Putting Memory in Its Place

Different tasks handled in separate brain locations

By William J. Cromie

Gazette Staff

Think of your memory as a huge library. Bits and pieces of recollections are stored among billions of cells scattered through your head like millions of books on thousands of shelves.

A library like that needs an index and a librarian to help you find what you're looking for. Harvard scientists believe they have found the places in the brain where such activities go on.

They have also traced the origin of names and thoughts that suddenly and unintentionally pop into your head.

"Our research reveals at least two components to memory," says Daniel Schacter, professor of psychology. "One is the actual recollection. The other is the effort you make trying to remember, whether you recall anything or not."

Brain scans of eight healthy people, done by Schacter and his colleagues at the Medical School, show that the effort to remember stimulates activity in the front of the brain, particularly on the right side.

Actually recalling past events "lights up" a small area in the middle of the brain between the temples. This region is centered around the hippocampus. The brain is roughly the size of a grapefruit, and the hippocampus is about the size of a small seed in the shape of a tiny sea horse.

The hippocampal region boasts nerve connections to the cortex, the wrinkled gray topping of the brain where memory is stored and learning takes place -- the thinking cap. "Many of us believe this region acts like an index," Schacter explains. "It retrieves sights, sounds, and words located in different parts of the cortex and pulls them together into a memory."

When an accident or disease destroys the hippocampus and its surrounding tissue, people no longer can remember anything new, but they recall events before the hippocampal light went out.

Evidently, two separate memory systems exist. Schacter showed this dramatically when he taught a woman whose hippocampal region was ravaged by encephalitis how to enter data from documents into a computer.

To do this, Schacter took advantage of nonconscious memory, the part that is gone but not forgotten. Called implicit memory, this system is responsible for those images and ideas that occasionally pop into your head for no apparent reason.

Brain scans reveal that such memories, which bypass the hippocampus, often come from the back of the head, the area over the back of you neck. This is the brain's visual center and "mind's eye," so those unintentional recollections often take the form of pictures.

Explicit vs. Implicit

Schacter and colleagues recently completed a series of experiments in which they managed a tricky feat, separating implicit and explicit memories. "Explicit" refers to conscious memories routed through the hippocampal index.

People in the study did different types of intellectual tasks with their heads in a scanner that reveals which areas of the brain are most active during the tasks. They learned a list of words, then later were asked to recall the words. Memories of the words caused blood to flow into the hippocampal area, particularly the right side, evidence that this is a seat of recollections.

In another test, subjects saw words and had to determine how many "t" junctions they saw, that is, how many times lines crossed at right angles. "X" provides another example of such a junction.

When researchers tested the latter group on recall of words rather than junctions, they remembered few words. At the same time, their brain scans showed increased activity in the front part of their brains, the site of effort to remember. When subjects managed to retrieve a word, the scans revealed increased blood flow to the visually oriented back of the brain.

What words they did recognize probably came from nonconscious memory because the participants had not actually learned the words. Such implicit memories are put on the shelves of the cerebral library but not recorded in the hippocampal index.

Schacter's team retrieves these recollections by a process known as "priming." People who took the implicit test searched a word such as "tablet" for "t" junctions. Later, researchers asked them to complete a three-letter string such as "tab----." Thirty to 40 percent of the time, they complete the word as "tablet," rather than "table," "tabby," or "taboo."

Although participants didn't explicitly learn the word "tablet," it registered in their brains unconsciously as they concentrated on finding "t" junctions.

"Our data support the idea that priming occurs independently of the hippocampal formation; instead it depends on other brain systems," Schacter explains. "Another conclusion is that the hippocampus is more concerned with the actual recollection of the past than with the effort involved in trying to remember events or people."

Possible vs. Impossible

In other experiments, conducted with colleagues at the University of Arizona, Schacter scanned the brains of people looking at drawings of possible and impossible three-dimensional objects. As you might guess, their memories for possible objects, even ones not seen before the experiment, were richer than for impossibles.

The hippocampal area lit up when a person remembered possible shapes they had seen before. No such light-going-on-in-the-brain occurred for impossibles, even those viewed previously. All the situations, however, registered in the brain's frontal region, over the eyes.

That kind of activity pattern indicates that decisions about possible or impossible, new or familiar, are made in the front of the head. The left side of the hippocampal region may be involved in identifying something as familiar or novel, and the right part in actually retrieving memories.

"The two studies provide a nice correlation between memories of words and those of objects," Schacter comments.

Other experimenters have discovered that the richer a memory, the greater the amount of blood flow to the hippocampal area.

All this evidence begs the conclusion that your thoughts go through the hippocampus to find a memory in the same way you go to a library index to find information. The hippocampal region puts your brain cells in touch with areas in the cortex where bits and pieces of your past life are scattered as words, scenes, or sounds.

When remembering requires a difficult or prolonged search, the frontal cortex gets involved and guides the effort.

"Implicit memories probably are stored in the same or similar brain subsystems as explicit memories," Schacter speculates. "The former may focus on isolated slivers of past perceptions not brought together into a full-blown recollection. Also, an implicit byte that pops to mind may fit into or trigger the construction of an explicit memory."

Schacter and others are now trying to uncover details of what goes on in the frontal cortea, two lobes of thinking tissue at the front of the brain. "Part of this region might be inhibiting activity in areas not concerned with a memory," he thinks. "Another part might be involved in guiding the search, yet another in deciding on alternative routes of memory pursuit."

Schacter also works with University of Arizona colleagues to distinguish between false and genuine memories. For example, are suddenly recalled rapes, robberies, and abuse real memories, or have they been implanted by suggestion? "Our team has induced false memories of words in a group of subjects, and we are now looking at what kind of activity shows up in the hippocampal and other areas," he says.

The false memories are induced by showing people words such as "bitter," "sour," "candy," and "chocolate." Later, researchers ask them if "sweet" appeared on the list. Those who reply "yes" have a false memory whose brain signature can be compared with that of words actually on the list.

Schacter also works with Medical School researchers Marilyn Albert, Nathaniel Alpert, Scott Rauch, and Cary Savage on how memory changes with age.

"In the elderly, we see hippocampal activity similar to that in young people," Schacter notes. "But clear differences appear in the frontal lobes, especially in areas concerned with language output. We interpret this to mean the brain is not as effective in its attempts at recollection. In other words, their memories are starting to fade."