Memory is fickle. Sometimes, people have great confidence in the
accuracy of a particular memory...even if it can be proven that the event
they claim to remember never actually happened. For example, a
psychologist named Ulric Neisser likes to tell a story about one of his
false memories:
Pearl Harbor was bombed on the day before Neisser's 13th birthday. On
that day, he was sitting in his living room listening to a baseball game
on the radio. Suddenly, in the middle of the game--right in the middle of
a play--the broadcast broke off and a reporter came on the air with an
announcement: Pearl Harbor has been bombed. The United States is at
war with Japan. Then Neisser sprinted upstairs to tell his mother the
news.
Neisser vividly remembers the fear and the
excitement that he felt then. He remembers the room clearly, he says, even
though he only lived in the house for a year. He remembers what the radio
looked like. He even remembered which teams were playing that day. The
details were so clear and the emotions so strong that he had absolute
confidence in this memory for over 40 years.
Then one day he realized the problem. Pearl Harbor was bombed on
December 7. There's no baseball in December. The event could not
have happened as he remembered it. Nonetheless, he still remembers it
that way, even though he knows intellectually that it's wrong.
Of course, anecdotes, even multiple
anecdotes, aren't enough. A lot of people have done systematic research
on this phenomenon, though. For example, Beth Loftus, a psychologist who does a lot of work in this field,
showed a bunch of people a videotape of a car accident. She asked
them one of two questions:
- Where was the old lady standing when the cars
collided?
- Where was the old lady standing when the cars
smashed?
Then she asked them if there had been any broken glass in the picture.
(There hadn't been.) Subjects who received question 1 generally got this
question right. Subjects who received question 2 almost always got it wrong--and they had great confidence that there had in fact been broken
glass.
All humans can have false memories, but kids are particularly
vulnerable. In the McMartin Preschool Trial, for example, kids were
induced by an overly aggressive and incompetent psychologist to have false
memories of molestation. People then started wondering just how
accurately kids could report such experiences. Obviously, it's hard to
conduct a systematic study of abused kids, because different kids had
different experiences, and it's often difficult to figure out what really
went on. So a psychologist named Peter Ornstein came up with a brilliant
experiment. He decided to interview three-year-olds, five-year-olds, and
seven-year-olds about the Well Child Checkup--the annual physical exam
that most kids get. This checkup involves a genital exam, a blood
test (which requires a fingerstick), and a urinalysis (which requires
the kids to urinate into a cup). Molesters do similar things to their
victims, so this is as close as you can ethically get to the situation in which you're really interested.
Ornstein's research asked two questions: first, what would kids tell
you about these invasive events; and second, to what extent could they be
induced to tell you about something that didn't happen? He found
that three-year-olds would volunteer information about the urinalysis and
the blood test if you simply asked them to tell you what happened.
Seven-year-olds, on the other hand, wouldn't tell you about it unless you
asked directly. He also found, though, that you could get three-year-olds
to tell you just about anything, especially if you asked a leading
question ("He tapped your foot with a hammer, didn't he?").
Five-year-olds were in between. Seven-year-olds usually couldn't be persuaded tell you
anything that didn't happen. So he concluded that therapists and police
investigators need to ask direct questions, but not leading questions.
It's a tricky subject because the stakes are so high--you don't want to
let some scumbag get away with abuse, but you also don't want to ruin an innocent man's life by accusing him falsely. All the research I've seen
suggests that we shouldn't convict people on recovered memories
alone...unfortunately, sometimes no other evidence is available.