Brain Science: Early Childhood Testimony Unreliable

A neuroscience review casts doubt on early memories. We talk to the researcher.

Childhood memory should be considered carefully, according to a new study.

A commonly held belief that traumatic childhood memories are accurate  misleads judges and jurors in criminal cases, according to a scientific review released on Wednesday.

Neuroscience and behavioral research into memory cuts directly against confidence in eyewitness testimony expressed by police and jurors in surveys, writes psychologist Mark L. Howe of City University London.

"Memory often serves as the key or only evidence in the courtroom," Howe writes in Nature Reviews Neuroscience. Yet many police officers, judges, and jurors in North America and Europe "are naive when it comes to understanding how memories are formed, how they become distorted over time, and how stress and emotion affect remembering."

Until the age of eight or nine, most people don't have a sense of memory that is developed enough to reliably recall more than the bare outline of events, particularly stressful ones, the review concludes. (Also see "Memory: Remember This.")

Perhaps the most notorious criminal case involving issues of dubious childhood eyewitness testimony was the McMartin Preschool trial of the 1980s. The case  involved sexual abuse accusations at a Manhattan Beach, California, day-care facility.

All charges were dropped at the end of the trial in 1990. A psychologist suggested that memories of abuse had been implanted in the minds of children by investigators, a view that swayed most of the jurors.

Nevertheless, despite what research has shown about the malleability of memory, law enforcement officials are still alarmingly overconfident about testimony from childhood, Howe suggests.

Childhood Amnesia

Memory develops slowly through infancy, behavioral research shows, with most memories before 18 months completely lost to people later in life, a phenomenon known as "infantile amnesia." Memories remain sparse and lacking in detail until age eight. (Also see "Braingames: Use It Or Lose It".)

That means that honest courtroom testimony by adults about early childhood events would contain few details. A richly detailed recounting of events is more likely have been embroidered or created by the witness, perhaps unconsciously.

Shortly after events, "although young children are frequently correct in the basic facts of what happened, their narratives do not contain many of the additional details," Howe says. Witness testimony that includes specifics—the color of someone’s clothes or the weather on a particular day—is probably suspect. Even true memories become distorted over time.

The belief that memories of stressful events are retained more vividly than others also goes against what psychologists have found, says Howe. "Overall, the fact that an event was stressful or traumatic is not a good predictor of a child's subsequent memory for that event."

Still, more than half of judges, jurors, and police officers in a 2006 survey agreed that "traumatic experiences can be repressed for many years and then recovered." Less than a quarter of memory experts agreed.

Malleable Memories

Overall, Howe writes that our everyday picture of memories as fixed and stable is wrong. Such beliefs about memory are fundamental, however, to the whole idea of eyewitness testimony in courtrooms. Partly for that reason, psychologists and neuroscientists have become increasingly critical of eyewitness testimony in the last decade.

National Geographic asked Howe to elaborate on some of the points he made in the review, which is part of a series on neuroscience and the law presented by the science journal. His written responses, lightly edited for clarity, follow:

How big a problem is it that adult witnesses are recounting childhood memories as testimony in the courtroom? Are many cases decided by eyewitness testimony of this type dubious?

All cases involving historic childhood sexual abuse involve memory evidence, and where there is no other corroborating evidence (e.g., medical records) it serves as the only evidence. So all cases of historic childhood sexual abuse are decided, in whole or in part, on memory evidence.

The issue, as discussed in the article, is the age at which memories were formed for the alleged abuse, the type of information contained in the complainant's narrative, the length of time this information has been in memory, and the intervening experiences that may have influenced memory during this interval.

What are the brain processes that make this kind of eyewitness testimony most unreliable? Is it the storage of memories? Or is it their retrieval? Or some combination?

It is a combination and has to do with the age at which information was encoded, how well that information can be stored, and what happens to that information during the decades of retention prior to testifying.

Do you see the remedy in educating law enforcement about the limits of this testimony? Or should there be policies limiting its use?

I view this as more of an education issue. When memory is the main or only evidence, those involved in the justice system (e.g., police and triers of fact) need to know how memory works because their naive beliefs about how it works are frequently wrong.

What do you see as the telltale signs of dubious testimony?

Because there is no litmus test discriminating true and false memories, and because decades-old memories encoded when we were children are often fragmented and decontextualized, whenever there is a worry about the extent of detail (particularly for peripheral information) and the content (particularly if there are concepts there that would not be known by a child of that age) of a complainant's narrative, then there are serious concerns about whether such memories can be used to convict.

You make the point that reconsolidation and reconstruction of memories takes place unconsciously, distorting memory. What does this say about all of our fondest memories? Are they at all reliable if we’re constantly burnishing them?

Difficult to say for certain, but my guess is that we should treat many of these memories with a (rather large) grain of salt.

There is considerable evidence showing that memory does not always serve us well if what we are trying to do is recapture an accurate or verbatim glimpse of the past. Indeed, we can often misremember even the core of events that have happened to us.

Memory serves us best by abstracting the relevant meaning from our experiences, not the exact details of every experience, and this allows us to form a worldview, a lens through which we interpret our present and anticipate the future.

Follow Dan Vergano on Twitter.