Information, misinformation and memory formation

Information, misinformation and memory formation

“We are our memories” it is often said. Remembering is something each of us does on a daily basis in several ways and for several reasons. We can “remember” things we have to do on a particular day. Or we can remember past events that either involve us or not.

Today we will tackle cases of remembering past events in which we are not personally involved but in which our remembering may be quite important. Such a case is the case of witnessing. Witnessing has particular importance in the juridical domain, when people are asked to give their description of an event that they assisted, in order to resolve in court “what really happened” during this event. What witnesses attest is really crucial, for their claims play a role in the final decision of the court. But how accurate are the attestations of witnesses? And how accurate can they be?

Elisabeth Loftus and her colleagues have run a long series of studies, for over 30 years now, showing the ease with which the witnesses’ memories can be influenced and crucially distorted. In a set of psychological experiments (Loftus, 1975) she asked university students to watch short film clips of a specific incident and then asked them several questions about this incident. The goal was to see how asking specific questions can modify and distort people’s memories of the event they had watched.

In one study (Loftus, 1975, experiment 2) the clip was an extract from a film displayingeight demonstrators entering into a class and disrupting it. After a noisy confrontation the demonstrators finally left the classroom. The participants of the study were split in two groups. Both groups received a questionnaire on the incident after they had watched the clip. And there was only one question in those questionnaires that was different for the two groups. The first group was asked “Was the leader of the four demonstrators who entered the classroom a male?” The second group was asked “Was the leader of the twelve demonstrators who entered the classroom a male?”  As you may have noticed, there is a particularity about these two questions; that is they presuppose something that is wrong. Presupposition is a very interesting mechanism on which we are very often based when we use language. It consists in the fact that in linguistic communication we implicitly stick to conditions that must hold for our statements to be appropriate in a specific context, and thus make sense. To make things clearer, imagine that I tell you “The past two years that I have been living in Brussels have been magnificent!”. In this case I presuppose that I am living  in Brussels and, that I have been doing so for two years. If for the past two years I have lived in Berlin, then my statement does not make sense and is obviously false. Now, importantly, presuppositions also play a role when we ask questions. When I ask you what is the name of your cat, then I presuppose that you have a cat. If you don’t, my presupposition is false in this context and my question does not make sense if addressed to you. Coming back to the two questions that the students in Loftus’ experiment were asked, they are both based on a false presupposition: the first one presupposes that the demonstrators were four, and the second one that they were twelve. The actual demonstrators were eight! So let’s see, what these false presuppositions can do on people’s memory of the incident they have watched with their own eyes. When participants came back to the psychology lab one week later they were asked again multiple questions concerning the incident they had watched, among which a simple one: “How many demonstrators did you see entering the classroom?”. All participants had watched the same clip –meaning that they had seen eight demonstrators. Yet, surprisingly, the two groups had been influenced by this one question with the false presupposition they had been asked one week ago. People asked about twelve demonstrators answered on average that there were 8.85 demonstrators, while people asked about four demonstrators answered that there were 6.40 demonstrators. This difference may look small, but statistical analyses show that it is significant. This means that it can not be due to chance, but that it is an effect of the different question that the two groups had been asked after watching the incident.

These results are interesting because they show how our memory of an incident we have witnessed with our own eyes can be influenced by information that we receive after we have assisted at the event. The most intriguing element, is the way in which this information can be implanted in people’s minds. Participants in the above-described study were not directly told that the number of demonstrators in the clip they had watched was four or twelve. This information was indirectly suggested to them just because they were asked a question that could not hold if the false presupposed information (the false number of demonstrators) was not correct. And this indirectly suggested information was somehow integrated in their representation of the event. This, in turn, shows the very subtle ways in which our memories and thoughts can be influenced and altered, crucially ways that we cannot easily detect. We are not talking about a case in which someone witnesses an incident and  after having assisted at that incident he meets another witness who explicitly claims that some details of the incident are different. We are talking about information that has been indirectly suggested to the subjects and which has the power to alter their initial memory of the incident.

Now you may say that this is not so big deal: the students’ memories were a bit biased in estimating the number of demonstrators by the fact that they were asked a subsequent question misleading them in this respect. But what if such subsequent questions can make us see things we never actually saw? This is what another study by Loftus (1975, experiment 3) shows. In this study, participants saw a clip about a car accident and were then asked ten questions about this accident. As in the study described above, the participants were split in groups of two, and each group was asked one question about a white car involved in the accident that was different in the two groups. The first one was asked “How fast was the white sports car going while traveling along the country road?” but the second group was asked “How fast was the white sports car going when it passed the barn while traveling along the country road?” This second question, was based on a wrong presupposition, namely there was no barn appearing in the film whatsoever. Participants of both groups filled-in a second questionnaire one week later, asking them among others “Did you see a barn?” Only 2.7% of the first group answered “yes” to that question, while a 17.3% of the second group that was asked the question presupposing the existence of a barn reported having seen it. Although 17% is not a big percentage in itself, again statistical analyses showed that this difference is statistically significant: it is big enough to permit us to infer that the second group was influenced in their answer by the false presupposition. This result corroborates and extends the finding of the previous study. Here participants were not simply biased in their answers by a false presupposition, but were further lead to remember having seen a barn they actually had not. This effect of a false presupposition on people’s memories is even more consequential. Imagine for example this phenomenon occurring in a real-world situation where witnesses are asked whether they have seen the perpetrator’s weapon, in a case where the perpetrator did not actually have a weapon.  The results of the study just outlined suggest that this question has the potential to make witnesses more likely to report having seen a weapon, while the poor aggressor had not at all premeditated his action! The consequences for the perpetrators sentence become obvious, highlighting the importance of the malleability of memory in real-life contexts.

As to a psychological aspect of the results of these studies, they are also interesting in showing how our minds work. Participants watched an incident and they afterwards received information that was implicitly transmitted to them, through presuppositions. Crucially, these presuppositions seem able to enter their memorial representations of the events that were initially formed based on a visual stimulus, i.e. the clip they watched. So on the one hand, we have the visual representation that their mind formed once while they watched the incident. And on the other hand, we have information that is contained in a linguistic stimulus that they process (the questions they were asked after having watched the clip) which is integrated in the visually formed initial representation. This process suggests how our mind can integrate pieces of information that is linguistically transmitted to us, with information that we acquire visually. For one, the ease with which such presuppositions can change our visual memories show the power of language and the subtle ways it can be used to influence our view of the world. The implications for real life phenomena are, I believe, obvious…


Loftus, E. (1975). Leading questions and the eyewitness report. Cognitive Psychology, 7, 560-572.

Myrto Pantazi

Postdoctoral Fellow

Postdoctoral Fellow

Leave a Reply

Your email address will not be published.