Brain Buys (8 page)

Figure 2.1 Storage and retrieval: The two circles connected by dashed lines represent the “raisin” and “grape” nodes with no link between them. In the process of our learning that raisins are grapes, both nodes are activated. This coactivity is hypothesized to strengthen the connections between them as a result of synaptic plasticity (strong connections are represented by dark lines with inverted triangles). During retrieval, when only the raisin node is directly activated, the same connection between the raisin and grape node is used—potentially further strengthening it. (The small gray arrows represent the direction of the flow of activity.)

Like a handwritten message before the ink dries, initial memories are vulnerable and can be disturbed by a number of factors. For example, learning new information can interfere with the long-term storage of recently acquired information. Think of trying to learn a friend’s telephone number 10 minutes after having memorized your new cell phone number. Some drugs and electroconvulsive therapy also impair the formation of new memories. Animal studies have taught us that when drugs that block protein synthesis are administered to rats shortly after they learn how to navigate a maze, the rats forget the solution. The reason this class of drugs interferes with the formation of new memories is that the long-term potentiation of synaptic strength requires the synthesis of new proteins within neurons. When applied immediately after synapses are potentiated as a result of Hebbian plasticity, drugs that inhibit protein synthesis also reverse the increase in synaptic strength—the synaptic memory.
⁷
The observation that both our actual memories and “synaptic memories” (changes in synaptic strength) are similarly susceptible to erasure by protein synthesis inhibitors was one of the first pieces of evidence that the latter underlie the former.

When protein synthesis inhibitors are administered to animals hours or days after a learning experience, there is no loss of memory. Similarly, when people being treated for depression are given electroconvulsive therapy, only their memories of what happened shortly before the therapy are lost. The transition from an early stage in which memories are vulnerable to erasure to a later stage where they are much more resistant to erasure is referred to as
consolidation
.
⁸
As the ink dries, the changes in synaptic strength seem to shift from a temporary to permanent media. But what does this process correspond to at the level of the synapses? It seems to be in part a shift from a synaptic memory that relies on biochemical reactions in the synapse, to more permanent “structural” changes that initially require protein synthesis.
⁹
Animal studies suggest that, like speed daters, many synapses in our brains are exploratory in nature—temporary hookups between a pre- and a postsynaptic neuron. Longlasting learning seems to be accompanied by structural changes in the brain’s wiring diagram in the form of the permanent stabilization of these once-nomadic synapses.
¹⁰

The notion of memory consolidation has been very influential in psychology and neuroscience. There is, however, evidence that in some instances “consolidated” memories are not as immutable as once thought. Specifically, in some cases consolidated memories appear to again become vulnerable to erasure by drugs, trauma, or interference from other memories
¹¹
—a process termed
reconsolidation
. As we will see in Chapter 5, a rat easily learns to express fear in response to a specific sound, by exposing it to a situation in which sound is associated with a shock. If a protein synthesis inhibitor is administered 24 hours after learning, little or no effect of this treatment is observed on the rat’s memory. The rat still behaves fearfully. Interestingly, however, if you later give the rat the drug while it experiences a “reminder” in the form of the tone by itself (in the absence of a shock), some amnesia is induced—that is, the rat behaves as if it is less fearful of the sound. In other words, the reactivation of an older memory somehow makes it susceptible to erasure once again. Although we do not understand the precise mechanisms underlying so-called reconsolidation, these findings further demonstrate that storage and retrieval are not distinct processes.

Updating memories is an essential feature of human memory, and reconsolidation may provide a mechanism by which old memories are revised.
¹²
As we follow the career of our favorite Hollywood actress over the years, her face changes from sighting to sighting; her hair-style and color change, a few wrinkles may appear and then mysteriously disappear. Similarly, every time I see my cousin, his face looks a bit different, perhaps it’s a bit rounder and the hairline has receded. Whenever we see someone, our memory is updated a little bit. This is, of course, an unconscious process, one in which retrieval (recognition of my cousin) seems inextricably linked with storage and updating of the memory (next time I see him my brain expects to see his last incarnation). Updating memory, and the fuzzy boundary between storage and retrieval, is a valuable feature in a changing and dynamic world. But this same flexibility can contribute to serious mnemonic mistakes. Particularly if the original template was not well established, “updating” can overwrite the original memory, as happened in the case of Jennifer Thompson, or the students who as a result of a misleading question substituted the original memory of a stop sign with a yield sign.

MAKING AND MAKING-UP MEMORIES

Memory bugs that cause us to mix up words or confuse the faces of people we do not know very well are probably easy to relate to. If you have not caught yourself making one of these errors, you have likely accused a friend of having done so. Human memory, however, can also fail in a far more spectacular fashion, beyond simply melding or overwriting information. In some cases entirely new memories can be fabricated, apparently from scratch.

Perhaps one of the best documented examples of extreme false memories was a string of cases related to repressed memories reported in the 1980s and early 1990s. The seeds of these false memories were sometimes in the form of dreams, and these seeds were often cultivated into “real” memories by a therapist or counselor, sometimes over the course of years.
¹³
The cases often involved women accusing their parents of sexual abuse, resulting in severed family ties, depression, and criminal charges. In one case, a 19-year-old woman, Beth Rutherford, went to her church counselor for help coping with stress. After months of counseling Beth uncovered “repressed” memories of atrocious acts of sexual abuse by her father. The subsequent accusations eventually led her father to lose his job as a minister, and made it difficult for him to find any other job whatsoever.

As in other such cases, Beth later recanted her memories, in part because she faced hard evidence contradicting her accusations. Among numerous facts that showed that the uncovered events could not have taken place was that she was still a virgin, as revealed by a gynecological exam performed at the suggestion of an attorney.
¹⁴
Beth later said, “At the end of 2½ years of therapy, I had come to fully believe that I had been impregnated by my father twice. I remembered that he had performed a coat hanger abortion on me with the first pregnancy and that I performed the second coat hanger abortion on myself.”

The recall of events that happened to you is called
autobiographical
(or
episodic
)
memory
, and, along with semantic memory, it is a type of declarative memory. Falsely recalling that your father raped you is an incredibly extreme example of a fabricated autobiographical memory. But how reliable are our memories of what did or did not happen to us in the past? Controlled experiments have revealed that children are particularly susceptible to autobiographical errors. This should not come as a surprise to those of us who are skeptical of our own childhood memories. I have memories of my invisible friend named Cuke when I was five years old, but are they accurate? Are they truly my own? Or were they created as I heard my mother retell stories about me and my imaginary buddy?

In another study by Elizabeth Loftus and colleagues, children between ages three and five were asked to think about whether they had experienced certain events. Two were events they had indeed experienced within the last 12 months (such as a surprise birthday party or a trip to a hospital for stitches), whereas two others were events that the experimenters knew the child had never experienced (a ride in a hot air balloon and having their hand caught in a mouse trap and a subsequent trip to the hospital). The children were questioned up to 10 times over 10 weeks. Children reported with more than 90 percent accuracy those events they had truly experienced. But approximately 30 percent of the time they also reported having experienced one of the fictitious events.
¹⁵
Interpreting these results is complicated since the proportion of false responses did not increase over the interview sessions. Perhaps the results did not always reflect false memories, and the children were instead learning the boundary between telling the truth and what they believe adults want to hear. But in any case it is clear that one must be very careful about relying on the testimony of children. This lesson was learned the hard way after a number of “mass-molestation” cases. In 1989 seven employees of the Little Rascals preschool in North Carolina were accused of molesting 29 children. One of the owners was imprisoned and sentenced to 12 consecutive life sentences on the basis of the testimony of the children, which included stories about flying on spaceships and swimming with sharks. The case would be comical if it had not ruined so many lives. It started months after the police attended a seminar on “satanic ritual abuse,” and may have taken flight after a teacher slapped a student, which progressively escalated into therapy sessions and police interviews tailored to extract information regarding sexual abuse. The children initially denied there was any sexual abuse, but eventually fed therapists and investigators bizarre and inconceivable stories of abuse. The court case lasted 10 years, and at the time it was the most expensive in the history of North Carolina. In the end charges were dropped against all defendants.
¹⁶
The most obvious brain bug in this case likely had little to do with false memories, but with
biases
of the therapists and police investigators, who were willing to ignore the massive amount of data that went against their hypothesis and embrace shreds of evidence consistent with their beliefs—and who taught the children to construct narratives that fit the distorted expectations of those in charge.

WHERE IS THE DELETE COMMAND?

The mechanisms underlying extreme examples of false memories, such as Beth Rutherford’s belief that she had been sexually abused by her father, are complex and undoubtedly depend on specific personality traits, as well as the presence of a “therapist” capable of mnemonically abusing a psychologically susceptible individual. In fact, there is little evidence that traumatic memories can be repressed and later recovered with the help of a therapist. Examples of childhood sexual abuse are not easily forgotten. Most of the cases of sexual abuse by Catholic priests that emerged in the late 1990s were brought by victims who had memories of their abuse. They did not involve uncovering repressed memories, but rather the motivation and means to make the events public. Similarly, concentration camp survivors agree that forgetting the horrors they suffered and witnessed was never an option. While some of these victims may have “compartmentalized” these memories to avoid dwelling on them so they could attempt to carry on a normal life, it is impossible to forget the unforgettable. In the words of the psychologist Daniel Schacter, one of the foremost experts on human memory:

If it were possible, many people believe that the ability to permanently repress or erase traumatic memories could help treat and cure the consequences of many forms of psychological traumas. Victims of sexual abuse and violence are often haunted by their own memories and suffer from anxiety, depression, fear, and difficulties engaging in normal social interactions. It is perhaps unfortunate, then, that another difference between human memory and a hard drive is the absence of a delete command.

Humans do forget things, which is a type of deletion, but we do not have much say as to what we erase. Scientists are currently experimenting with behavioral and pharmacological methods that would at least dampen, if not erase, the intensity of emotionally charged memories, such as being raped or experiencing the horrors of battle. These studies attempt to take advantage of the notion of reconsolidation; the hope is that immediately after the recall of a traumatic experience, the memory will once again be labile and susceptible to erasure by drugs or even by new nontraumatic memories. Unfortunately, however, there may be an expiration date for reconsolidation; that is, after months or years, memories may no longer exhibit reconsolidation.
¹⁸
Furthermore, if new treatments prove successful in erasing some memories, it is unlikely that, even if it were desirable, it will ever be possible to delete specific memories as depicted in the film
Eternal Sunshine of the Spotless Mind
.

One day in 2006 I woke up and was informed that some obviously very powerful people had decided that Pluto was no longer a planet. After a lifetime of being told that Pluto was a planet my brain had created strong links between the neural representation of “planets” and the celestial object “Pluto.” Thus, in a semantic priming task, the word
Pluto
would probably speed up my reaction time to “planet.” But now I was being told that this link was incorrect. The brain is well designed to form new links between concepts, but the converse is not true: there is no specific mechanism for “unlinking.” My brain can adjust to the new turn of events by creating new links between “Pluto” and “dwarf planet,” “Pluto” and “Kuiper Belt Object,” or “Pluto” and “not a planet.” But the Pluto/planet link cannot be rapidly erased and will likely remain ingrained in my neural circuits for the rest of my life. And there may come a day late in my life in which I will revert to my first belief and insist to my grandchildren that Pluto is a planet.