Our memories are not always reliable. But sometimes they’re rich, textured and vivid — even if they didn’t happen. Research has suggested false memories often have the descriptive, multisensory elements of real memories, a fact that obviously poses both interesting questions about memory itself and difficulties for those relying on eyewitness encounters for evidence.
But beyond the question of how people remember is another quandary: are we, as observers, able to tell whether someone’s memory is true or false? It’s a question tackled by UCL’s Julia Shaw in a new study published in Frontiers in Psychology — and she finds that not only are we susceptible to having memories planted, we’re not very good at working out when someone else’s memory is false either.
Shaw has worked on false memory and our susceptibility to it before. In 2015, she co-authored a paper that claimed to have successfully implanted false memories of having committed a crime in 70% of participants; these crimes ranged from petty theft to the much more serious assault with a weapon. Videos of these participants recalling true and false memories formed the stimuli for the new work.
(It’s important to note that the findings from Shaw’s 2015 study have been subject to much debate, with some psychologists arguing that the coding scheme used in the study led to wildly inflated results. However, all videos of false memories included in this paper were also classified as false memories in a re-analysis by critics.)
In the first study, 124 participants saw videos of someone recalling both a real and a false memory of events that had (supposedly) happened during their adolescence. In one condition, participants saw the person recalling an emotive false memory (e.g. being attacked by an animal, being bullied or the death of a parent) while in the other condition, the false memory related to committing a crime (e.g. an assault with a weapon). Participants were told before watching the videos that “all, some, or none of the videos” involved memories of real accounts. Afterwards they were told one memory was real and one was false, and were asked to identify which was the false one.
In the second study, participants were shown the same videos — only this time, one set of participants watched the video with sound, one listened just to the audio, and a final set watched the video without audio.
The results suggest that participants are not better than chance at classifying false memories — about 57% of false memories were identified in the first study, and only 44% in the second. Numbers were similar across conditions: criminal false memories were identified 55% of the time in study one and 44% in study two. In study two, accuracy was highest when participants were given video with audio (53% accuracy) and worst with just audio (32%) .
Shaw argues that the study’s results demonstrate that false memories look real to an observer: even though participants knew that one memory was false and one was true, they were still unable to accurately tell the difference.
This could have serious implications for judges, police officers, lawyers and others involved with gathering evidence and interviewing eyewitnesses. Shaw points to judges in particular, who should “never assume they can tell when someone has a false memory … and should consider the entire process to see if there was any risk of contamination of a defendant or witness’ memories.”
The results should almost be beside the point — judges, lawyers and police officers should already try to assume they can’t always tell whether someone’s memory is accurate or not. So it will be interesting to see exactly how these findings are taken by legal and law enforcement officials. Memory, for better or worse, is a key part of evidence; being cautious before we make any firm conclusions about it may be vital to ensuring that evidence is as watertight as possible.