Problems with memory are at the core of many psychological disorders. For example, people suffering from both clinical-level depression and PTSD often have difficulty remembering details of specific memories, especially for happy experiences. The memory distress is called overgeneralized autobiographical memory (OGM). A therapist might ask a depressed person showing overgeneralized autobiographical memory (OGM) to recall a recent happy experience. The depressed person might answer, “When I was visiting my friends last weekend,” but then be unable to recall or describe any particular events or interactions during that visit that were enjoyable or rewarding. For another example, people suffering from OCD experience less confidence in the accuracy of memories they retrieve than people without the disorder. This uncertainty about memory can lead to obsessive thoughts about whether they turned off the stove or paid the electric bill when it was due. People with OCD also tend to show a bias to retrieve threatening memories. Nearly every major psychological disorder you will study in this course has some aspect of memory that is either a symptom or a process that maintains the disorder or more often both.

You may also have learned by now that remembering and thinking about past events—either recent or long ago—is the basis of most forms of psychotherapy. The psychodynamic therapy developed by Sigmund Freud is almost entirely based on remembering actual experiences or recent dreams. Even newer forms of therapy, like Cognitive-Behavioral Therapy (CBT), involve a great deal of memory work.

It may seem that research laboratories in universities and medical centers are a long way from psychotherapists’ offices, but professional therapists keep up with new developments in basic research and they often collaborate with researchers in bridging the gap between new theories and the application of those theories in the real world. A great example of the basic research-applied research connection is the development of therapies that can change the emotional impact of some memories without erasing or otherwise distorting them.

Memory Consolidation

Figure 1. Older theories on memory said that memories were stored like printed books, but new research suggests that they are not so set.

Until the early part of the 21st century, most people thought of memories—particularly memories of personal events, technically known as autobiographical memories—as mental representations that become relatively stable and unchangeable very quickly. We knew that memories do not stabilize immediately, however, because brain trauma (e.g., a concussion) or certain drugs could interfere with people’s ability to recall events immediately before the trauma or administration of the drug. The neural processes that occur between an experience and the stabilization of the memory for that experience is called consolidation. Consolidation is complex, with some consolidation processes taking minutes to hours and other consolidation processes taking weeks, months, or even years. For the rest of this reading, we will concern ourselves with the quick part of consolidation that occurs in the hours and days immediately after an experience.

The idea of consolidation does not rule out forgetting. Memories can fade—that is, lose details—or become impossible to retrieve. In the reading on memory, you also learned that misinformation that a person hears shortly after an event can be incorporated into the memory. But the idea is that the final version of the memory is fixed once it has consolidated within a few hours. This late-20th century theory says that memory is like a book. When it is first printed, the ink must dry (the consolidation process that takes up to a few hours), but when that has occurred, the contents of the book don’t change. The ink may fade over time or you may have trouble finding it in your library, but the contents of the book never change, no matter how often you pull it out to read it.

Reconsolidation

Around the beginning of the current century, our understanding of memory was shaken by new research, first in animal labs, but later with humans.^[1] The study that initially caught the attention of memory scientists was a study using rats as subjects by Karim Nader, Glenn Schafe, and Joseph Le Doux of New York University in the year 2000. The research taught their animals a fear memory by pairing a particular sound with a mild, but unpleasant shock using classical conditioning.^[2] The researchers found that they could change a memory that had already been consolidated if they did just the right things at just the right time.

Figure 2. Research on consolidation supports the idea that memory is saved somewhat like a computer file: the original file is there, but that file can be modified and re-saved.

What Nader and his colleagues found was that memories become open to changes for a brief period of time when they are retrieved. For a few hours, the changed memories can be disrupted (e.g., by trauma to the brain, by drugs, and by other means), but once they have reconsolidated, they become the new version of the memory. This newer theory of memory says that our memories are not really like books, which don’t change after the print has dried. Now memory is more like a computer file that is updated without saving the original. You originally create the memory (consolidation) and store it away. When you retrieve the memory, you can change some information in the file, but this new version now becomes the memory. Many researchers believe we don’t have a backup version of the original memory. All we have is the new, modified memory of the event.

Reconsolidation: In the Basic Research Lab

The theory of reconsolidation has changed the way we think about the stability and accuracy of memories, but a scientific theory must be more than interesting or novel: it must be supported by careful research. There is now an impressive body of research about reconsolidation. We have already mentioned experiments with rats by Karim Nader and his colleagues, but we will go into more detail on a study by Elizabeth Phelps, a highly respected psychologist who is one of the leaders on modern neuroscience of emotion and cognition. The study we will discuss is by Dr. Phelps, Daniella Schiller (now an associate professor of psychiatry at Mt. Sinai hospital in New York), and some of their colleagues.

You may remember learning about classical conditioning. Ivan Pavlov discovered how classical conditioning works when he trained dogs to salivate when they heard a bell (click HERE to review classical conditioning). Dr. Phelps and her colleagues classically conditioned volunteer research participants to fear a shock. They allowed this learning (i.e., the conditioned fear response) to consolidate, and then figured out the way to eliminate the fear response.

To start, we are going to look at what happened in one of the control conditions, which will give you an idea about what normally happens with this kind of fear learning.

Note: we used emoticons in the exercise above, but the actual dependent variable in the study was a physiological measure of fear: skin conductance. When we are scared, our sweat glands respond by producing sweat, sometimes a lot, sometimes a little, but always some. The perspiration on our skin changes the way that electricity moves across the skin, and these changes can be detected and measured, even if the changes are very subtle. This is the skin conductance response (called SCR). Detection of changes in skin conductance is simple, requiring only some detectors on your fingers, and it is painless.

Day 2 has been successful. The person is no longer afraid of the yellow box. But, we’re still not quite done. We need to test for spontaneous recovery. Let’s go to Day 3.

Spontaneous recovery is one of the big problems with extinction training. You can get rid of a response for a while, but the response can return over and over again. According to the researchers—Dr. Phelps and Dr. Schiller—the problem may be that the person has two memories: one where the yellow box means a shock is coming, and another that means the yellow box equates to no shock. These two memories are both available, so when a yellow box happens to retrieve the first memory (yellow box = shock), the fear response returns.

So how can we change the first memory without creating a new memory? Here is a second condition in the experiment. We’re going to call this group the “10-Minute Group,” and we’ll explain why shortly.

The first step involves the same process as in the control group and involves conditioning the subject to “fear” a yellow box.

Now let’s go to Day 2. Remember from the control group that Day 2 involves extinction, which is the process of unlearning the fear response. But for this new group, we’re going to try something different to see if we can replace their original memory without creating a new memory.

Memory Reactivation

This time, before we begin the process of extinction, we are going to get the person to think about the shock experience—that is, we want them to retrieve the full fear memory—before they start extinction. Once the full memory is reactivated, there is a 10-minute delay, and then the subjects go through the same extinction trials that the Control Group subjects experienced on Day 2.

This reintroduction of the yellow box on Day 2 is the one event that did not happen in the control condition you read about earlier. It turns out that this reactivation step is crucial to preventing spontaneous recovery.

After the extinction process has been completed on Day 2, the question is this: will the person show spontaneous recovery of the fear response on Day 3? If they do show spontaneous recovery, then our new procedure (reinstatement of the memory on Day 2) has failed to produce the change in memory that we hoped for.

The last step is to again test for spontaneous recovery.

So far, the experimenters have shown that fear can be learned (Day 1), extinguished (Day 2), and then spontaneously recover (Day 3) for the control condition. By contrast, the reactivation condition shows that, if the full memory is activated on Day 2 just before extinction, then the fear response does not spontaneously recover.

However, our journey is not quite complete. The experimenters claim that a reactivated memory acts like a new memory: it is open to change for only a brief time and then it becomes stable again. So the Day 2 extinction process should only work to change the original memory for a short while—at most, a few hours. If the memory is reactivated, but extinction is delayed for a few hours, then the memory should not be changed because it has had time to reconsolidate.

The final experiment tests this idea. The only difference between this new group and the last group is the time delay on the second day. Rather than waiting 10 minutes between reactivating the memory and extinction, the experimenters waited six hours. After six hours, the fear memory should no longer be active and extinction should not change the memory.

This experiment is important because it serves as a control to help us determine if “rewriting a memory” is actually the correct interpretation of the results. In this experiment, the memory is reactivated (just like in the 10-minute group), but the memory is then allowed to deactivate over a six-hour delay. If there is no spontaneous recovery in this condition, then rewriting memory is not a particularly convincing explanation for the results. If there is spontaneous recovery of fear, then the theory that we are actually rewriting a memory is more convincing.

So let’s see what happens.

The procedure on Day 3 is the same for all three groups, but the responses are different. Participants in the two control conditions (control group and six-hour group) both act the same: they both show spontaneous recovery of the fear response. Those in the reconsolidation treatment condition (the 10-minute group), however, show no spontaneous recovery of the fear response.

Interpreting Results

Let’s take another look at the results of the study by Schiller, Phelps, and their colleagues. The y-axis on the graph below shows the skin conductance response of the subjects. Higher values indicate higher levels of fear.^[3] You will be adjusting the lines, so move them up to indicate more fear and down to indicate less fear. The x-axis shows the end of Day 1, after successful fear conditioning, and the first trial on Day 3, when spontaneous recovery is being measured.

We have placed the circles for Day 1 in their correct positions. The fact that they sit high on the graph reflects the fact that all three groups of participants were successfully conditioned on Day 1 to fear the yellow box. The differences among the three lines are not statistically significant.^[4] Your task is to grab the circles on the right and move them to the appropriate positions for the results of the experiment. You can move them up or down or leave them where they are. When you have entered your solution, you can look at the actual results.

Remember, spontaneous recovery means that the person returns to the fear level they had learned earlier, on Day 1. No spontaneous recovery means that the fear response (high levels of skin conductance) had been eliminated. Lower fear is shown if the dots get closer to the x-axis.

Try It

Instructions: Click and drag the circles on the right (Day 3) to where you think they should be to reflect the results of the experiment. When you’re done, click the link below to see the actual results.

Click here to see the results.

The figure above shows the actual results from the experiment. The green line (control group) and the blue line (six-hour group) show slight declines in fear level, but not much. These two groups are not significantly different statistically on either Day 1 or Day 3. The fact that these two groups showed high levels of fear on Day 3 is consistent with spontaneous recovery of the fear response after extinction on Day 2.

The red line (10-minute group) drops dramatically from Day 1 to Day 3. This means that the fear these subjects learned on Day 1 and then had extinguished on Day 2 remained extinguished on Day 3. There was no spontaneous recovery of the fear response. These results are consistent with the idea that a learned fear response can either stay strong across several days (see the two control conditions) or it can be eliminated (see the reinstatement treatment condition) if new learning takes place under just the right conditions (i.e., while the fear memory is still active).

Keep in mind that one experiment doesn’t convince anyone—certainly not experienced scientists. But, when many similar experiments are conducted and they generally give consistent results, then scientists become increasingly confident that the results are not just due to chance, but that they are seeing something real. Go online (for example, use Google Scholar) and search for “memory reinstatement” and you will find many studies that are related to the one you have just studied. Together, these experiments suggest that memories can be altered. In fact, every time we retrieve a memory, it is possible that we alter details or emotional elements of the memory. Our memories may change across our lifetimes in profound ways.

What is the practical value of this research?

At the very end of the video, you heard Dr. Phelps (from an interview in 2009) explain the potential for turning this research into a useful procedure for therapists:

So, you know, at this point, how this works in the clinic is going to be all speculation. But what this data suggests might happen in the future is: if you come into the clinic with a fear-related disorder, like a phobia or PTSD, if we can understand how these memories are re-stored when they are retrieved, much as we did in this study, we then may be able to time our therapeutic interventions in such a way where we aren’t creating new learning that’s overriding those earlier memories but actually rewriting them, in a sense. If we can time that correctly so we can target these mechanisms, perhaps we’d have a more effective, long-lasting outcome.

One of the goals of this research, then, is to give therapists a way of working with memory disorders. Of course, rather than creating a fear as the researchers did, therapists work with people who experience debilitating fear-related memories that came from experiences, often traumatic ones, in their lives. The therapist’s job is to help the person overcome the disabling experiences of fear. In most cases, they would like to reduce the emotional impact of the experience, which is part of the memory itself, without actually changing the facts that are remembered.

This application of reconsolidation theory to therapy is already underway. Here are the basic steps in this therapy:

• REINSTATEMENT: Have the person retrieve the memory. Be sure that the retrieval is emotionally powerful. If the person avoids fully reactivating the memory in its complete painful form, then reduction of the emotional impact will be impossible. The emotion may be fear or anxiety or some other strong negative response.
• REDUCTION OF EMOTIONAL IMPACT: While the memory is active and painful, the therapist acts to reduce its impact. There are two approaches to this, using the example of a phobia (irrational fear) to illustrate the method:
  • EXTINCTION OF THE FEAR RESPONSE: In a therapy session, a person with a phobia (e.g., fear of spiders or dogs or heights) might (a) have the fear response reactivated (have them stand near a spider or dog or on a high perch) and then, (b) through continuous or repeated exposure to the source of fear with support from the therapist and experience of no bad consequences (not getting bitten or not falling), show a reduction of the fear response.
  • DRUGS THAT BLOCK FEAR MEMORY: In a therapy session, a person with a phobia (e.g., fear of spiders or dogs or heights) might (a) have the fear response reactivated (have them stand near a spider or dog or on a high perch) and then, (b) the person is given propranolol, a drug that inhibits the storage of emotional aspects of a memory.
• REPETITION ACROSS DAY OR WEEKS: For a deep-seated problem, it is very unlikely that a single session will eliminate or even substantially reduce the automatic negative emotional response. The process of reinstatement followed by either extinction or drug intervention is necessary for effective treatment.

As you can see from the video, therapists can now use the new insights coming from research on reconsolidation of memory to help in their treatment of people with disorders that include memory dysfunctions. The video showed treatment of a phobia, but reconsolidation therapy has also been used with some success with people suffering from PTSD.

The reconsolidation research discussed is this exercise is just one example of the relationship between basic research taking place in scientific laboratories and practical application of discoveries about the mind and brain in the real world. Psychology in the 21st century owes a great deal to researchers in the 20th century, but old dogma is constantly being updated and even overthrown in favor of better ideas that come from deeper understanding of the causes of human behavior.