“Who you gonna believe, me or your lying eyes?” Eyewitness testimony, trauma, and assessing ‘credibility’ of refugees

Anyone who has had anything traumatic happen to them may have difficulty articulating it. “It all happened so fast,” some might say. “It was all a blur.” It may become difficult to remember exactly what happened; or, the actual memories may be flashes of sensations: images, sounds, tastes, feelings, impressions. It may be difficult to remember exactly what happened, and it may be difficult to describe to someone else.

Eyewitness testimony has been proven to be quite unreliable in terms of presenting a true version of accounts. Trauma also impacts how we form memories and how we relate them, including what details we remember. This has huge implications for asylum-seekers, who are often asked to give detailed, precise accounts of their persecution, and are then considered “not credible” if there are exaggerations, discrepancies, or incoherency in their accounts.

Limitations of eyewitness testimony

BBC did a series called Eyewitness, where they stage a fictional crime in front of a pub-full of test subjects, whom they then ask for eyewitness testimony as to the circumstances of the crime and the murderer. As it turns out, it’s harder than it seems to correctly identify the assailant and describe the crime. Take a look:

There is are two interesting TED talks that discuss the limitations of eyewitness testimony, and also how we can manage to form false memories – remember things that did not happen, or remember things differently than how they actually happened.

As Elizabeth Loftus, a psychologist who worked on a the case of a man who was wrongly convicted of rape based on the faulty memory of the victim, describes it, “Many people believe that memory works like a recording device, you record the information, then call it up and play it back when you want to answer questions or identify images. But decades of work in psychology have shown that this just isn’t true. Our memories are constructive, are reconstructive. It works more like a Wikipedia page: you can go and change it – but so can other people.”

Leading questions, interrogations and torture

Refugee status is determined by a process that involves an in-depth interview about the person’s circumstances – how and why and when they fled their country, and why they may not want to return. Yet, the way the questions are phrased has a substantial impact on the answer received. Leading questions, which suggest to the interviewee the answer that the interviewer expects, or plants in their mind a certain thought which taints the validity of the answer.

For example, in asking the question. “is this the man that you saw murder the shopkeeper?” the interviewer has suggested that (a) the culprit was a man; (b) the witness definitively saw them; (c) the police believe him to be the culprit; (d) and the word ‘murder’ implies deliberation.

Another example of a leading question might be to ask a witness, “Which one of these people is the one you saw committing the crime?” you suggest to the witness that the guilty party is among the suspects. If you ask instead, “Do you recognize anyone from this lineup? The person who committed the crime may or may no be present.” Leading questions, discussion with other witnesses, and external information impact the content of the testimony.

False memories and suggestability

In an attempt to test the reliability of eyewitness testimony, and susceptibility to suggestion, Loftus and Palmer did a study in 1975 on about 400 participants who watched videos of fast-moving events, and established that leading questions can lead witnesses to describe scenes that they did not see.

In one study, researchers studied a military training exercise where participants, all members of the military, were subjected to a kidnapping/hostage situation (fun fact: they have these trainings for humanitarian workers too!). The participants were aggressively interrogated for 30 minutes and then were asked to identify who had been the person interrogating them. Researchers found that giving suggestive information about one or another of the suspects results in the “victims” mistakenly identifying someone who doesn’t even physically resemble the actual perpetrator – in 85% of the cases!

Describing Lofus’ work the article notes,

“The study of misinformation and false memories have consistently shown that human beings are highly susceptible to suggestion. Much of the work has focused on creating or changing people’s memories of the past. Loftus gave several humorous examples of memories that her team has been able to plant in substantial portions of the people in their studies, including convincing people that they had gotten sick from eating strawberry ice cream or the Pluto character at Disneyland had licked “their ears disturbingly and uncomfortably” when they were young children.”

Some things to keep in mind with respect to how memories work:

  • Reconstructive memory: people store information in the way that makes the most sense to them. We try to fit what we remember with what we really know and understand about the world.  As a result, we quite often change our memories so they become more sensible to us, and fill any missing bits with something that makes sense to us, or allows us to make sense of the event – which means that we can remember things that did not actually happen.
  • Time: Memories change over time. They naturally decay over time, they can become distorted, or they may become more vivit. People remember more details with repeated recalls, and also with new detail added.It can be extremely difficult to recall accurately separate incidents that were repeated.
  • Weapon Focus: In events where there is a weapon involved, people tend to focus on the weapon itself to the exclusion of other details.
  • Specific dates and facts: “Memory for facts is better than memory for the source of those facts, so asking someone how they came toknow a fact is not a good way of testing knowledge. Memory for temporal information such as dates, times, frequency, duration and sequence; proper names; verbatim of verbal exchanges; peripheral information;and the appearance of common objects are notoriously unreliable and may be difficult or impossible to  recall.” (source, p 58)
  • Anxiety / Stress: Impacts how people form memories and how they recall them.

As the 2013 report on assessing credibility, Beyond Proof, Credibility Assessment in EU Asylum Systems, notes,

“A wealth of research in the field of psychology reveals that there is a wide-ranging variability in a person’s ability to record, retain, and retrieve memories. Some people appear to recall memories more easily than others. Indeed, many people struggle to recall facts and memories of past events. Moreover, psychological research has consistently shown that memories of even the most important, traumatic, or recent life events can be difficult to retrieve and recall with any accuracy. Inconsistency, loss of detail, and gaps in recall are a natural phenomenon of the way a person records, stores, and retrieves memories.”

Trauma and memory

For refugees, who have left their home fleeing a fear of persecution or death, there is a substantial likelihood that they have experienced traumatic events, and particularly the events leading to flight may well have been traumatic.

Psychological literature indicates that memories of traumatic events differ significantly from normal memories. There is ample evidence that the need to cope with traumatic experiences affects memory. There is also a substantial body of research that demonstrates the effects of trauma on recall and behaviour.

The ECHO report gives a very comprehensive look on how trauma affects an applicant’s ability to relate their asylum claim. It’s worth reading the whole report, but particularly chapter 2. Here’s a very long excerpt:

 “Those who have suffered traumatic events often display avoidance symptoms; that is, they avoid thinking and talking about the event, and/or avoid situations that might trigger a recall. This is a normal survival strategy, which would need to be suppressed to facilitate disclosure of all relevant information in an asylum interview. As such, it may be extremely difficult, very distressing and potentially detrimental for the applicant to disclose such traumatic memories. Moreover, the applicant may not even be conscious that he or she is avoiding triggers or situations that could cause traumatic memories to recur. Avoidance may explain an applicant’s apparent refusal to answer a question, omission of relevant information from testimony, vagueness and apparent inconsistencies if relevant facts are recalled later in the asylum process.

Studies have also shown that applicants who have lived through traumatic events may experience dissociation.Dissociation can happen either at the time of the traumatic event or later when recalling it. Dissociation at the time of the traumatic event may hinder the person’s encoding of the event in memory. The applicant may experience dissociative amnesia – that is, an inability to remember some or all aspects of the trauma, because the event, or aspects of the event, was never initially encoded. Dissociation may be a reason why there is a lack of detail, vagueness, incoherence, or gaps in an applicant’s recall. Dissociation may also occur at the moment a person is asked to recall a traumatic event. The person may appear distracted and detached, and/or appear unwilling to cooperate. […]

Memories of traumatic experiences can be qualitatively different from other autobiographical memories. An applicant may have no memorized verbal narrative of the trauma that occurred, but only sensory impressions such as emotions, sensations, sounds, smells, or visual images like flashbacks and nightmares. Such memories are not evoked voluntarily, but they are provoked by triggers or reminders of the traumatic event. When triggered, the individual may relive an aspect of the experience as though it is occurring in the present. Therefore, an applicant who has experienced trauma may be unable to produce a coherent verbal narrative because none exists; this may mean only fragments or impressions of the experience may be related. Since sensory impressions are not evoked voluntarily, it is also possible that recall may be different in different interviews.”

Assessing credibility

Based on the above discussion, it might be fair to say that, “We can’t reliably discover true memories from false memories”. The goal of the interview and assessment, however, shouldn’t be to obtain the “truth”, rather to obtain the version of events as the applicant understands them with as little outside interference as possible.

Fear and lack of trust, cultural backgrounds, education, gender, age, sexual orientatoin or identity, stigma and shame, and other aspects of the applicant’s background may also impact the applicant’s ability to tell their story, or the way that they tell it. There is quite a lot that has been written on evaluating credibility, and there is not space here to recap. But some of the factors that adjudicators consider when assessing asylum claims, such as sufficiency of detail and specificity, internal and external consistency (consistency within the statement and also compared to the statements of family members and available external evidence), and plausibility can all be impacted by the intersection of trauma and memory. The kinds of inconsistencies and discrepancies that might lead an adjudicator to believe that the person might not be credible:

In a study of Kosovan and Bosnian refugees, researchers found that “Discrepancies between an individual’s accounts were common. For participants with high levels of post­traumatic stress, the number of discrepancies increased with length of time between interviews. More discrepancies occurred in details peripheral to the account than in details that were central to the account.
Conclusion The assumption that inconsistency of recall means that accounts have poor credibility is questionable. Discrepancies are likely to occur in repeated interviews. For refugees showing symptoms of high levels of post­traumatic stress, the length of the application process may also affect the number of discrepancies. Recall of details rated by the interviewee as peripheral to the account is more likely to be inconsistent than recall of details that are central to the account. Thus, such inconsistencies should not be relied on as indicating a lack of credibility..” (Source)

Distorting the evidence – accusatorial interrogation and torture

It has also been established that the style of the interview or interrogation can impact the results obtained. TV police procedural dramas like to portray “good cop/bad cop” or aggressive interrogation “techniques”. As it turns out, “accusatorial methods also significantly increased the likelihood of obtaining a false confession – a rather medium-to-large effect that is consistent with many cases of wrongful conviction in the United States.” (And about 75 percent of the wrongful convictions in the United States were convictions based on so-called eyewitness testimony). There is also mounting evidence to support the conclusion that evidence from torture is unreliable – people will say anything to stop torture – and much of the information obtained by torture not actionable, already provided prior to torture, or not reliable information. According to the CIA torture report, torture yielded false confessions and information previously obtained through alternative sources – and there is a lot of the scientific research supporting this conclusion.This confirms the assertion by a 2006 Intelligence Science Board report that very succinctly denied that torture was effective:

“The scientific community has never established that coercive interrogation methods are an effective means of obtaining reliable intelligence information. In essence, there seems to be an unsubstantiated assumption that ‘compliance’ carries the same connotation as ‘meaningful cooperation.’ ”

The case of Russell Williams, a former colonel of the Canadian Air Force who was a serial rapist and murderer who confessed after about four hours of interrogation, is a useful contrast to assertion that torture is needed to elicit confessions. Much of the video of the interrogation is available on youtube – below only an excerpt describing a couple of the tactics used by the interviewer.

Some sources

J Baxter, J Boon, C Marley, ‘Interrogative Pressure and Responses to Minimally Leading Questions’, Personality and Individual Difference, vol. 40, no. 1, 2006, p 87–98.

D Bögner, C Brewin, J Herlihy, ‘Refugees’ Experiences of Home Office Interviews: A Qualitative Study on the Disclosure of Sensitive Personal Information’, Journal of Ethnic and Migration Studies, vol. 36, no. 3, 2009, p 519–35.

D Bögner, J Herlihy, C Brewin, ‘Impact of Sexual Violence on Disclosure during Home Office Interviews’, British Journal of Psychiatry, vol. 191, no.1, 2007, p 75–81;

J Cohen, ‘Questions of Credibility: Omissions, Discrepancies and Errors of Recall in the Testimony of Asylum Seekers’, International Journal of Refugee Law, vol. 13, no. 3, 2001, p 293–309

K A Deffenbacher, S D Penrod, B H Bornstein and E K McGorty, A Meta-Analytic Review of the Effects of High Stress on Eyewitness Memory’, Law and Human Behaviour, vol. 28,
no. 6, 2004, p 687–706

M Eastmond, ‘Stories as Lived Experience: Narratives in Forced Migration Research’, Journal of Refugee Studies, vol. 20, no. 2, 2007, p 248-64;

H Evans Cameron, ‘Refugee Status Determinations and the Limits of Memory’, International Journal of Refugee Law, vol. 22, no. 4, 2010, p 469–511 at p. 506.

G Gisli, The Psychology of Interrogations, Confessions and Testimony, Chichester: J Wiley and Son, 1992;

J Herlihy, S Turner, ‘Asylum Claims and Memory of Trauma: Sharing our Knowledge’, The British Journal of Psychiatry, vol. 191, no. 1, 2007, p 3–4 at p. 3;

J Herlihy, S Turner, ‘The Psychology of Seeking Protection’, International Journal of Refugee Law, vol. 21, no. 2, 2009, p 171–92 at p. 181

E Loftus, ‘Planting Misinformation in the Human Mind: A 30-year Investigation of the Malleability of Memory’, Learning & Memory, vol. 12, no. 4, 2005, p 361–6.

T Valentine, J Mesout, ‘Eyewitness Identification under Stress in the London Dungeon’,
Applied Cognitive Psychology, vol. 23, no. 2, 2009, p 151–61;

B Tversky, E J Marsh, ‘Biased Retellings of Events Yield Biased Memories’, Cognitive Psychology, vol. 40, no. 1, 2000, p 1–38.



Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s