DISCOVER MAGAZINE
January 1997
When Memories Lie
Deciding whether memories are true or false is a
matter of considerable legal controversy. What is
beyond controversy, however, is how easily false
memories can be formed. Consider, for example,
these words: candy, cake, sugar, taste. Keep
those words in mind for a few minutes and then
ask yourself: Was sweet one of the words on the
list? The chances are pretty good that you'll
recall--falsely--that yes, indeed, sweet was one
of the words on the list. What, then, is the
difference between remembering cake and
remembering sweet? In trying to answer that
question, a group of Harvard researchers last
August captured the first glimpse of the brain as
it retrieves both true and false memories.
First, several lists of semantically related
words (for example, candy, cake, sugar, taste)
were read aloud to 12 adult subjects. Ten minutes
later, the subjects were shown words from each of
the lists and asked if they had been among the
words the subjects had heard. As they were
thinking, their brain activity was monitored by
PET scans. (This technique detects areas where
increased blood flow--and presumably increased
mental activity--is occurring.) Then the subjects
were shown another list, containing only words
similar--but not identical--to the words on the
original lists (for example, sweet, frosting,
sticky). Any recognition of words on this list
would constitute a false memory. Again the
researchers monitored brain activity.
The subjects were slightly better at recognizing
words they had indeed heard than at recognizing
that they had not heard other words. What
intrigued the researchers, though, were the
patterns of brain activity during the two tests.
Both tests showed activation in the medial
temporal lobe, a region deep within the brain
that has been shown to be involved in forming
memories of recent events. But there was also a
telling difference. In the first test, when
subjects were asked to recall words they had
actually heard, an additional region--high in the
temporal lobe and close to the surface of the
brain--was active.
This region is thought to process sounds, says
Daniel Schacter, a cognitive neuroscientist at
Harvard, and its activity may reflect the
auditory memory of how the words sounded when
they were read aloud. In contrast, a memory of
sensory detail wouldn't exist for words on the
"false" word list, because the subjects hadn't
been read those words. Past studies have shown,
not too surprisingly, that true memories tend to
be supported by more physical and sensory details
than false memories. While subjects were
pondering whether they'd heard a "false" word
before, Schacter notes, their brains sometimes
showed activity in the frontal cerebral
cortex--the decision-making center--as if they
were frantically searching for sensory evidence.
The wild-goose chase probably originates in the
medial temporal lobe, which seems to be active
whenever we try to recall something, and which
may, Schacter suspects, be the source of the
human predisposition to false memories. Its job
may be to store or retrieve associations--among
the various sensations of a single experience,
say, or among words that usually go together--and
then tell us where in the higher regions of the
brain to look for the associated bits. That
allows us to reconstruct a package of remembered
information when presented with one piece. It may
also prompt us to look for and even find memories
that aren't really there, because they are
similar to ones that are. "When someone is saying
yes to false items like sweet," says Schacter,
"they're probably thinking that really feels
familiar--there were words like sugar on the
list. And ultimately the memory for the gist
carries the day."
PET scans are not detailed enough to reveal the
differences in activity between a brain that is
falsely remembering a word and one that is
correctly recognizing that the word is new. And
they are certainly not likely to serve as some
kind of false-memory detector for therapists or
criminal justice officials--who tend to be
interested in events that happened more than a
few minutes ago. "We don't know if we would even
see this activity if we asked our subjects to
remember the words several days later," Schacter
says. "The best way to think about this study is
as a first foot in the door for building an
understanding of the biology of illusory
memories."
--Sarah Richardson