By Tanisha Jhaveri
Think about a childhood story you have. There may be some details that you recall more vividly, while others are a little fuzzy. Now imagine trying to recount this story to someone. You will probably have to fill in some details that you don’t fully remember. For example, if the events took place in the morning, you might assume that you ate breakfast, despite not specifically remembering so. This subconscious process of introducing new pieces of information can alter your own recollection of the memory. Oddly, this can mean that the more often you reconstruct a memory, the less accurate it may become.
The psychological theory of reconstructive memory proposes that memory recall is an active process through which information is reconstructed. That is, when we remember things, we are actually rebuilding our memories from partial pieces of information stored in our brains.
In 1974, Loftus and Palmer conducted a two-part study to investigate the reliability of eyewitness testimony as it relates to reconstructive memory. In the first part of the study, they split 45 students into five groups and showed them videos of car crashes. After this, participants in one group were asked to recall the speed of the cars before they crashed: “At what speed were the cars travelling when they collided?” However, in the other groups the word “collided” was replaced by either “hit,” “smashed,” “bumped” or “contacted.” Interestingly, the stronger the verb used, the higher the average estimate of the speed was. For example, the group shown the word “smashed” on average estimated a speed of 40.5 mph, while the group shown “contacted” estimated the speed to be only 31.8 mph. The stark difference between these estimates strongly suggests that our memory can easily be distorted by factors that occur after the event, during the process of reconstruction.
In the second part of the study, a different set of participants was divided into three groups, shown the same videos of car accidents, then given a questionnaire afterwards. This time, the first two groups were asked the same question about the speed of the car, but with the verbs “smashed” and “hit,” while the third group was not asked any questions as a control. Then, a week later, the participants were given another questionnaire in which they were asked whether they had seen any broken glass in the video. Although in reality, there was no broken glass, participants from the group that had the stronger verb “smashed” were more likely to recall seeing it. This fascinating result shows that factors after the event can not only distort memories, but also lead people to recall something that never actually happened!
These results have led us to reevaluate the reliability of eyewitness testimony, which, being reconstructed memories, can often be influenced by factors such as the phrasing of questions or individual biases. Twenty-five states in the United States have already implemented reforms to help improve the reliability of eyewitness testimony. Under some policies, for example, eyewitnesses are required to provide a statement articulating their level of confidence in their testimony. Double blind procedures are also used, in which neither the administrator nor the eyewitness know who the suspect is. This limits the presence of external factors that may distort the eyewitness’s recollection of events.
Reconstructive memory and the research conducted into it have given us valuable and maybe surprising insights into how our memory really works — and how fallible it can be. It has the potential to impact many fields, including law, medicine, and more.