Fall 1996 Volume 5, Number 4
Decades of psychological research have shown that memory,
though usually accurate, is sometimes unreliable. Yet little is known about
what happens in the brain when people retrieve accurate and illusory memories.
To find out, Daniel Schacter of Harvard University and Eric Reiman of Arizona
State University tested 12 volunteers, using positron emission tomography (PET)
to measure blood flow in particular brain regions. The volunteers heard lists
of related words, such as candy, sour, sugar, bitter, good, taste, etc. A few
minutes later, the investigators took scans to see what brain areas were active
when the volunteers recognized words they had heard earlier and when they
claimed to recognize similar words, such as sweet, that they had not actually
heard.
Several brain areas were active during both true and false recognition, including a region near the left hippocampus. In contrast, an area in the temporo-parietal region that stores information about a word's sound showed a difference: This area was active when people recognized words from the original list; it was inactive when they falsely recognized a related word. Regions in the frontal lobes were especially active during false recognition, perhaps reflecting the volunteers' efforts to decide whether that word had been on the list.
The PET technique cannot be used as a "lie detector" in court cases involving memory-based testimony, as the results reflect group averages and investigators cannot generalize from the data. But it offers a new way to study many questions about the truthfulness of memory.