This article was taken from the April 2012 issue of Wired magazine. Be the first to read Wired's articles in print before they're posted online, and get your hands on loads of additional content by subscribing online.
Jeffrey Mitchell, a volunteer firefighter in the suburbs of Baltimore in the US, came across the accident by chance: a car had smashed into a pickup truck loaded with metal pipes. Mitchell tried to help, but he saw at once that he was too late.
The car had rear-ended the truck at high speed, sending a pipe through the windscreen and into the chest of the passenger --a young bride returning home from her wedding. There was blood everywhere, staining her white dress crimson.
Mitchell couldn't get the dead woman out of his mind; the tableau was stuck before his eyes. He tried to tough it out, but after months of suffering, he couldn't take it any more. He finally told his brother, a fellow firefighter, about it.
Miraculously, that worked. No more trauma; Mitchell felt free.
This dramatic recovery, along with the experiences of fellow first responders, led Mitchell to do some research into recovery from trauma. He eventually concluded that he had stumbled upon a powerful treatment. In 1983, nearly a decade after the car accident, Mitchell wrote an influential paper in the Journal of Emergency Medical Services that transformed his experience into a seven-step practice, which he called Critical Incident Stress Debriefing (CISD). Its central idea: people who survive a painful event should express their feelings soon after so the memory isn't "sealed over" and repressed, which could lead to post-traumatic stress disorder (PTSD).
In recent years, CISD has become exceedingly popular, used by the United Nations, the US Department of Defense, the Federal Emergency Management Agency, the American Red Cross and the Israeli army. Each year, more than 30,000 people are trained in the technique in the US. After the 9/11 attacks, 2,000 CISD facilitators descended on New York.
A typical CISD session lasts about three hours and involves a trained facilitator who encourages people involved to describe the event from their perspective in as much detail as possible.
Facilitators are trained to probe deeply and directly, asking questions such as, "What was the worst part of the incident for you personally?" The underlying assumption is that a way to ease a traumatic memory is to express it.
The problem is, CISD rarely helps. In fact, recent studies show it often makes things worse. In one, burn victims were randomly assigned to receive either CISD or no treatment at all. A year later, those who went through a debriefing were more anxious and depressed and nearly three times as likely to suffer from PTSD.
Another trial showed CISD was ineffective at preventing post-traumatic stress in victims of violent crime, and a US Army study of 952 Kosovo peacekeepers found that debriefing did not hasten recovery, but it did lead to more alcohol abuse.
Psychologists have begun to recommend that the practice be discontinued for survivors of disasters. (Mitchell now says that he doesn't think CISD necessarily helps post-traumatic stress at all, but his early papers on the subject seem clear on the link.)
Mitchell was right about one thing, though: traumatic, persistent memories are indeed a case of recall gone awry. But as a treatment, CISD misunderstands how memory works. It suggests that the way to get rid of a bad memory, or at a minimum denude it of its negative emotional connotations, is to talk it out. That's where Mitchell went wrong. It wasn't his fault, really; this mistaken notion has been around for years. Once a memory is formed, we assume that it will stay the same. This, in fact, is why we trust our recollections. They feel like indelible portraits of the past.
When CISD fails, it fails because, as scientists have recently learned, the very act of remembering changes the memory itself. New research is showing that every time we recall an event, the structure of that memory in the brain is altered in light of the present moment, warped by our current feelings and knowledge.
That's why pushing to remember a traumatic event so soon after it occurs doesn't unburden us; it reinforces the fear and stress that are part of the recollection.
This new model of memory isn't just a theory -- neuroscientists actually have a molecular explanation of how and why memories change. In fact, their definition of memory has broadened to encompass not only the clichéd cinematic scenes from childhood but also the persisting mental loops of illnesses such as PTSD and addiction -- and even pain disorders like neuro-pathy. Unlike most brain research, the field of memory has actually developed simpler explanations. Whenever the brain wants to retain something, it relies on just a handful of chemicals. Even more startling, an equally small family of compounds could turn out to be a universal eraser of history, a pill that we could take whenever we wanted to forget anything.
And researchers have found one of these compounds.
In the very near future, the act of remembering will become a choice.
Every memory begins as a changed set of connections among cells in the brain. If you happen to remember this moment -- the content of this sentence -- it's because a network of neurons has been altered, woven more tightly together within a vast electrical fabric. This linkage is literal: for a memory to exist, these scattered cells must become more sensitive to the activity of the others, so that if one cell fires, the rest of the circuit lights up as well. Scientists refer to this process as long-term potentiation, and it involves an intricate cascade of gene activations and protein synthesis that makes it easier for these neurons to pass along their electrical excitement. Sometimes this requires the addition of new receptors at the dendritic end of a neuron, or an increase in the release of the chemical neurotransmitters that nerve cells use to communicate. Neurons will actually sprout new ion channels along their length, allowing them to generate more voltage. Collectively this creation of long-term potentiation is called the consolidation phase, when the circuit of cells representing a memory is first linked together. Regardless of the molecular details, it's clear that even minor memories require major work. The past has to be wired into your hardware.
That understanding of how memories are created emerged in the 70s. But what happens after a memory is formed, when we attempt to access it, was much less well understood. In the late 90s, Karim Nader, a young neuroscientist studying emotional response at New York University, realised that nobody knew. "My big advantage was that I wasn't trained in memory," Nader says. "I was very naive about the subject."
He began with the simplest question he could think of. While it was clear that new proteins were needed for the making of memories -- proteins are cellular bricks and mortar, the basis of any new biological construction -- were additional proteins made when those memories were recalled? Nader hypothesised that they were, and he realised that he could test his notion by temporarily blocking protein synthesis in a brain and looking to see if that altered recall.
His boss, neuroscientist Joseph LeDoux, wasn't optimistic. "I told Karim he was wasting his time," LeDoux says. To LeDoux, the reason was obvious: even if Nader blocked protein synthesis during recall, the original circuitry would still be intact, so the memory should be too. Any amnesia induced would be temporary. LeDoux and Nader made a bet: if Nader failed to permanently erase fear memories in four lab animals, he had to buy LeDoux a bottle of tequila.
He taught several dozen rats to associate a loud noise with a mild but painful electric shock. It terrified them -- whenever the sound played, the rats froze in fear. After reinforcing this memory for several weeks, Nader hit the rats with the noise again, but this time he then injected their brains with a chemical that inhibited protein synthesis. Then he played the sound again. "I couldn't believe what happened," Nader says. "The fear memory was gone. The rats had forgotten everything." The absence of fear persisted even after the injection wore off.
The secret was the timing: if new proteins couldn't be created during the act of remembering, then the memory ceased to exist. The erasure was also exceedingly specific. The rats could still learn new associations and they remained scared of other sounds associated with a shock that hadn't been played during the protein block. They forgot only what they'd been forced to remember while under the influence of the protein inhibitor.
The disappearance of the fear memory suggested that every time we think about the past we are delicately transforming its cellular representation in the brain, changing its underlying neural circuitry. It was a stunning discovery: memories are not formed and then pristinely maintained, as neuroscientists thought; they are formed and then rebuilt every time they're accessed.
After collecting his tequila, Nader hit the library in an attempt to make sense of his bizarre observations. "I couldn't believe that no one had ever done this experiment before," he says. "I thought, there's no way I'm this lucky." Nader was right. He had unknowingly replicated a 44-year-old experiment performed by a Rutgers psychologist named Donald Lewis, in which rats had been trained to be afraid of a sound -- associating it, again, with an electric shock -- and then had those memories erased by a separate electroconvulsive shock.
Lewis had discovered what came to be called memory reconsolidation: the brain's practice of recreating memories over and over again. "These guys had discovered it all way before me," Nader says. "But they had been left out of all the textbooks."
Nader was convinced that Lewis's work had been rejected unjustly. But no one wanted to hear it. He was shunned at conferences and accused in journal articles of "forgetting the lessons of the past". But Nader was so angry at his scientific opponents that he refused to let them win, and by 2005 other researchers had started to take his side. Multiple papers demonstrated that the act of recall required some kind of protein synthesis -- that it was, at the molecular level, nearly identical to the initial creation of a long-term recollection.
Once you start questioning the reality of memory, things fall apart pretty quickly. So many of our assumptions about the human mind -- what it is, why it breaks, and how it can be healed -- are rooted in a mistaken belief about how experience is stored in the brain. (According to a recent survey, 63 per cent of Americans believe that human memory "works like a video camera, accurately recording the events we see and hear so that we can review and inspect them later.") We want the past to persist, because the past gives us permanence. It tells us who we are and where we belong.
But what if your most cherished recollections are also the most ephemeral thing in your head?
Consider the study of flashbulb memories, extremely vivid, detailed recollections. Shortly after the 9/11 attacks, a team of psychologists led by William Hirst and Elizabeth Phelps surveyed several hundred subjects about their memories of that awful day.
The scientists then repeated the surveys, tracking how the stories steadily decayed. After one year, 37 per cent of the details had changed. By 2004 that number was approaching 50 per cent. Some changes were innocuous -- the stories got tighter and the narratives more coherent -- but other adjustments involved a wholesale retrofit. Some people even altered where they were when the towers fell. Over and over, the act of repeating the narrative seemed to corrupt its content. The scientists aren't sure about this mechanism, and they have yet to analyse the data from the entire ten-year survey. But Phelps expects it to reveal that many details will be make-believe. "What's most troubling, of course, is that these people have no idea their memories have changed this much," she says. "The strength of the emotion makes them convinced it's all true, even when it's clearly not."
Reconsolidation provides a mechanistic explanation for these errors. It's why eyewitness testimony shouldn't be trusted (even though it's central to our justice system), why every memoir should be classified as fiction, and why it's so disturbingly easy to implant false recollections. (The psychologist Elizabeth Loftus has repeatedly demonstrated that nearly one-third of subjects can be tricked into claiming a made-up memory as their own. It takes only a single exposure to a new fiction for it to be reconsolidated as fact.)
And this returns us to Critical Incident Stress Debriefing. When we experience a traumatic event, it gets remembered in two separate ways. The first memory is the event itself, that cinematic scene we can replay at will. The second memory, however, consists entirely of the emotion, the negative feelings triggered by what happened.
Every memory is actually kept in many different parts of the brain.
Memories of negative emotions, for instance, are stored in the amygdala, an almond-shaped area in the centre of the brain. (Patients who have suffered damage to the amygdala are incapable of remembering fear.) By contrast, all the relevant details that comprise the scene are kept in various sensory areas -- visual elements in the visual cortex, auditory elements in the auditory cortex, and so on. That filing system means that different aspects can be influenced independently by reconsolidation.
The larger lesson is that because our memories are formed by the act of remembering them, controlling the conditions under which they are recalled can actually change their content. The problem with CISD is that the worst time to recall a traumatic event is when people are flush with terror and grief. It's the opposite of catharsis. But when people wait a few weeks before discussing an event -- as Mitchell, the inventor of CISD, did himself -- they give their negative feelings a chance to fade. As a result, the emotion is no longer reconsolidated in such a stressed state.
Subjects will still remember the terrible event, but the feelings of pain associated with it will be rewritten in light of what they feel now.
LeDoux insists that these same principles have been used by good therapists for decades. "When therapy heals, when it helps reduce the impact of negative memories, it's really because of reconsolidation," he says. "Therapy allows people to rewrite their own memories while in a safe space, guided by trained professionals. The difference is that we finally understand the neural mechanism."
But competent talk therapy is not the only way to get at those mechanisms. One intriguing approach to treating PTSD that emerged recently involves administering certain drugs and then asking patients to recall their bad memories. In one 2010 clinical trial, subjects suffering from PTSD were given MDMA (ecstasy) while undergoing talk therapy. Because the drug triggers a rush of positive emotion, the patients recalled their trauma without feeling overwhelmed. As a result, the remembered event was associated with the positive feelings triggered by the pill.
According to the researchers, 83 per cent of their patients showed a dramatic decrease in symptoms within two months. That makes ecstasy one of the most effective PTSD treatments ever devised.
Other scientists have achieved impressive results with less extreme drugs. In 2008, Alain Brunet, a clinical psychologist at McGill University, Montreal, Canada, identified 19 patients who had been suffering for several years from serious stress and anxiety disorders such as PTSD. Their traumas included sexual assaults, car crashes and violent muggings. People in the treatment group were given the drug Propranolol, a beta-blocker that has long been used for conditions like high blood pressure and performance anxiety; it inhibits norepinephrine, a neurotransmitter involved in the production of strong emotions.
Brunet asked subjects to write a detailed description of their traumatic experiences and then gave them a dose of Propranolol.
While the subjects were remembering the awful event, the drug suppressed the visceral aspects of their fear response, ensuring that the negative feeling was somewhat contained.
One week later, all the patients returned to the lab and were exposed once again to a description of the traumatic event. Here's where things got interesting: subjects who got the placebo demonstrated levels of arousal consistent with PTSD (for example, their heart rate spiked suddenly), but those given Propranolol showed significantly lower stress responses. Although they could still remember the event in vivid detail, the emotional memory located in the amygdala had been modified. The fear wasn't gone, but it no longer seemed crippling. "The results we get sometimes leave me in awe," Brunet says. "These are people who were unable to lead normal lives, and yet after just a few sessions they become healthy again."
Recoveries are possible, but they aren't necessarily neat. One of Brunet's patients was Lois, a retired member of the Canadian military living in Kingston, Ontario. When Lois (she asked for her surname to be withheld) describes the tragic arc of her life, she sounds like a cursed character in the Old Testament. Sexually molested as a child, she married an abusive man, who would later hang himself at home. Years later, her teenage daughter was hit by a truck and died. "I'd been holding it together my entire life," she says. "But when I heard my child was gone I just started sobbing and couldn't stop. I felt this pain that I thought was going to kill me."
In early 2011, Lois learned about the experimental trials being conducted by Brunet. She immediately wrote him an email, begging for help. Last spring Lois began reconsolidation treatment at Brunet's hospital in Montreal. The routine was always the same: a nurse would give her Propranolol, wait for the drug to take effect, and then have her read her life story out loud. The first few weeks were excruciating. But then, after five weeks of therapy, Lois felt herself slowly improve. She would still cry when describing the death of her daughter but she would eventually stop crying. "That was the difference," she says. "I still remembered every-thing that happened, and it still hurt so much, but now I felt like I could live with it. "
Propranolol, of course, is an imperfect drug, a vintage tool commandeered for a new purpose. Despite Brunet's optimistic assessment, many of his patients remain traumatised, albeit perhaps less so. Although he is currently conducting a larger-scale, randomised PTSD trial with the beta-blocker, future therapies will rely on more targeted compounds. "These norepinephrine inhibitors are just what's available right now," LeDoux says. "They work OK, but their effect is indirect." What reconsolidation therapy needs is a drug that can target the fear memory itself. And here's the amazing part: the perfect drug may have already been found.
The chemistry of the brain is in constant flux, with the typical neural protein lasting anywhere from two weeks to a few months before it breaks down or gets reabsorbed. How then do some of our memories seem to last forever? It's as if they are sturdier than the mind itself. Scientists have narrowed down the list of molecules that seem essential to the creation of long-term memory -- sea slugs and mice without these compounds are total amnesiacs -- but until recently nobody knew how they worked.
In the 80s, Columbia University neurologist Todd Sacktor became obsessed with this mental mystery. His breakthrough came from an unlikely source. "My dad was a biochemist," Sacktor says. "He was the one who said I should look into this molecule, because it seems to have some neat properties." Sacktor's father had suggested a molecule called protein kinase C, an enzyme turned on by surges of calcium ions in the brain. "This enzyme seemed to have a bunch of properties necessary to be a regulator of long-term potentiation,"
Sacktor says. "But so did a bunch of other molecules. It took me a few years to figure out if my dad was right."
In fact, it took Sacktor more than a decade. He spent three years just trying to purify the molecule. What he discovered is that a form of protein kinase C called PKMzeta hangs around synapses, the junctions where neurons connect, for an unusually long time. Without it, stable recollections start to disappear.
While scientists such as Nader had erased memories using chemicals that inhibited all protein synthesis, Sacktor was the first to target a single memory protein so specifically. The trick was finding a chemical that inhibited PKMzeta activity. "It turned out to be remarkably easy," Sacktor says. "All we had to do was order this inhibitor compound from the chemical catalogue and then give it to the animals. You could watch them forget."
In an experiment, Sacktor and scientists at the Weizmann Institute of Science trained rats to associate, thanks to an injection of lithium, the taste of saccharin with nausea. After just a few trials, the rats began studiously avoiding the artificial sweetener. All it took was an injection of a PKMzeta inhibitor called zeta interacting protein, or ZIP, before the rats forgot all about their aversion. The rats went back to guzzling down the stuff.
By coupling these amnesia cocktails to the memory-reconsolidation process, it's possible to get even more specific. Nader, LeDoux and a neuroscientist named Jacek Debiec taught rats elaborate sequences of association, so that a series of sounds predicted the arrival of a painful shock to the foot. Nader calls this a "chain of memories" -- the sounds lead to fear, and the animals freeze up. "We wanted to know if making you remember that painful event would also lead to the disruption of related memories," Nader says. "Or could we alter just that one association?" The answer was clear. By injecting a protein synthesis inhibitor before the rats were exposed to only one of the sounds -- and therefore before they underwent memory reconsolidation -- the rats could be "trained" to forget the fear associated with that particular tone. "Only the first link was gone," Nader says. The other associations remained perfectly intact. This is a profound result. Although scientists have long wondered how to target specific memories in the brain, it turns out to be remarkably easy: all you have to do is ask people to remember them.
Right now, researchers have to inject their obliviating potions directly into the rodent brain. Future treatments, however, will involve targeted inhibitors that become active only in particular parts of the cortex and only at the precise time a memory is being recalled. The result will be a menu of pills capable of erasing different kinds of memories. These negative thoughts and feelings can be made to vanish, even as the rest of the memory remains intact. "We can get very specific about which associations we go after," LeDoux says. "And that's a very good thing. Nobody actually wants a totally spotless mind."
The astonishing power of PKMzeta forces us to redefine human memory. Whereas we typically think of memories as those facts and events from the past that stick in the brain, Sacktor's research suggests that memory is actually much bigger and stranger than that. In fact, PTSD isn't the only disease that's driven by a broken set of memories -- other nasty afflictions including chronic pain, obsessive-compulsive disorder and drug addiction are also fuelled by memories that can't be forgotten.
Sacktor is convinced that the first therapeutic use of PKMzeta inhibitors will involve making people forget not an event, but physical pain. For reasons that remain mysterious, some sensory nerves never recover from bodily injury; even after a wound heals, the hurt persists. Because these memories are made of exactly the same stuff as every other kind of memory, injecting an inhibitor near the spinal cord -- where, presumably, the sensation of pain is being stored -- and then somehow inducing or focusing on the pain, could erase long-term suffering, as if the nerves themselves were reset. PTSD is the emotional version of this problem. Instead of the pain coming from the spinal cord, it comes from the amygdala, where a trauma is encoded and just won't let go. It doesn't matter if the tragedy is physical or psychic: the treatment is the same.
The problem with eliminating pain, of course, is that pain is often educational. We learn from our regrets and mistakes; wisdom is not free. If our past becomes a playlist -- a collection of tracks we can edit with ease -- then how will we resist the temptation to erase the unpleasant ones? Even more troubling, it's easy to imagine a world where people don't get to decide the fate of their own memories. "My worst nightmare is that some evil dictator gets a hold of this," Sacktor says. "There are all sorts of dystopian things one could do with these drugs." Whereas tyrants have often rewritten history books, modern science might one day allow them to rewrite us, wiping away genocides and atrocities with a cocktail of pills.
Those scenarios aside, the fact is we already tweak our memories -- we just do it badly. We repeat stories until they're stale, rewrite history in favour of the winners, and tamp down our sorrows with booze. "Once people realise how memory works," says Nader, "a lot of these beliefs that memory shouldn't be changed will seem ridiculous."
Some ethicists and clinicians dispute whether this kind of therapy is acceptable. Researchers in the field counter that not treating suffering is cruel, regardless of the type of pain involved. We have a duty, they say, to take psychological pain seriously.
At the moment, of course, such treatments remain hypothetical, limited to the lab. PKMzeta inhibitors can zap rodent memories, but we can't ask the rats how they feel afterwards. Maybe they feel terrible. Maybe they miss their fear. Maybe they miss their morphine. Or maybe all they know is that they miss something. They just can't remember what.
Jonah Lehrer's latest book is Imagine: How Creativity Works (Canongate)
This article was originally published by WIRED UK
