefit of another social group. Testimonial oppression for Altanian (2020) refers to conditions and practices that systematically deprive speakers of recognitional epistemic goods. These recognitional epistemic goods consist of the speaker being afforded the right response to their epistemic agency and status as a knower when giving testimony. On Dotson’s account (2011), testimonial oppression can take many forms, one of which is ‘testimonial smothering’ which occurs when a speaker perceives their audience as unwilling or unable to gain the appropriate uptake of the proffered testimony. Dotson identifies three circumstances that routinely gives rise to testimonial smothering (A) circumstances when the content of the testimony is unsafe and risky, (B) circumstances where the audience demonstrates testimonial incompetence to the speaker in regard to the content of the testimony, and (C) in circumstances where testimonial incompetence follows from, or appears to follow from, pernicious ignorance. In such circumstances, it is likely that the speakers will be coerced to silence themselves, thus becoming testimonially smothered (Dotson 2011). Altanian (2020) argues that practices of silencing can be observed in cases of genocide denialism, as genocide denialism generates pernicious situated ignorance that cultivates testimonial incompetence specifically on the subject matter of genocide, making testimony on genocide unsafe and risky and thus creating the conditions for testimonial smothering.

Altanian (2020) takes denialism to refer to a larger, orchestrated ideological, political or cultural pattern of denial. By genocide denialism she refers to a historically rooted practices that justifies, authorises and imposes pernicious sets of beliefs and understandings by maintaining historically entrenched prejudices against former victims’ groups, mispresenting social relations, distorting historical facts as well as distorting social, moral and epistemic norms for inquiry. Altanian’s account (2020) of denialism elucidates some particularly interesting epistemological insights. The denialism she examines is not just the telling of known falsehoods or withholding of the truth with the intention to deceive and mislead others. Denialism in Altanian’s account (2020) refers to systematic efforts to distort evidence and create a new, wholly opposite understanding consisting of not only a lack of true beliefs but also the establishment and imposition of an alternative, distorted view of reality.

Conditions of genocide denialism legitimise distrust and scepticism towards testimony on genocide, thus systematically depriving testifiers of epistemic recognition. In her analysis of Turkish genocide denialism Altanian (2020) identifies two such mechanisms for epistemic misrecognition. Genocide denialism discredits the victim group as a whole by attributing negative characteristics to the group, for example through portraying the victim group as suggestible and supposedly having adopted inadequate hermeneutical resources to interpret their past, thus misrecognising their status as knowers. Genocide denialism also entails factual and normative distortions of the subject matter of genocide. Epistemic contributions are rejected insofar as they threaten the dominant understanding of the genocide. Thus, genocide denialism not only misrecognises social identities but also misrecognises and targets particular information, which makes it likely that all members of society are coerced to silence themselves on the genocide (Altanian 2020). The two are also argued to be mutually reinforcing as under conditions of pervasive ignorance about genocide, such as those in cases of genocide denialism, the testimony of those marginalised will appear implausible to testimonially incompetent hearers, which then feeds further distrust (Altanian 2020).

Similarly, Oranli (2020) argues that the Turkish denial of the Armenian genocide constitutes and produces pervasive and systematic epistemic harms. In an earlier paper, Oranli (2018) demonstrated that the denial of the Armenian genocide involves several forms of epistemic injustice such as testimonial injustice, hermeneutical injustice and the unjust distribution of epistemic resources. In a more recent paper, Oranlı (2020) draws on the work of Fricker (2007), Medina (2013) and Mills (2017) to demonstrate the interrelatedness of individuals who perpetrate genocide denialism, the ideology behind denialism and the institutions supporting genocide denialism. Oranlı (2020) suggests that to fully understand such an epistemically unjust regime one must trace both the individual, the institutional and the ideological. She concludes her argument by arguing that the responsibility for genocide recognition thus belongs not only to governments, but also to individuals themselves, as if they do not actively resist, they become actively ignorant deniers themselves.

Troublesome similarities between the cases discussed in previous sections and those elaborated upon by Altanian (2020) and Oranlı (2020) arise. In Altanian’s (2020) definition of genocide denialism we find the maintaining of entrenched prejudices against former victim groups and distorting historical facts, both of which are topics that have been touched upon previously in this article. A core concern highlighted previously was the potential of re-perpetuating epistemic injustices of the past by drawing upon archival material that reflects the narratives of the perpetrators, and thus forwarding those narratives rather than those of the victims. We also find Altanian (2020) concerned with the distortion of historical facts, which was the primary concern in the discussion of the epistemic harms of the destruction of archival materials in the last section. Particularly concerning is the relationship between destroying or hiding evidence of wrongdoing and denial. The British did so to hide the extent of abuse to protect their reputation despite extensive wrongdoing. One doesn’t need to look long to find examples from other contexts where extensive destruction of archival material took place and later denial of the gravity or extent of past wrongdoing emerged. For example, there is an extensive scholarly literature on the destruction of archives at the end of the Apartheid regime in South Africa (see for example: Harris 2002; Wouters 2020), and former president FW de Klerk denied that Apartheid was a crime against humanity as late as in 2020.³ Altanian (2017) argues in an earlier paper that strong archival records are important in order to settle controversies about the proper interpretation of the past but in contexts of distrusts and prejudice such as in contexts of genocide denialism, even strong archival records might be perceived as unreliable and lacking evidentiary status.

Song (2020) offers an account of a backward-looking epistemic responsibility which suggests that some morally responsible agents who committed epistemic injustices are liable to make epistemic amends for past epistemic injustices. Song’s (2020) primary interest is Japan’s history of military sexual slavery, and the subsequent denial of that history. Based on particular agents’ moral responsibility, Song (2020) argues that some moral agents who deny Japan’s military sexual slavery are liable to make epistemic amends for past epistemic injustice. Song (2020) defines denial quite generally as any attempt that deliberately minimises, trivialises, justifies, or negates well-established facts and accounts about an injustice and points out that all denial doesn’t involve epistemic injustices. She identifies two forms of denial, individual-led denial and state-led denial, and demonstrates that these may work in tandem, for example through the state-led form enabling the individual form.

Backward-looking epistemic responsibility as the way forward

Song (2020) distinguishes between backward-looking and forward-looking justifications of responsibility. Backward-looking justifications are focused on restoring, mending or repairing a certain status ex ante injustice, while forward-looking justifications are focused on how to make the world better from now on. These different justifications translate into different notions of responsibility. Backward-looking responsibility include notions such as accountability, blameworthiness and liability, while forward-looking responsibility refers to obligations going forward or virtues (Van de Poel 2011).

The literature on epistemic responsibility is primarily forward-looking (See for example Fricker 2007; Medina 2013). However, Ben Almassi (2018) argues for a remedial epistemic responsibility shared among perpetrators, victims and a third party to redress past epistemic injustices. Song’s (2020) account contrasts with Almassi’s (2018) in that she develops an account of a backward-looking epistemic responsibility that only some agents bear. Song (2020) argues that backward-looking epistemic responsibility is important to amend past epistemic injustices, and thus move closer to genuine epistemic justice. The discussion in this article adds further strength to Song’s (2020) argument as it elucidates how epistemic injustices of the past can be brought from the past to the present through archives, thus reinforcing the urgency to address them. Furthermore, Song (2020) argues that epistemic amends should come from specific moral agents for them to have a genuine restorative effect. To identify who is the bearer of this kind of epistemic responsibility, Song (2020) distinguishes between liable and non-liable agents. She outlines three independently necessary conditions for discerning who is liable to make epistemic amends: (I) causality, (II) autonomy, and (III) epistemic competence. The first condition, causality, posits that for an agent to be liable, he must directly cause epistemic harms. The second condition requires a moral agent to be autonomous. Song (2020) argues that an agent is epistemically autonomous if he is in a position to widen his perspective despite having adopted a certain worldview. The third condition stipulate that the agent must be epistemically competent. Epistemic competence for Song (2020) entails that the moral agent is reasonably expected to have access to epistemic resources, and crucial evidence for epistemic judgement, as these are taken to be necessary for the agent’s capacity to make critical epistemic judgements.

Song (2020) suggests that acknowledgement is one possible way to make effective epistemic amends. Acknowledgement for Song (2020) is a process where the responsible agents admit their wrongdoing as something that should not have occurred, admit their responsibility, establish a thorough recognition of what the wrongdoing entailed, and commit to concrete change to prevent such wrongdoings from happening in the future. Acknowledgement would thus consist of three steps: a responsible agent’s thorough recognition of (I) what the injustice entailed for the victims, (II) that the injustice should not have occurred, followed by a (III) genuine commitment to not repeat the injustice. Acknowledgment in the cases of denial as an epistemic injustice on Song’s (2020) account entails admitting the victims’ experiences as the historical truth by recanting past denials, paying appropriate credibility to victims’ testimonies, and committing to no longer contribute to, and sustain false-knowledge and collective ignorance. Such acknowledgment may ‘unmask the truth’ and repair the bad-informant status that the victims were subjected to (Song 2020). Acknowledgement would thus constitute an important step to repair epistemic harms. However, Song (2020) notes that such acknowledgment can only repair certain types of epistemic harms and that other complementing accounts of backward-looking epistemic responsibility based on culpability and blameworthiness should be developed (Song 2020).

Similarly, Teresa Godwin Phelps (2002) has argued for the importance of acknowledgement in her work on truth commissions. She argues that the victims of oppression and torture experience a loss of their ‘story’, that oppressive regimes silences and fragments the oppressed, and through that silencing creates its own, false ‘master narrative’. It isn’t until the truth is revealed and acknowledged that this false narrative can be destroyed and rewritten. Thus, Godwin Phelps (2002) argues that for just retribution to take place, victims of oppression need an opportunity to tell their stories and reshape their own experiences in their own words, and that the state actively hears and acknowledges the story, thus recognising the worth of the victim.

This account of epistemic amends elucidates a starting point for what is ethically required of those who have perpetuated epistemic injustice in the past. It sheds light on what is ethically required of those who have perpetuated the epistemic injustices identified in Melanson’s (2020) work on epistemic injustice in archives, and with some qualifications it offers us a framework for thinking through what those responsible for destroying and hiding the colonial archives in Kenya can do to make epistemic amends for their actions.

While several types of epistemic injustice, such as testimonial injustice and an unjust distribution of epistemic resources, are apparent in the case of the ‘migrated archives’ I remain relatively agnostic at this point in relation to any further perpetuation of epistemic injustice. It seems very likely that further forms of epistemic injustice and oppression, such as testimonial smothering, trace their origins to, or have been reinforced, by the case of the ‘migrated archives’, especially considering the similarities with the cases of denial of past wrongdoing as an epistemic injustice discussed by Altanian (2020), Oranlı (2020) and Song (2020). However, as Song notes, not all cases of denial constitute epistemic injustice and the Kenyan case undoubtedly deserves a more thorough examination in that regard than what is possible within the scope of this article. All the same, the destruction and hiding of the archives and the subsequent denial produced significant epistemic harms. Further to this, those responsible for the destruction of the archives, and the later denial, are responsible to make amends, particularly so considering the unjust conditions of the colonial context that these events took place in.

Song’s (2020) framework maps onto the ‘migrated archives’ case quite well. While evidence demonstrating who took crucial decisions and how they were implemented has been found in the retained files (Anderson 2015), potentially making it possible to identify individual perpetrators that hold an individual backward-looking epistemic responsibility, I will focus on the responsibility of the British government to demonstrate the usefulness and action guiding potential of Song’s (2020) framework. I assume that the British government fulfils the necessary conditions of liability as it seems reasonable to suggest that the British government caused the epistemic harms, is autonomous and is epistemically competent. The British government thus has a backward-looking epistemic responsibility and that acknowledgement on the part of the British government would be one possible way to make effective epistemic amends. This would entail a thorough recognition of what the injustice entailed for the victims, for example having to face British denial of the extent of past abuse and repression, being denied reparations, and a loss of possible knowledge about the past. The British government would also have to recognise that the injustice should never have occurred and make a genuine commitment to not repeat such injustices. As Song (2020) suggests, this would also entail the recognising the victims’ experiences as historical truth by recanting past denials, paying appropriate credibility to victim’s testimonies, and committing to not perpetuate false-knowledge and collective ignorance. Further, Song (2020) points to measures against the perpetuation of false-knowledge and ignorance such as including the histories of wrongdoing in history textbooks and in museums which appear apt for the case of the ‘migrated archives’ too.

Unmasking the truth about past abuses that the British worked hard to mask and making the truth part of collective memory would go some way towards addressing the epistemic harms incurred upon the victims. Doing so would entail having to rely extensively on non-archival sources and taking the oral testimonies of the victims seriously. This would necessitate a move away from the traditional Eurocentric primacy assigned to written accounts over oral accounts in the recording of history, to centre the experiences of the victims. Further, this becomes a necessity due to the extensive destruction of archival materials that took place, and even more so with the concerns of epistemic injustices and oppression being perpetuated within and through the use of archives. However, as Song (2020) warns this would only constitute a small step on the road towards epistemic justice. Even if the British government made epistemic amends in the manner suggested in this article, it would only constitute one small piece of the enormous puzzle that is addressing the injustices of the colonial past.

Conclusion

In this essay, I demonstrate that epistemic injustices and epistemic harms can be perpetuated through archives. Specifically, I argue that the destruction or hiding of archives can cause long-lasting epistemic harms and constitute complex and significant ethical challenges. The case of the ‘migrated archives’ is a clear example of how actions in the past can have long-lasting epistemic consequences and can cause contemporary epistemic injustices and harms related to our knowledge of the past. The perpetrators of such harms and injustices are argued to have a backward-looking epistemic responsibility and are liable to make epistemic amends. Following Song (2020) I propose acknowledgement as one possible way to make effective epistemic amends. In doing so I am suggesting a first step towards addressing epistemic harms and injustices related to our knowledge of the past. However, it is important to remember that this would only be one out of many necessary steps in addressing epistemic injustice and that further individual, institutional and ideological changes are necessary.

Acknowledgements

I thank Christopher Day, Eleonora d’Annibale, Rosalie Waelen, Francesca Sharman and two anonymous reviewers for their support and thorough feedback.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Notes on contributors

Karl Landström

Karl Landström is a PhD student at the Centre for Trust, Peace and Social Relations at Coventry University. His research focuses on research ethics, social epistemology and philosophy of social science, with a particular interest in epistemic oppression, epistemic injustice and post- and decolonial theory.

Notes

1 Originally published as Rachel McKinnon.

2 Thank you to an anonymous referee for raising this point.

3 As reported in BBC (2020, February 18). FW de Klerk and the South African row over apartheid and crimes against humanity. BBC News. Retrieved from. https://www.bbc.co.uk/news/world-africa-51532829 4.35pm, January 4th, 2021.