Mar 14 2008

False Memories in the Courtroom and Elsewhere

Scientists and skeptics are keenly aware of the nature and limitations of human memory (or at least they should be). While so-called “believers”, and the public at large, generally put unjustified faith in the accuracy of memories, especially their own. This often constitutes the gulf that separates believers and skeptics on many issues.

Many people remember being abducted by aliens or seeing ghosts. Advocates of dubious medical treatments often site stories of people who were apparently cured by the treatments. The stories often seem very compelling, the facts all line up, the conclusion seems obvious. Yet skeptics will easily shake their head and say, “I just don’t believe it.” This frustrates the believers no end. How can the skeptics dismiss what so many people have experienced, they wonder.

The reason is simple – human memory is incredibly unreliable, and most people grossly underestimate the extent to which their own memories can be altered and fabricated.

The issue is of very practical importance in the courtroom. Eyewitness testimony is still commonly relied upon as key evidence in trial, including murder trials. This is despite the fact that for years there has been evidence from memory researchers that eyewitness testimony is unreliable. Now another assumption of the courtroom regarding memory has apparently fallen. New research suggests that the testimony of children may be more accurate than that of adults – the exact opposite of prior assumptions in the courtroom.

Researchers Valerie Reyna and Chuck Brainerd of Cornell University have conducted research that shows that children are less likely to produce false memories than adults. They have developed a theory of memory to explain this apparent paradox, a theory which is concordant with my reading of other neuroscience research in this area. They propose that that humans actually store two kinds of memory, which they call verbatum trace and gist trace. Verbatum trace memories are a more accurate and unaltered version of the details of what was experienced. Gist trace memories, however, focus on the meaning of the memories and not on the details.

Other researchers, studying true vs false memories with fMRI scanning, show that different parts of the brain are involved in retrieving detailed (verbatum) memories from emotional (gist) memories and also that there is different activation with true vs false memories in some parts of the brain but not others. What this means is that there is an underlying neuroanatomical correlation to what the memory experts are discovering clinically. Our brains store different types of memories that serve different purposes – one for detail, and another for emotion and meaning.

So why would children have more reliable memories than adults? Because their meaning-based gist memory is less well developed. Their brains are less cluttered with wisdom and meaning, so their memories are more concrete and therefore detail oriented. They are therefore less prone to false memories, because false memories begin with the desired meaning (a gist memory) and then backfill the details as necessary.

While I buy this explanation, I think it needs to be modified. In the 1980’s there was an epidemic of false child abuse cases brought to court, largely on the false testimony of children. Upon review of video-taped testimony it was determined that much of the testimony extracted from the children was coerced by interviewers who were either poorly trained or excessively eager. The children, in retrospect, were largely telling the adults what they thought the adults wanted to hear – even to the point that they made up elaborate tales about satanic cult abuse, cannibalism, and other atrocities. The difference is that the children did not necessarily believe the tales they were telling – they weren’t false memories they were just encouraged fantasies.

That aside, when not encouraged to fantasize, children give more verbatum-based testimony than adults. An adult is more likely to remember that a man accused of murder is a threatening criminal, and then fabricate entire details to support that gist memory – like that he brandished a weapon and made threatening comments. These false memories for details that our brains happily invent to support and amplify our meaning-based memories are indistinguishable to us from accurate memories.

What does all this mean? In the courtroom it means that there is further evidence to be suspicious of eyewitness testimony, especially if it is contradicted by objective evidence. Also, witnesses are not either lying or telling the truth – they may be honestly relating a false memory. It may be possible to encourage witnesses to access their verbatum memories rather than their gist memories – for example by asking emotionally neutral questions, and encouraging them to focus on details rather than emotions, and to put themselves in the events they are trying to recall.

For skeptics this is further evidence that memories can’t be trusted. It nicely helps explain why so many people “remember” a standard alien abduction scenario under hypnosis. They have or are being given the gist memory – the abduction – and then are encouraged to back fill the details. The brain happily complies – voila: a false memory.

But there are more subtle manifestations as well. As a physician I have learned to be highly skeptical of the details that patients give me as part of their medical history. I have learned to triangulate – ask multiple questions from different angles, questions that are designed to encourage memory for details and to appear neutral (at least during the history taking) regarding the meaning of the details. It is well known that patients will tend to simplify stories to enhance their apparent meaning, or that they will link the onset of symptoms to discrete events. Because we have an historical record (the medical chart) that we can compare the patient’s history to, physicians will typically see the process of false memory formation for themselves.

Here is a typical scenario: A patient reports to me that her neurological symptom began after she fell and hit her head six months ago. I will often then ask a detailed question without indicating why I am asking (although it is often obvious) such as – did you have this symptom last year at this time? Patients will often give me conflicting reports when I do this, so then I know I have to ask more questions in order to fill in the real picture. Or, I may read her records and see that she complained of the symptoms two years ago to her primary doctor, despite now claiming that they did not occur until she hit her head six months ago. The patient is not lying, the details of their memory have simply been altered subconsciously to enhance the meaning of a gist memory – perhaps that they were doing well, without any problems, until they had the injury.

A similar thing can happen in response to a treatment the patient believes helped them. They may report that their symptoms were continuous and unremitting, and then shortly after they took the miracle treatment their symptoms completely resolved. However, it’s possible that their symptoms come and go at random and have no real relationship to the treatment they took, or perhaps the symptoms were going away on their own before they started the treatment. But the patient has a gist memory – this treatment cured me of my symptoms – and the details of her memory morph over time to enhance the meaning of that memory.

This is one reason (among others) that anecdotal evidence is not reliable. Scientific data is recorded as objectively as possible under controlled conditions and in detail. Anecdotes often rely upon memory -and the reason people remember and tell others their anecdotes is because they have meaning to them, and that meaning trumps the accuracy of the details.

As the psychology and neuroscience of memory advances it is moving more and more in the direction of concluding that memories are malleable, they can be fabricated, and human memories appeared to have evolved more to tell us a meaningful story than to faithfully record details. Hopefully this message can be beaten into the public consciousness. As the ancient Greeks said – “know thyself.” Humans would do well to understand the nature of being human. It will help us administer appropriate justice in the courtroom. It helps scientists conduct better research. It helps clinicians take more accurate medical histories. It can help everyone in their everyday lives.

And it can help skeptics explain why we are so skeptical.

23 responses so far