Knowing your past: Trauma, stress, and mnemonic epistemic injustice

Abstract

There is strong psychological evidence suggesting that sometimes social and institutional structures cause people to experience, or exacerbate existing, trauma and severe stress. Evidence further suggests that trauma and stress can lead autobiographical memories to become disorganized and distorted. In this way, social and institutional structures can cause significant harm by denying some individuals access to a specific kind of knowledge; knowledge about their personal past. When people are being denied access to this kind of knowledge, their objective, basic interests in good epistemic agency, capacity for autonomy, and general well-being, are curtailed. In this paper, we argue that these memory distortions therefore constitute a distinctive form of mnemonic epistemic injustice: some people are unjustly disadvantaged as epistemic agents by being avoidably and foreseeably denied access to epistemic goods required to support their objective interests, due to social and institutional structures that cause some of their memories to become distorted or disorganized.¹ They are denied something that they are entitled to, that is, freedom from the imposition of stress and trauma that brings epistemic costs and other damage to their objective needs.

Moreover, this injustice can be further compounded in cases where trauma and stress make it harder for an individual to be believed because their testimony contains untruths due to memory errors. Memories that are distorted and disorganized often exist alongside core memories about important events that are accurate. When it is assumed that certain core aspects of a person's account of their own experiences are false there can be an additional epistemic injustice that compounds the initial injustice of having one's memories distorted. The compound epistemic injustice that we describe can be experienced by people who are speaking untruths (due to memory errors), as a response to the untruths that they are speaking. It can even happen in cases where the hearer responds reasonably to the untruths of the speaker when denying their account credibility. These compound epistemic injustices therefore differ significantly from standard cases of testimonial injustice where a person has their testimony dismissed as lacking credibility when there is little or no good reason to believe that they are speaking untruths (Fricker, 2007). The compound injustices that we describe are interesting because they can be jointly caused by two or more different unjust features of social and institutional structures, that is, those features that cause the memory distortions and those that lead the memory errors to be misinterpreted. They show how different features of social and institutional structures can conspire to make it especially difficult for marginalized individuals to be believed.

We illustrate these points via the case study of asylum seekers in the United Kingdom. Asylum seekers experience high levels of trauma and stress relative to the general population not only because of events that they have often experienced in their country of origin, but also because of hardships that they encounter in the process of seeking refuge and claiming asylum. The heightened vulnerability to undergoing memory distortion and disorganization is the avoidable and foreseeable consequence of deliberate policies, such as, for example, those aiming at deterring asylum seekers from entering the UK. Moreover, this initial injustice can be compounded when asylum seekers must articulate their need for protection in the asylum process.

The paper makes a significant contribution to the literature on epistemic injustice by explaining how social and institutional structures can cause serious epistemic harms and wrongdoing by negatively impacting an individual's memory. The paper also shows how this epistemic injustice can be compounded when an individual provides testimony that contains falsities, even in cases where the hearer responds reasonably to the falsities—introducing the notion of a compounding epistemic injustice. At the same time, it highlights a significant aspect of the plight of asylum seekers.

In Section 2, we begin by showing how social and institutional structures can lead people to experience memory distortions, with specific attention paid to the UK asylum system. Next, in Section 3, we argue that these memory distortions are cases of epistemic injustice. We then proceed in Section 4 to show how these initial injustices can be compounded when individuals are required to provide an account of their needs. Finally, in the conclusion, we highlight implications of this discussion for how epistemic injustice in general should be conceived.

Let us begin, then, by considering evidence suggesting that social and institutional structures can lead people to experience memory distortions. We will begin as we intend to go on, by considering the experiences of asylum seekers as exemplars of this broader social phenomenon.

Now that we have reason for thinking that social and institutional structures like the asylum system can cause memory disorganization and distortion, let us consider how this can constitute epistemic injustice. Epistemic injustice occurs when a person is wrongfully harmed in their capacity as an epistemic agent (Fricker, 2007). An epistemic harm can be wrongful for a number of reasons (Fibieger Byskow, 2020). One distinctive class of wrongs is when an agent is “ingenuously downgraded and/or disadvantaged in respect of their status as an epistemic subject” (Fricker, 2017: 53). This is sometimes referred to as discriminatory epistemic injustice (Fibieger Byskow, 2020; Fricker, 2017). Our claim is that in cases where people undergo stress and trauma due to social and institutional structures in ways that negatively impact their ability to remember the past, they are disadvantaged with respect to their status as an epistemic subject because of the epistemic harms that they experience. They are disadvantaged with respect to their ability to remember aspects of their past in ways that can be crucial to supporting further non-epistemic objective needs. In the process, they are denied something to which they are entitled: freedom from the avoidable imposition of trauma and severe stress that distorts their memory and prevents them from achieving objective needs like capacity for autonomy and well-being. This, we argue, is a mnemonic epistemic injustice. While no one is entitled to accurate memories of their personal past, people are entitled to not have our memories knowingly distorted, given the serious harms this can cause. In addition, we also suggest that there are similarities between the phenomena found in our memory cases and distributive epistemic injustice when institutional and social structures create inequalities in opportunities to access autobiographical memories.

In Section 3, we have argued that when people experience memory distortion and disorganization due to trauma that results from social and institutional structures, this constitutes epistemic injustice. The goal of the current section is to show how this injustice can be compounded when individuals are required to provide an account of their needs.

The first thing that is crucial to note is that where people undergo trauma and stress, the memory distortions that they experience tend to be localized. As we saw in Section 2, under conditions of trauma and stress, memories can present differently from standard autobiographical memory. People may experience memories that are disorganized, where specific events become disjointed from their context and contextual information (e.g., about what happened immediately before or afterward) can become confused (Ehlers & Clark, 2000; Sachschal et al., 2019), and this can have a significant impact on their epistemic agency, autonomy, and well-being. Nonetheless, people who have experienced trauma tend to be able to vividly and accurately remember the core aspects of traumatic events (Herlihy et al., 2012; McNally, 2005). The localized nature of the memory distortions means that it will often be a mistake to generalize from observations of memory disorganization that occurs due to experiences of trauma and stress to the conclusion that the person who is displaying the memory errors is unable to provide an accurate account of core aspects of traumatic experiences that evidence their need for help or protection. Take, for example, an asylum seeker who has experienced traumatic events that led them to flee. They may provide an account of their traumatic experiences which is in some ways disorganized, and may contain some contextual errors, but they are unlikely to misremember the nature of the threat that led them to seek asylum, or its severity. These will be core details about their past, so likely to be remembered accurately and vividly.

What this suggests is that where people who have undergone trauma provide testimony, for example, about past experiences that evidence their need for future help or protection, it is likely that some details of their accounts—including core details of the most important events that are remembered—will be dismissed although they are true. Those responding in this way to evidence of errors in memory would be acting in a way fitting with psychological findings suggesting that people respond to evidence of even minor or peripheral errors present in testimony by discounting the whole of the testimony (Borckardt et al., 2003). In the case of asylum seekers, when they are required to provide an account of their past experiences to evidence their need for asylum, core details of their account of their need for protection may be distrusted although they are true. There is empirical evidence demonstrating precisely this effect: that is, that asylum seekers are treated as generally lacking credibility on the basis of even minor or peripheral factual errors. Often they are treated as lacking credibility because there are inconsistencies in their stories, but untrue claims can also lead them to be dismissed as lacking credibility (Amnesty International, 2013; Asylum Aid, 2011). In a report to the Home Office, the UNHCR (2006) noted that it had “observed a large number of cases where one statement deemed by the case worker to be untrue […] is relied upon to dismiss the credibility of the entire claim” (UNHCR, 2006: 9).

To capture this type of situation, we borrow the term “credibility deficit” from Fricker's (2007) account of testimonial injustice. Fricker describes how prejudice on the part of a hearer can lead a speaker to be given less credibility than they deserve when they are attempting to convey knowledge. Fricker says, and we agree, that credibility deficits can occur where a person displays signs that may under other circumstances indicate a lack of credibility. In the types of case that we are concerned with here people suffer credibility deficits in relation to specific parts of their testimony—those core aspects of their testimony demonstrating their need for help and protection that are true—because of evidence of errors in other aspects of their memory (see Puddifoot, 2020, 2021a for examples of a similar effect in people who have not experienced trauma).

At this point we have argued that individuals who have undergone trauma and stress due to social and institutional structures sometimes experience mnemonic epistemic injustice. We have argued that where people undergo memory distortion and disorganization due to trauma or stress, but they can nonetheless provide an accurate account of core details of their experiences, those core details can be given less credibility than they deserve—a credibility deficit can occur. Now we aim to show that the initial epistemic injustices are compounded by a later epistemic injustice when the credibility deficit happens. In other words, we need to show that the credibility deficit should be understood to be an epistemic injustice.10

Why, then, should credibility deficits experienced by individuals in response to their memory errors be judged to be epistemic injustices that compound the initial epistemic injustice they have undergone? This depends on the type of case that is being considered. We can return to the asylum case to see why.

Take a case where an asylum seeker is providing their account of their past experiences in their country of origin, with the aim of evidencing their need for protection in the country to which they have arrived. They meet with an asylum case worker who is tasked with making an evaluation of the strength of their asylum claim. The case worker harbors prejudicial stereotypes associating the asylum seeker with untrustworthiness or unreliability. This may be the “bogus asylum seeker” stereotype, or a stereotype relating to some other aspect of the asylum seeker's social identity (e.g., their religious identity), or an intersectional stereotype relating to multiple aspects of the asylum seeker's social identity (e.g., Muslim and LGBTQI+ asylum seekers). Psychological research strongly suggests that the presence of stereotypes like these increases the chance that falsities contained in the accounts of asylum seekers will be attended to and remembered (Puddifoot, 2017a, 2017b, 2021a; Bodenhausen, 1988; Cohen, 1981; Levinson, 2007; Signorella & Liben, 1984; Stangor, 1988), and that the falsities will be incorrectly attributed to a wider disposition of the asylum seeker to untrustworthiness (Puddifoot, 2017a, 2017b, 2021a; Duncan, 1976). Let us stipulate, as seems plausible given these empirical results, that in our example the case worker focuses on the errors contained in the testimony, incorrectly assuming that these are an indicator of a broader unreliability, due to the prejudicial stereotypes that they harbor. Focusing in this way on these specific false details of the asylum seeker's account leads the case worker to dismiss core aspects of the testimony that are accurate.

This case is in important respects like Fricker's (2007) prototypical case of testimonial injustice, in which a person presents accurate information, but it is not treated as credible due to the hearer's prejudice. In our case, a person also presents accurate information—their accurate account of the core details about their past experiences—but it is not treated as credible due to the prejudicial stereotypes harbored by the hearer. We would argue that the similarities between our target cases and Fricker's merit our cases being treated as an example of compound epistemic injustice, where an initial injustice is compounded by a further epistemic injustice.

It is important, however, to note that there is one significant difference between our cases and prototypical cases of testimonial injustice. In the prototypical cases, the prejudice suffices to lead to the dismissal of the testimony. A person could say something completely true, in a credible manner, but nonetheless have their testimony dismissed due to prejudice relating to their social identity. In contrast, in the cases we are discussing here the prejudice does not obviously suffice for the credibility deficit. The case worker might not have dismissed the core parts of the asylum seeker's story if there were not aspects of their testimony that were false.

There are two directions that one might take once this difference is acknowledged. One might argue that wherever there is evidence that there are falsities contained within testimony, there cannot be epistemic injustice, even if a hearer makes a prejudiced response to the falsities. Alternatively, one might argue that the notion of epistemic injustice ought to be broadened to include cases where speakers provide testimony that contains untruths and hearers have prejudiced responses to these untruths. There are good reasons for adopting the second of these options. It seems highly undesirable to adopt a position (like the first option) according to which a thinker who responds in a prejudiced way to minor or peripheral errors in a person's testimony, assuming that other aspects of the testimony are false partly due to negative stereotypes, must be just, or neutral with regards to justice, but cannot be unjust. To see this point, simply imagine that you are an older person who tends to momentarily misremember the names of your grandchildren. If someone were to respond to evidence of this localized error by dismissing other parts of your testimony, because of a prejudicial belief that older people are forgetful, they would not be doing something epistemically unjust according to the first option. However, it seems that the notion of epistemic injustice should be able to capture cases of this type. Therefore, it would be better to adopt an expansive notion of epistemic injustice according to which there can be epistemic injustices when hearers respond with prejudice to falsities contained in a speaker's testimony (i.e., option two).

Based on the similarities between the credibility deficits we are describing here and prototypical cases of testimonial injustice, and on this claim that the main difference between our target credibility deficits and prototypical cases should not preclude our cases from being classified as cases of epistemic injustice, we conclude that where a hearer's prejudices lead them to dismiss core and accurate parts of a speaker's testimony in response to trauma-induced memory errors there can be epistemic injustice. This form of epistemic injustice can be suffered by anyone affected by trauma induced memory distortions, whether these were caused by social institutions or bad luck. But most importantly for the purposes of this paper this epistemic injustice can also compound an initial mnemonic epistemic injustice.

We now have one argument in support of the claim that the initial mnemonic epistemic injustice described in Section 3 is sometimes compounded when a person is required to provide an account of their future needs based on their past experiences. But we mentioned earlier that there are different ways that credibility deficits which happen after a speaker has undergone memory errors due to trauma can be epistemic injustices. Let us now consider a second way.

The first type of case, where a person makes a prejudiced response to memory errors, is an interpersonal epistemic injustice: one person makes an unjustly prejudiced assessment of the credibility of certain parts of another person's testimony. However, there can be cases where a speaker suffers a credibility deficit not because of prejudice on the part of the hearer but instead because of a failure of institutional practices, policies, and procedures. Here, we argue, there is an institutional epistemic injustice that compounds the initial distributive epistemic injustice (cf. Blomfield 2021).

Returning to the asylum case once again we can see how there can be failures of institutional practices, policies, and procedures that can lead people to experience credibility deficits due to their trauma-induced memory errors. It is no secret that asylum seekers experience trauma, or that memory can be negatively impacted by trauma. There is no lack of evidence showing how asylum seekers are likely to provide accounts that contain errors even if they are in a good position to provide an accurate account of core details relevant to their asylum claim. In fact, this idea is reflected in guidance or guidelines produced by the Common European Asylum System (EASO, 2018: 75), the UK Home Office (Home Office, 2015: 13), and the UNCHR Handbook on Procedures and Criteria for Refugee Status (UNHCR 2019: Paragraph 199).11

This means that those involved in making assessments of the credibility of asylum claims could be given adequate guidance and training on how to judge whether falsities contained in a testimony should lead the core details to be dismissed. Given that the goal of the asylum process is to produce correct judgments about whether people are in need of protection—something that it is vital to get correct—it seems that institutions that administer the asylum process (e.g., the UK Home Office) have strong epistemic and moral duties to ensure that their staff are given adequate advice and training about this issue. In the absence of this adequate training and advice, even unprejudiced asylum case workers, actively attempting to make a correct judgment about whether an asylum seeker is telling the truth, may systematically and predictably give less credibility than is deserved to testimony. They may respond reasonably, and without malice or intention to catch anyone out, to evidence of false details in an asylum seeker's account by dismissing core pieces of the testimony that are highly likely to be accurate. In such cases, we argue, there is an institutional epistemic injustice. The credibility deficit constitutes an epistemic injustice, and the injustice finds its source in the institutional failure to provide adequate guidance and training to asylum case workers.

It is worth noting, again, that the injustices described in this section may be experienced by anyone who suffers memory related epistemic harms due to trauma or stress, regardless of whether it was caused by social and institutional structures or bad luck, if the hearer responds with prejudice or ought to have been provided with appropriate training to detect these kinds of memory errors. But there seems to be something particularly troubling about cases where social and institutional structures cause memory distortions and disorganization, that is, mnemonic epistemic injustice, and, subsequently, either the same or different social and institutional structures subject the sufferers to testimonial-type epistemic harms in response to such memory errors.

In previous sections, we argued that people who undergo trauma or stress due to social and institutional structures can experience what we call mnemonic epistemic injustice. This section has aimed to show that when this epistemic injustice occurs prior to a person giving testimony about their need for future help or support, the initial epistemic injustice can be compounded if core details of their testimony are accurate but are dismissed as lacking credibility. Although we have focused on the case study of the asylum system, similar phenomena may be found in other institutions where people who have experienced trauma or stress due to the nature of social and institutional structures are required to provide testimony to evidence their need, for example, in the welfare system or where people are applying for support due to domestic abuse. The initial injustice is structural or institutional, that is, due to features of social or institutional structures that impose trauma. The compounding epistemic injustice can be either interpersonal, for example, due to the prejudice of the hearer, or institutional, if it is due to institutional failings.

We have argued that there are epistemic harms that occur when people experience trauma or stress that negatively impact their memories, and that these harms can constitute mnemonic epistemic injustice. The injustice occurs because people are unfairly disadvantaged as epistemic agents by being denied something that they are entitled, that is, freedom from the imposition of stress and trauma that brings epistemic costs and other damage to their objective needs. On its own, mnemonic epistemic injustice is serious, because it constitutes severe harms to individuals' objective interests in good epistemic agency, capacity for autonomy, and general well-being. But we have argued that this epistemic injustice is sometimes compounded when individuals subject to this kind of injustice are required to testify about their need for help or protection. In these cases, the initial mnemonic epistemic injustices are compounded by hearers' responses to memory errors that people experience due to trauma.

One implication of our argument for the understanding of the boundaries of epistemic injustice is that people can experience epistemic injustices that are extremely like testimonial injustice because of other people's responses to evidence that they are speaking untruths. The focus of attention in the testimonial injustice literature has tended to be on how true beliefs that people attempt to communicate can be dismissed due to prejudice. Jennifer Lackey's (2020, 2021, 2022) work on agential testimonial injustice is an exception, highlighting how people are sometimes only believed when they say things that are untrue. We have argued here that people can also experience epistemic injustice in cases when they are disbelieved when (and because) they say false things.

A further implication is that hearers can be involved in a compounding form of epistemic injustice while responding in a reasonable way to evidence that they have available to them. If a hearer is operating within the structures of an institution in which they are not given adequate guidance and training that allows them to give appropriate weight to evidence of errors, they can respond in a reasonable way to evidence of errors but thereby be implicated in epistemic injustice. For example, asylum case workers who notice errors, such as inconsistencies in the chronology of an account given by an individual asylum seeker may reasonably conclude that the errors indicate that the account is untrustworthy. More specifically, they may reasonably conclude this if they have not been given adequate guidance or training about how memories are influenced by trauma.

In short, our argument highlights how social and institutional structures can, metaphorically speaking, conspire to create an epistemically inhospitable environment for marginalized individuals. Where there is an initial mnemonic epistemic injustice that is compounded, the initial injustice may be the result of one part of a social or institutional structure while the compounding injustice is the result of the operation of another. This seems to be precisely what is the case for those asylum seekers who have stressful and traumatizing experiences in the asylum system, which lead their memories to become disorganized and distorted, and then face credibility deficits when they provide their account to asylum case workers.

A significant number of issues need to be addressed to tackle the epistemic injustices we have identified. Not only is it necessary to tackle the prejudice of people who hear the testimony of others who have experienced trauma and stress, it is also necessary to ensure that institutions give their staff guidance on how memory errors are consistent with a person being able to provide a strong account of the core details of their experiences. Both these strategies would be required to eradicate the compounding epistemic injustice. But to tackle the initial mnemonic epistemic injustice far more radical changes are needed, that is, changes to social and institutional structures to reduce the heightened risk that trauma and stress is experienced by individuals existing within them.

There are no conflict of interests.