Culture and credibility: the assessment of asylum seekers’ statements

ABSTRACT

Asylum seekers typically have a different cultural background from the immigration officials interviewing and assessing their statements. Yet the role of culture in eliciting and evaluating asylum seekers’ statements has remained largely unexplored. This article addresses the intersection of memory and culture in the context of the asylum determination process. To that end, we draw on the literature on eyewitness memory, which has so far largely been studied separately from credibility assessments in asylum procedures. We integrate the academic literature on the two subjects by discussing memory aspects prone to cultural variations through the lens of credibility assessments in the asylum determination procedure. We highlight the need for further research, given that most studies on memory have involved participants from Western, Educated, Industrialised, Rich and Democratic (WEIRD) societies, ignoring 88 per cent of the world population. More insight into cultural differences within memory and reporting will allow for the creation of culturally nuanced credibility assessment tools that better accommodate the majority of the world population.

KEYWORDS:

• Culture
• memory
• cross-cultural
• asylum seekers
• credibility assessment

In recent decades, more and more people report their memories in cross-cultural contexts, such as asylum interviews and international criminal cases. Differences in how people from different cultural backgrounds remember and talk about events pose challenges for all involved (Anakwah et al., 2020; Jobson, 2009; Wang, 2021). The cultural setting in which someone grows up influences how memories are encoded, stored, retrieved and reported. Reporting memories in cross-cultural contexts is not just a one-way street, as culture affects how memory statements are evaluated as well (e.g. by police interviewers or judges). A specific context in which these three elements – culture, memory and the evaluation of memory statements – come together is in the asylum determination procedure. In this article, we systematically integrate and apply insights from the eyewitness memory literature, focusing on cultural aspects, to the applied context of asylum credibility assessment.

Eyewitness memory has been researched extensively (for a recent review, see Odinot et al., 2021) but studies have predominantly involved participants from Western, Educated, Industrialised, Rich and Democratic (WEIRD) societies, ignoring 88 per cent of the world population (Apicella et al., 2020; Henrich et al., 2010; Rad et al., 2018). Importantly, asylum seekers are typically from non-WEIRD societies, while immigration officials are from WEIRD societies (Eurostat, 2022). More recently, there have been calls for psychological research outside of WEIRD societies (Brady et al., 2018; Gelfand et al., 2017; Hope et al., 2022), and legal psychologists have started to assess the influence of culture on eyewitness memory (e.g. Anakwah et al., 2020) but there is still much ground to cover. Illustrative is Schacter’s (2022) recent update of his seminal work (Schacter, 1999) on the seven sins of memory, focusing on applied contexts. Although Schacter’s work is relevant to various applied contexts, including eyewitness memory, its focus is solely on WEIRD countries (see Vredeveldt & De Bruïne, 2022, for a discussion of cultural aspects of memory sins).

Similarly, credibility assessments in the asylum determination procedure have only recently become the subject of psychological research (Selim et al., 2022; Skrifvars et al., 2022; Van Veldhuizen et al., 2017). So far, the influence of culture on eyewitness memory on the one hand, and credibility assessments in asylum procedures on the other, have been studied separately, with an occasional reference to each other (e.g. asylum context as an example of a cross-cultural interview or cultural differences as an example of an important aspect to consider in credibility assessment). The aim of the current paper is to integrate the academic literature on the two subjects by discussing aspects of memory statements that vary across cultures through the lens of credibility assessments in the asylum determination procedure. We discuss the credibility indicators often used to assess asylum seekers’ statements and how cultural differences in memory statements complicate the assessment of those indicators.

Culture is difficult to define. Drenk et al. (2023) found only six explicit definitions of culture in their review of 87 articles on the role of culture in witness testimony. In the current paper, we use Hofstede’s (1980) definition: ‘the interactive aggregate of common characteristics that influence a human group’s response to its environment’ (p. 21). According to Hofstede (1980, 2001), there are six cultural dimensions in which societies differ from each other. Three of those will feature in our discussion: collectivism-individualism (the extent to which societies value autonomy as opposed to group cohesion), power distance (the extent to which people accept hierarchy and unequal power distribution in society) and uncertainty avoidance (the extent to which societies tolerate uncertainty and ambiguity).

Credibility assessments in the asylum context

The focus of this article is on the asylum determination procedure in the European Union (EU). EU member states must assess and decide upon the eligibility of each applicant to be granted asylum (Qualification Directive, 2011). Seeking asylum is a human right and is granted to protect persons who face persecution in their country of origin (United Nations General Assembly, 1948; United Nations High Commissioner for Refugees (UNHCR), 2001). To obtain refugee status, applicants must reach the level of conviction required and establish their fear of persecution with ‘a reasonable degree of likelihood’ (European Asylum Support Office (EASO), 2015, p. 21). Applicants have the duty to substantiate their claims. They must convince immigration officials that they risk threats to their life or safety if they return to their country of origin to gain international protection (e.g. Gorlick, 2003; Sweeney, 2009; UNHCR, 1998, 2013; Van Veldhuizen et al., 2017). Immigration officials from the host state evaluate the statements and corroborative evidence provided by asylum seekers to prove their identity, country of origin and qualifying circumstances (Qualification Directive, 2011). Usually, such a determination process consists of conducting interviews, assessing the credibility of the claim and evaluating the foundation of the alleged threats to life or safety (Selim et al., 2022; Selim et al., 2022; UNHCR, 2013). Based on that evaluation, immigration officials decide whether an asylum seeker is eligible for international protection (i.e. refugee status).

The asylum applicants’ knowledge of their home country is an important aspect judged by officials in their credibility assessment (e.g. Byrne, 2007; Cohen, 2001; Herlihy et al., 2012; Kagan, 2002). Applicants are commonly asked about their place of birth, their journey to Europe and incidents or persecution that may have taken place in their home country. These questions come close to those asked in eyewitness interviews about the scene of the crime and witnessed incidents (Van Veldhuizen et al., 2018). In effect, asylum seekers can be considered to be eyewitnesses of their own lives during asylum interviews (Herlihy et al., 2012; Herlihy & Turner, 2009).

Evaluating the credibility of asylum claims has proven difficult. Since documentary evidence (e.g. an identity document or an arrest warrant) is often absent, decisions in these cases are typically based solely on the applicant’s own statement in an asylum interview (e.g. Gyulai et al., 2013; Herlihy et al., 2012; Noll, 2005a, 2005b). Their assessed credibility depends on their ability to provide detailed and consistent statements about their identity, origin and past experiences, for which they are required to use and communicate their autobiographical memory (Cameron, 2010; Herlihy et al., 2012; UNHCR, 2013; Van Veldhuizen et al., 2018).

In practice, European immigration officials often use five credibility indicators when assessing statements provided during an interview (Gyulai et al., 2013; UNHCR, 2013), discussed further below. However, researchers dispute whether these indicators can validly discriminate between truthful and false statements (Granhag et al., 2005; Herlihy et al., 2010; Skrifvars et al., 2022). The discussion, however, hardly accounts for cultural variations within those credibility indicators, despite the relevance of empirical research on cultural differences in autobiographical memory (e.g. Herlihy et al., 2012; Jobson, 2009; Jobson & O’Kearney, 2008; Wang & Conway, 2004).

In the vast majority of cross-cultural studies, Western societies (e.g. Western Europe and USA) were compared with East Asian societies (e.g. China and Japan) (Veillard, 2017), leaving a knowledge gap and possibilities for cross-cultural studies involving other non-WEIRD societies. From a practical point of view, certainly for asylum applications in the EU, comparing Western to East Asian societies is not as relevant as comparisons with people from other societies. In 2021, the top three countries of origin of asylum seekers in the EU were Syria, Afghanistan and Iraq (Eurostat, 2022). Further, sub-Saharan African countries accounted for over one-fourth of asylum-sending states to Europe in 2021 (Eurostat, 2022). It would therefore be more practically relevant to examine cultural differences between people from Western Europe and people from the Middle East and Sub-Saharan Africa.

Culture and credibility assessments

Obtaining high-quality reports from asylum seekers is crucial for the successful functioning of the system, and no less important, for the future of the applicant. After the interview, immigration officials must assess the credibility of the asylum claim. Now, their cultural background will affect their evaluation of the asylum seeker’s statement. Immigration officials hold beliefs and expectations about what a genuine asylum claim looks like (e.g. Granhag et al., 2005; Herlihy et al., 2010). Their beliefs and expectations have been shaped by their own cultural background.

Immigration officials generally do not receive feedback on their decisions, which means that immigration officials cannot learn from potential mistakes in previous decisions (DePaulo & Pfeifer, 1986; Ekman & O’Sullivan, 1991; Vrij & Semin, 1996). It also means that their own biases and assumptions will generally not be disproved and are therefore likely to remain unchanged. Immigration officials may also resort to stereotypes and heuristics because of the uncertain decision-making context in combination with a high workload and time pressure (i.e. stressful working conditions), to reduce cognitive complexity (Lustig et al., 2008; Tversky & Kahneman, 1974). Indeed, research on the asylum determination process shows a lack of objectivity and motivation in asylum decisions (Maegherman et al., 2018; Millbank, 2009; Skrifvars et al., 2022; Sweeney, 2009). After analysing decisions in UK asylum cases, Herlihy et al. (2010) concluded that asylum judges based their evaluations on assumptions about behaviour and truthful accounts that do not align with findings from psychological research. Similar assumptions have been found in evaluations of Australian Refugee Review Tribunal (Dowd et al., 2018) and Finnish immigration officials members (Skrifvars et al., 2022).

People are generally not good at detecting deception (e.g. Bond et al., 2013; Bond & DePaulo, 2006; Hartwig & Bond, 2011; Masip, 2017) but when attempting to detect deceit in people with a different cultural background, people perform even worse in terms of lie detection accuracy (Bond et al., 1990; Bond & Atoum, 2000; Vrij, 2000 for an overview see, Taylor et al., 2014). A widely accepted explanation is the ‘expectation violation model’, in which scepticism and suspicion arise from behaviour that does not meet expectations (e.g. Bond et al., 1990; Levine et al., 2000; Taylor et al., 2014). In addition, professionals are less likely to believe that others are telling the truth than the general public (Masip, 2017; Meissner & Kassin, 2002; Selim et al., 2022). Taken together, this contributes to a ‘culture of disbelief’ in asylum contexts (Selim et al., 2022, p. 2).

Typically, there are five credibility indicators used to assess the information elicited in asylum interviews: (a) level of detail and specificity, (b) internal consistency, (c) consistency with information from external sources, (d) consistency with statements from other witnesses, and (e) plausibility (Gyulai et al., 2013; UNHCR, 2013; Van Veldhuizen, 2022). As mentioned, the validity of these indicators is considered poor (see e.g. Granhag et al., 2005), but there is also discussion about whether the indicators can be used objectively and independent of each other (Maegherman et al., 2018). Importantly, there is little research on how someone’s cultural background may affect those indicators and, ultimately, the credibility assessment.

Detail and specificity

The use of level of detail and specificity as credibility indicator is problematic because people from collectivistic societies tend to provide less detailed and specific reports than people from individualistic cultures (e.g. Anakwah et al., 2020; Jobson, 2009; Wang & Ross, 2005). Differences in level of detail and specificity in memory reports can arise at different stages, discussed further below.

Encoding

Before details can be reported, they must be encoded. Findings from cross-cultural research show cultural variations in cognitive processes (Allport & Pettigrew, 1957; Gutchess & Sekuler, 2019; Hudson, 1960; Segall et al., 1966). For instance, there are cultural differences in visual perception. People from rural African cultures perceive depth in pictures less frequently and are less susceptible to rectangular illusions than people from Western societies (Allport & Pettigrew, 1957; Hudson, 1960; Segall et al., 1966; for an overview of cross-cultural studies on perception and illusions see, Deręgowski, 2017). That is likely due to their reduced exposure to two-dimensional representations and rectangular shapes in daily life (Kilbride & Robbins, 1969). Since the 1960s, hardly any empirical research has been conducted in this field, identifying a knowledge gap in the contemporary literature. Deręgowski (2017, p. 51) observes: ‘The great wave of research activity concerning cross-cultural differences in perception of illusions has now, as the list of references shows, abated. This does not mean that all the questions have been answered. It may mean that the value of such studies to understanding of perceptual processes is underestimated.’ We are aware of only one recent study in which Hudson’s (1960) findings were replicated. De Bruïne et al. (2018) found that sub-Saharan African asylum seekers (originating from both rural and urban areas) were significantly less likely to perceive depth in pictures compared to a matched Dutch control group.

The importance of cultural variations in the encoding of an event should not be underestimated in the asylum context. Immigration officials often test the applicant’s knowledge of the place, local customs and common objects of the specific country, area and town the asylum seeker claims to originate from (Van Veldhuizen et al., 2017). The officials may show pictures of landmarks or objects, aerial pictures, or Google Maps to establish the region of origin. Asylum seekers may not be able to transform such two-dimensional images into the three-dimensional locations and landmarks they represent. If asylum seekers have difficulty interpreting pictures or maps, they will be unable to point out locations or identify landmarks. That does not mean that they do not know them or are lying about where they are from, but may simply mean that the asylum seeker is relatively unfamiliar with two-dimensional representations.

Another aspect that depends on someone’s cultural background is how time is perceived (e.g. Anders, 2011; Combs, 2010, 2017; Kagan, 2002). Observations in international criminal cases show that witnesses from rural areas do not know where and when they were born or how long they have been married, simply because they do not routinely keep track of calendar dates or give estimates in hours, days, or months. The same holds for other estimations, such as distances between places, the size of objects and the number of people present at an incident (Anders, 2011; Combs, 2010, 2017). If such information is not encoded or encoded differently, it cannot be retrieved and may cause misunderstandings in an interview. Van Veldhuizen (2017) reported an example of immigration officials who cast doubt in a particular case on an asylum seeker’s story because he reported the day of his brother’s murder incorrectly: a day too late. It later transpired that the brother was murdered after sunset, which is considered the next day in the asylum seeker’s culture. The story was initially considered implausible, and the application was denied, but once the cultural context was explained, the asylum seeker was granted asylum.

People’s cultural background also influences what they pay attention to in a scene. Witkin and Berry (1975) argued that individuals can have a field-dependent (i.e. focus on environment and context) or field-independent (i.e. focus on individual elements) cognitive style, depending on whether the individual comes from a society that values conformity (collectivistic non-WEIRD societies) or autonomy (individualistic WEIRD societies), respectively. Similar patterns have been observed in studies on cross-cultural preference for analytical or holistic perceptual processing (e.g. Gutchess et al., 2006; Kitayama et al., 2003; Masuda & Nisbett, 2001; Miyamoto et al., 2006). Masuda and Nisbett (2001) found that American participants focused more on individual fish in an underwater scene (i.e. analytical perceptual processing), whereas Japanese participants focused more on the background and relations among the fish (i.e. holistic perceptual processing). These cultural differences in perception may be explained by the ancient philosophies underlying them: The ancient Asian philosophies emphasise balance and harmony rather than independence, whereas the opposite is true for the ancient Greek philosophy often seen as inspiration for Western cultures (Nisbett et al., 2001; Norenzayan & Nisbett, 2000; Varnum et al., 2010).

Within the asylum context, immigration officials tend to believe that someone who is genuinely from the claimed region of origin is able to give a detailed description of frequently encountered objects, such as local currency or identity documents (Van Veldhuizen et al., 2017). Considering the limitations of human memory, it is unlikely that such an assumption holds for people from any culture. Moreover, our hypothesis would be that asylum seekers (often from societies high in collectivism) encode fewer details of individual objects because of their holistic perceptual processing and field-dependent cognitive style. That may make it even more difficult to describe objects in great detail, potentially leading immigration officials to doubt their credibility.

To conclude, what elements are perceived and how these elements are encoded is related to someone’s cultural background. Cultural differences in encoding, such as familiarity with two-dimensional representations, the perception of time and space and differences in perceptual processing, should be carefully considered in asylum interviews and assessments of credibility.

Retrieving and reporting

In an asylum interview, memories are retrieved and reported. What aspects are retrieved and how they are reported varies between cultures (e.g. Anakwah et al., 2020; Jobson, 2009; Leal et al., 2018; Wang, 2021). In cross-cultural psychological research, differences in memory reports’ length, specificity, emotionality, and focus have been found.

People from individualistic societies are more likely to provide lengthy, specific, and emotionally elaborate memories (Herlihy et al., 2012; Jobson & O’Kearney, 2008; Wang, 2001, 2021). Their independent self-construal is associated with a focus on memories that relate solely to themselves (e.g. graduation day; Jobson & O’Kearney, 2008; Marian & Kaushanskaya, 2004; Wang, 2001). In contrast, individuals from collectivistic societies are more likely to focus on memories concerning collective activities, general routines, and emotionally neutral events (Herlihy et al., 2012; Jobson & O’Kearney, 2008; Wang, 2001). Their interdependent self-construal is associated with a focus on memories that relate to their social identity and relationships (e.g. a family birthday; Jobson & O’Kearney, 2008; Marian & Kaushanskaya, 2004; Wang, 2001). These different styles of retrieving and reporting memories can lead to misunderstandings if immigration officials hold assumptions or expectations of what a genuine report would look like based on their own cultural background. For example, an asylum seeker’s report about a raid on the village may provide little detail about what the attackers looked like, and instead focus on the collective flight of the villagers in an emotionally neutral way. That reporting style could lead to suspicion in immigration officials who may conclude that the event is fabricated because it was not reported in a detailed, emotionally elaborate manner with a focus on individual actions.

Asylum interviews often involve descriptions of traumatic events. Research shows that traumatic memories tend to be less specific than memories of everyday events (Brewin, 2011; Johnson & Thompson, 2008). Further, culture influences retrieval and reporting of traumatic memories (e.g. Humphries & Jobson, 2012). Trauma memory has been examined cross-culturally in only a few studies (e.g. Graham et al., 2014; Humphries & Jobson, 2012; Jobson, 2011; Jobson & O’Kearney, 2006), which have shown both similarities and differences between individualistic and collectivistic cultures (for an overview, see Vredeveldt et al., 2023). Many asylum seekers experience depression and post-traumatic stress disorder (PTSD) (Fazel et al., 2005). Studies in WEIRD samples have shown that PTSD and depression are associated with overgeneral autobiographical memory retrieval (for a review, see Brewin, 2011). Graham et al. (2014) assessed through clinical interviews whether that overgenerality extended to asylum seekers from diverse cultural backgrounds. They concluded that asylum seekers experiencing PTSD and depression also provide less specific memory reports than those without PTSD or depression. There is, however, little empirically supported knowledge on how someone’s cultural background relates to the qualities of trauma memories. More research is needed on how culture affects trauma memory especially in the cross-cultural context of the asylum determination process.

Cross-cultural communication

Retrieving and reporting memories are not isolated acts, but rather a result of intercultural communication (cf. Skrifvars et al., 2022). Successful intercultural communication depends on a multitude of factors (for an overview, see Taylor et al., 2014). For example, Hall (1976) proposed a distinction between low- and high-context communication. Low-context communication, typical for individualistic societies, is direct and relies on the explicitness of the communicator. High-context communication, typical for collectivistic societies, is more implicit and relies on the context, leaving many things unsaid; ‘letting the culture explain’ (Leal et al., 2018, p. 193). Several studies comparing low-context and high-context samples have found that high-context participants reported significantly fewer details than low-context participants (e.g. British vs. Arab vs. Chinese: Leal et al., 2018; and British vs. Arab: Vrij et al., 2021). The researchers coded only details that they considered relevant to the to-be-remembered event, so it is not clear whether high-context participants reported more information about context that was deemed irrelevant by the researchers. Another example of differences between high- and low-context communication is whether or not personal pronouns (e.g. I or we) are used. There are several high-context cultures, such as Japan, that do not use pronouns (for an overview, see Tabata & Vrij, 2023). If an asylum seeker does not use pronouns during the interview, it may become unclear to the immigration official who the actor is in the applicant’s story. For instance, an applicant reports: ‘because (I was) supposed to meet (him)’ (cf. Kashima et al., 2014, p. 52), leaving the immigration official wondering who was supposed to meet whom. In the best case, the immigration official asks for clarification, but in the worst case, the statement is considered insufficiently specific and therefore noncredible.

Cultures also differ in power distance; the extent to which people accept hierarchy and unequal power distribution in society (Hofstede, 2001). In cultures with high power distance, people are less likely to express their views to authority figures (Anakwah et al., 2021; Ghosh, 2011). Anakwah et al. (2021) conducted a controlled eyewitness experiment comparing Dutch (low power distance) and Ghanaian (high power distance) mock witnesses and found that mock witnesses from low power distance societies provided more details than those from high power distance societies, which may reflect a reticence to report in participants from a high power distance culture.

Power distance may also influence a person’s response style. De Bruïne et al. (2018) found that sub-Saharan African participants were more likely to answer ‘yes’ on a recognition task than Dutch people, regardless of whether they had actually seen the object before. The desire to agree with the experimenter seemed to outweigh the desire to give an accurate response (known as an acquiescence response style; Cheung & Rensvold, 2000).

Another factor that may affect cross-cultural communication is the extent to which societies tolerate uncertainty and ambiguity, known as uncertainty avoidance (Hofstede, 2001). Individuals from cultures with a high level of uncertainty avoidance tend to rely strongly on their set rules and ways of doing things. It may be seen as an attempt to control as much of the uncertainty as possible (Hofstede, 2001). There is little empirical research on the role of uncertainty avoidance in investigative interviewing in cross-cultural settings. However, Giebels et al. (2017) assessed the role of uncertainty avoidance in crisis negotiation and concluded that more formal communication by police negotiators could be beneficial when communicating with suspects high in uncertainty avoidance. They suggest that uncertainty avoidance is particularly relevant to complicated and risky interactions in which the parties are trying to make sense of each other’s intentions and position, such as a negotiator-perpetrator dialogue.

We argue that uncertainty avoidance is therefore relevant to the high-pressure context of the asylum determination process, where the asylum seeker is trying to convince the immigration official of an acute need for international protection. Kagan (2002, p. 393) provided an example of a Somali asylum seeker who continued to answer ‘no’ to questions about the leading people from the organisation to which he said he belonged and the tallest building in Mogadishu. It later transpired that he was aware of whom those prominent people were and that he was able to identify the tallest building in Mogadishu. He did not feel comfortable, however, answering positively to these questions, since he did not have a personal relationship with these people and never stepped inside that building. Uncertainty avoidance may have been an underlying factor in the discrepancy between the culturally normative expectations of the immigration official asking those questions and the asylum seeker’s interpretation of those questions based on his own cultural background. Thus there is a need for more cross-cultural research examining the role of uncertainty avoidance in investigative interviewing.

The frequent use of interpreters in asylum interviews can complicate matters even further. Evans et al. (2019) provide an overview of challenges related to the use of interpreters in investigative interviews. For example, interviewees provide fewer details when interviewed through an interpreter than when interviewed in their native or second language (Ewens et al., 2016). In addition, in their literature review on credibility assessments of asylum claims based on religion, Selim et al. (2022) identify three potential issues related to the use of an interpreter in asylum interviews: (1) mistrust and concerns about confidentiality; (2) the lack of specific cultural or religious knowledge; and (3) the risk to unintentionally change the questioning style into a more confrontational style. All three issues could potentially lead to less detailed and specific statements, either as a result of reluctance of the asylum seeker to disclose because of mistrust or disrupted rapport building or the interpreter proves unable to translate with sufficient specificity.

Clearly, differences in communication style can lead to miscommunications. Statements from high-context cultures may be judged less credible based on the credibility indicators in the asylum determination process, since they lack the level of detail, specificity and relevance that is expected by immigration officials from low-context cultures. The use of interpreters in asylum interviews complicates matters even further since it may reduce the level of detail and specificity even further.

Level of detail as a deception cue

Immigration officials generally seek to discriminate between truthful and fabricated asylum claims. Two meta-analyses on deception detection identified level of detail as one of the most robust deception cues, indicating that truth-tellers typically provide more details than liars (Amado et al., 2016; DePaulo et al., 2003). The findings about level of detail from the deception detection literature may, however, not be applicable to non-WEIRD societies. For example, Taylor et al. (2017) found that North Africans provided more rather than fewer contextual details when lying about a personal experience. Leal et al. (2018) compared British, Israeli-Arab and Chinese participants and found truth-tellers to be significantly more detailed than liars across all cultural groups, but found that British participants overall reported significantly more details than Chinese and Arab participants. Vrij et al. (2021) similarly found that truth-tellers provided more details than liars in British as well as Arab samples, but British participants overall provided significantly more details than Arabs. Although it appears that within cultural groups truth-tellers generally provide more details than liars, both studies warrant caution when using level of detail as deception cue between cultural groups, because non-WEIRD participants invariably reported fewer details than WEIRD participants.

A tool used for veracity assessments is Criteria-Based Content Analysis (CBCA; Steller & Köhnken, 1989). It involves 19 criteria observed more frequently in genuine statements than in fabricated ones (for a review, see Vrij, 2005). One criterion is contextual embedding, which involves ‘references to time and space’, for example ‘I saw her for the last time last week at the market’ (cf. Vrij, 2005, p. 5). According to the CBCA, an account is considered more truthful if it includes references to time and space (Vrij, 2005). Yet, assessing the veracity of a statement in a cross-cultural setting based on contextual embedding is problematic, as people in some cultures do not routinely keep track of calendar dates or time (see e.g. Kagan, 2002), as previously discussed. Illustrative is the Rwandan witness described by Combs (2009, p. 243) who could only recall that the rally (which the defendant allegedly attended) occurred sometime during the rainy season.

In conclusion, using the level of detail and specificity as an indicator of credibility in the cross-cultural asylum context raises problems. Asylum seekers’ statements might come across as deceitful if their statements lack the expected level of detail and specificity, even though this lack could be explained by the asylum seeker’s traumatic experiences, their collectivistic background and communication style, or the involvement of an interpreter.

Consistency

Three of the five credibility indicators relate to consistency, which appears to be a central element in the asylum determination process. Both laypeople and legal professionals often believe that inconsistency is a cue of deception (e.g. Akehurst et al., 1996; Granhag & Strömwall, 2000; Vredeveldt et al., 2014). More specifically, Granhag et al. (2005) found that Swedish immigration officials believed that deceptive statements are less likely to be consistent over time than truthful ones. However, the common belief that inconsistency is a cue to deception is not entirely in line with current knowledge from memory research. Research shows that witness statements with relatively many contradictions are generally no less accurate overall than statements with few contradictions (Fisher et al., 2013; Smeets et al., 2004). Thus, the fact that a statement contains contradictions does not automatically mean that the entire statement is inaccurate. In addition, research shows that almost all witnesses in a second interrogation omit information and add new information. In short, all forms of inconsistencies occur in statements by witnesses who (attempt to) tell the truth.

Inconsistencies in truth-tellers and liars

Vredeveldt et al. (2014) conducted a systematic review of different types of inconsistency in truth-tellers and liars, which may also be observed in the asylum determination process. Vredeveldt and colleagues distinguish between four types of consistency: (a) within-statement consistency, (b) between-statement consistency, (c) within-group consistency, (d) statement-evidence consistency. Within the five criteria of credibility assessment in asylum procedures (Gyulai et al., 2013; UNHCR, 2013; Van Veldhuizen, 2022), within-statement and between-statement consistency seem to have been merged into one indicator: internal consistency. Within-group consistency corresponds to consistency with statements from other witnesses, and a parallel can be drawn between statement-evidence consistency and consistency with information from external sources. Vredeveldt and colleagues concluded that liars are typically either equally consistent or even more consistent than truth-tellers, with the exception of statement-evidence consistency. A widely accepted explanation for that is the ‘repeat-versus-reconstruct hypothesis’ introduced by Granhag and Strömwall (1999). The hypothesis states that liars carefully repeat their stories to avoid being caught, stimulating consistency, while truth-tellers reconstruct their stories relying on their genuine memories, thereby increasing the likelihood of inconsistencies.

In the asylum context, statements are expected to align with Country-of-Origin Information (COI) reports (i.e. consistency with external sources). COI reports can be based on fieldwork or desk research, and often rely on other immigration services’ reports, non-governmental and media reports. These reports contain information about, for example, security situations, political systems, religion, and human right legislation, but also customs, history and traditions (e.g. Bodström, 2022; Dauvergne & Millbank, 2003; Gibb & Good, 2013). Immigration officials generally use the information from the COI reports for knowledge tests to establish credibility (Van Veldhuizen et al., 2017). Yet, the COI reports are subject to criticism for their alleged objectivity in the asylum determination process (e.g. Bodström, 2022; Gibb & Good, 2013). Another point of criticism is the rigorous interpretation of these reports. In an example provided by Bodström (2022), an asylum claim was judged to be inconsistent with external sources because it was inconceivable to the judges that an extremely violent militia (as described in the COI report) would just have a conversation with a group of villagers without harming them (as described by the asylum seeker).

Regarding statement-evidence consistency, liars tend to be less consistent than truth-tellers (Vredeveldt et al., 2014). This type of consistency may therefore be useful in the credibility assessment of asylum seekers’ claims. The Strategic Use of Evidence (SUE) technique may be of interest in asylum procedures, since it tends to increase differences in statement-evidence consistency between liars and truth-tellers (Hartwig et al., 2005; Hartwig et al., 2006; Jordan et al., 2012). However, the SUE technique first needs to be tested cross-culturally and in the asylum context to explore its usefulness in this setting. In addition, the COI reports must be of high quality if they are to be used to assess statement-evidence consistency.

Inconsistencies in cross-cultural contexts

There are various reasons for inconsistencies, especially in the asylum context (for an overview see, Cohen, 2001). For instance, inconsistent statements can be a result of normal functioning of memory such as forgetting (e.g. Schacter, 2022) or hypermnesia (e.g. Bluck et al., 1999). It can also be a result of external factors, such as leading questions (e.g. Loftus, 1975a) or the presence of an interpreter mistrusted by the interviewee (Selim et al., 2022). Further, exposure to news coverage and discussions with other witnesses results in the incorporation of new information into existing memories, referred to as memory conformity or social contagion (Gabbert et al., 2003; Roediger et al., 2001; Vredeveldt et al., 2016; Vredeveldt et al., 2017).

Every time we retrieve a memory, the content is altered in some way (Loftus, 1975b; Treanor et al., 2017). Culture influences how memories change over time before an asylum interview takes place. Through frequent retellings of memories to family members, friends and institutions (e.g. the South African Truth and Reconciliation Commission hearings and Rwandan Gacaca courts), people build collective memories that form part of their cultural, social and personal identities (Boyer & Wertsch, 2009; Hirst & Echterhoff, 2012; Wang, 2001).

Social contagion can consist of accurate information but also of inaccurate information. Susceptibility to misinformation, potentially leading to inconsistencies, may be linked to the concept of self-construal (see previous section Retrieving and Reporting; Anakwah et al., 2022; Petterson & Paterson, 2012). Petterson and Paterson (2012) assessed the role of independent versus interdependent self-construal (on an individual level) in susceptibility to misleading post-event information in a co-witness paradigm with Australian participants. After witnessing an event, participants discussed it with a ‘co-witness’ (in reality an experimental confederate), who introduced both correct and incorrect information about the event. The results showed that those with higher levels of independent self-construal were significantly less likely to accept post-event misinformation suggested by a co-witness. No significant relation was found between interdependence and memory conformity, perhaps because all participants were from an individualistic society. Conducting a similar study cross-culturally, comparing a WEIRD and non-WEIRD sample, would provide more insight into cultural differences in susceptibility to misinformation in a co-witness paradigm.

We are aware of only one study assessing cross-cultural differences in susceptibility to post-event misinformation. Anakwah et al. (2022) provided mock witnesses from Ghana and the United Kingdom (UK) with post-event misinformation, in the form of a media report, about an event they had witnessed on video. The researchers found no significant differences in misinformation endorsement during free recall. However, Ghanaians endorsed more misinformation on the recognition task than the UK participants. Translating that finding to the asylum context, we hypothesise that asylum seekers from collectivistic backgrounds are at higher risk of endorsing misinformation, because most of the questions in asylum interviews rarely allow for free recall. Van Veldhuizen et al. (2018) and Skrifvars et al. (2020) found that approximately four-fifths of the questions in the analysed Dutch and Finnish cases were closed and fact-checking. In fact, there was no opportunity for applicants to provide a free recall during their application (Van Veldhuizen et al., 2018). The greater likelihood of endorsing misinformation together with the type of questions asked, could potentially lead to a statement that reflects what an asylum seeker believes the interviewer wants to hear rather than what actually happened.

There are other possible reasons for inconsistencies that may specifically occur in the asylum context. First, late disclosure of sensitive information could lead to internal inconsistency (i.e. between-statement inconsistency) (Bögner et al., 2010; Herlihy & Turner, 2006). Second, asylum seekers may fail to recall trauma memories in the same way in repeated interviews (Herlihy & Turner, 2006), because mental health problems affect the consistency of asylum seekers’ statements (Bögner et al., 2007; Herlihy et al., 2002; Herlihy et al., 2010; Herlihy et al., 2012). Herlihy et al. (2012) assessed inconsistency in autobiographical memory of Bosnian and Kosovar asylum seekers and found that discrepancies were common. Herlihy and colleagues also found that for those showing high levels of post-traumatic stress, the length of the application process increased the number of discrepancies.

In conclusion, scientific literature cautions against the rigid use of consistency as a cue to deception, as there is little empirical evidence to suggest that liars are always less consistent than truth-tellers. In addition, there is a need for empirical research outside of the WEIRD context to provide insight into cultural differences in consistency between truth-tellers and liars as well as susceptibility to misinformation.

Plausibility

There is little consensus on the exact definition of plausibility, but subjective terminology such as ‘reasonable and likely’ is commonly used to define the concept (UNHCR, 2013, p. 176). Scientific literature as well as training manuals and policy documents contain warnings about the subjectivity of these definitions (Gyulai et al., 2013; Maegherman et al., 2018; UNHCR, 2013). Immigration officials’ assessments of what is reasonable or likely are influenced by their personal experiences and knowledge, accumulated in a particular cultural context (Tversky & Kahneman, 1974). In an example presented by Van Veldhuizen (2017), an asylum seeker claimed that he had blown up a flat truck tire with his mouth to cross the river on it. That story was initially judged as implausible, and his claim was denied. At the appeal hearing, he showed up with an empty truck tire and blew it up on the spot, which eventually led to the approval of his claim.

Religious and sexual orientation claims

The assessment of plausibility is likely to be especially problematic in specific identity claims (e.g. religious or sexual orientation claims), since they are heavily influenced by culture. For example, according to Selim et al. (2022) review of asylum claims based on religion, applicants whose religious beliefs were unfamiliar to asylum officials (e.g. fear of witchcraft) faced a greater risk of being found noncredible. Selim and colleagues highlighted that such implicit assumptions on what constitutes religion are formed by asylum officials’ own cultural background (see also Kagan, 2010).

In 2020, a Dutch newspaper, NRC, published a critical article about the functioning of the Dutch Immigration and Naturalisation Service (IND) (Rengers & Kuiper, 2020). Particularly striking was the section on inappropriate questions (‘ongepaste vragen’). According to Rengers and Kuipers Dutch immigration officials ask asylum seekers inappropriate questions, for example about same-gender sexual behaviour. In the article, a Dutch asylum lawyer provided an example of such questions: asylum seekers have been asked whom they had slept with and what sexual acts had been performed. Such intimate and sexually explicit questions may lead to short and vague answers due to discomfort and shame (Jansen & Spijkerboer, 2011; Selim et al., 2022). In their literature review on credibility assessment in asylum claims on sexual orientation, Selim et al. (2022) discuss the problematic nature of assumptions made by asylum officials regarding sexuality. These assumptions often reflect the officials’ cultural background, while applicants with different cultural backgrounds may manifest their sexual orientation in different ways. For example, Berg and Millbank (2009) describe an asylum seeker who kept referring to his sexual orientation as ‘his problem’. That was taken as evidence that the applicant was actually not gay.

Bias correction tools

Subjectivity and intuitive judgements are likely to affect any investigation. For example, Hill et al. (2008) found that police officers’ intuitive beliefs of guilt affected their interviewing strategy as well as their evaluation of evidence. Similarly, Van Veldhuizen et al. (2017) concluded that an initial snap judgment of plausibility regarding the asylum story predicted the final decision of the immigration officials. That points to the presence of confirmation bias in the asylum determination process, as has also been found in criminal investigations (e.g. Ask & Granhag, 2005; Hill et al., 2008; Kassin et al., 2003; Nickerson, 1998). We argue that the development and use of bias correction tools, such as the alternative scenario method, would contribute to a more objective assessment in the asylum context. The alternative scenario method, for instance, invites immigration officials to devise, assess and compare at least two scenarios. In this way, they must always construct alternative explanations for their findings, which may disconfirm their first impression. In addition, the officials must actively and explicitly assess to what extent the findings discriminate between the scenarios (Van Koppen & Mackor, 2020; Van Veldhuizen, 2017; Vredeveldt et al., 2022).

In the asylum determination procedure, the main question is whether the applicant has a genuine fear of persecution in the country of origin. In an origin claim, the two scenarios would thus be that the applicant is from country X or the applicant is not from country X. If the applicant cannot identify, for example, a water well located in their claimed town of origin from a picture, one explanation is that the asylum seeker is lying about their origin, but there are also several alternative explanations to explore. For instance, one scenario could be that the applicant is familiar with the water well but has difficulties interpreting a two-dimensional representation (i.e. the picture). Another scenario could be that the applicant cannot identify the water well because women collect the water in their community. Alternatively, the applicant does recognise the water well but does not feel comfortable identifying it because it would be an inappropriate place for men to be familiar with (cf. Van Veldhuizen, 2017, p. 193). With the possible scenarios in mind, the immigration official can now ask additional questions to examine the probability of each.

In sum, plausibility as credibility indicator is problematic because of its subjective nature, which is exacerbated in the cross-cultural context of the asylum determination procedure. It is therefore important to assess possible bias correction tools, such as the alternative scenario method. Finally, immigration officials should keep an open mind and ask for clarification when in doubt, especially in plausibility judgements.

Conclusion

The asylum context is a perfect example of a setting where memory, culture and credibility assessment come together. There is a growing body of literature on cultural differences in eyewitness memory and on psychological assumptions that are relevant for the asylum determination process. We have integrated the findings by discussing relevant literature on aspects of memory statements that vary across cultures through the lens of credibility assessments in the asylum determination procedure.

In conclusion, we highlight three observations. First, while we applaud cross-cultural studies, we note that many researchers have singled out one of the cultural dimensions (e.g. power distance or collectivism) and then explain any differences found based only on that cultural dimension. For example, the finding that group A (e.g. from a sub-Saharan African country) reports fewer details than group B (e.g. from a Western European country) has been explained by collectivism in studies focusing on that topic, but the exact same finding has been explained by power distance or high-context communication in studies focusing on those topics. To gain more insight into the underlying factors of cultural differences in memory statements, we recommend that future studies employ a full cross-over design comparing countries that score differently on multiple cultural dimensions (e.g. Country 1 with high collectivism and high power distance, Country 2 with high collectivism and low power distance, Country 3 with low collectivism and high power distance, and Country 4 with low collectivism and low power distance).

Second, using nation-level dimensions to account for individual-level behaviour has been criticised, and there are serious doubts about the validity of the measurements to capture those dimensions on an individual level (for an overview see, Sharma, 2010). Some efforts have been made in developing more reliable and valid individual-levels scales for Hofstede’s dimensions (Erdem et al., 2006; Sharma, 2010; Yoo et al., 2011). The validity of these recently developed individual-level scales will need to be tested in the context of asylum procedures to explore their usefulness.

Third, a bird's eye view of the body of literature discussed in this article suggests that some reported cultural differences may generalise across wider populations, whereas other cultural differences seem highly specific to the particular context. An example of a more generalisable and perhaps systematic difference is the finding that people from more rural societies, who are not as frequently exposed to maps and pictures, typically have more difficulty transforming two-dimensional representations into three-dimensional representations. An example of a more specific and perhaps idiosyncratic difference is that in some cultures, sunset marks the beginning of a new day. Both systematic and idiosyncratic differences need to be incorporated in practical guidelines for investigative interviewing, but this may be more difficult to achieve for idiosyncratic differences. We recommend developing open-source cultural background documents for investigative interviews, preferably per region, which cover as many systematic and idiosyncratic differences as possible. The development of such cultural background documents requires a continuous effort from scholars to gain more specific insights into the particularities of specific cultural groups. At the same time, it should be acknowledged that a large degree of intra-cultural variability exists as well.

To conclude, we make an urgent call for researchers to investigate the influence of culture on memory statements more precisely, by moving beyond the simple comparison of collectivistic and individualistic societies toward a detailed mapping of systematic and idiosyncratic aspects of the region under study. We recommend that researchers work closely with practitioners to map out their needs regarding cultural background documents per context (e.g. asylum determination procedures, international criminal cases, local police interviewing). Without considering cultural differences, it is frankly impossible to properly interview or assess an asylum applicant.

Acknowledgements

We thank Krista King for her assistance with organising our literature database and providing feedback on previous versions of the article.

Disclosure statement

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

Additional information

Funding

The authors have received funding from the European Research Council (ERC) under the European Union’s Horizon 2020 research and innovation programme (grant agreement no 802080).