In the movie Paycheck, opening Christmas Day, a crack reverse engineer helps companies steal and improve upon the technology of their rivals, then has his memory of the time he spent working for them erased. The story, based on Philip K. Dicks sci-fi thriller of the same name, is set in the near future, but such selective memory erasure is still highly speculative at best. asked neurobiologist James McGaugh of the University of California at Irvine, who studies learning and memory, to explain what kinds of memory erasure are currently possible. For more information, see his book Memory and Emotion: The Making of Lasting Memories, released in 2003.

Scientific Early in Paycheck we see the main character get several months' worth of his memories erased by having individual neurons zapped. Is that possible?

JM: No. First of all there is no evidence of memories being stored individually. And even if they were stored in individual neurons, no one would know where they were. What we know an awful lot about are the brain systems that are involved in storing memories. Your memories of this conversation, for example, are stored rather diffusely in the brain. They're not going to be stored in a couple of neurons someplace that anybody could easily identify.

SA: But haven't surgeons poked people's brains in certain spots and made them recall specific things?

JM: [In the 1950s] Wilder Penfield up at the Montreal Neurological Institute was doing surgery for people who had brain seizures, and he had to stimulate the brain and have people talk to make sure that he wouldn't eliminate speech areas, for example. He found that he could evoke some things that appeared to be like memories, but it's more likely that he was just evoking [an impression of] something, not a specific memory.

SA: Are there any ways to erase memories by stimulating the brain?

JM: The dominant evidence that goes back over 50 years is that one can block or certainly reduce memories formed within the past several hours by treating human or animal subjects with electro-convulsive shock. But it's nonselective; whatever happened in that past several hours will be gone. And that's rather gross stimulation applied to the skull. What Larry Squire at UC San Diego has shown is that if human subjects are repeatedly given electro-convulsive shocks (several times a week for several weeks), they will have impaired global memory that goes back many months, but that memory will gradually recover. He did this in the late 1980s.

SA: Are there any more selective ways to erase memory?

JM: If one work with the hippocampus, one can selectively remove animals' memories of places where they have received training. In the Morris water maze, for example, animals are trained to swim from a variety of regions [in] a six-foot tank to an invisible platform located about two centimeters below the surface of the water. That kind of learning requires the hippocampus. If the hippocampus is blocked electrically or chemically within a few hours after animals have been trained to go that spot, they will not remember it the next day. So that would be an example of a place memory that could be influenced by discrete stimulation of a specific region of the brain. This doesn't mean the memory is permanently stored there. It means that the hippocampus is involved in the processing of that information, which is ultimately stored someplace else. And it's not something that could be done by electrical stimulation applied outside of the brain except for electro-convulsive shock, which activates the entire brain.

SA: How do we know memories aren't stored in the hippocampus?

JM: If subjects are taught something and then the hippocampus is removed several weeks later (in an extreme case), the information is retained. Take the very famous case of HM. He has anterograde amnesia--he is unable to learn new knowledge about the world--because he doesn't have his anterior hippocampus. He has retrograde amnesia that goes back for maybe a year, but for earlier times--a year longer and back to childhood--the memories are pretty much intact. So that says that these regions cannot be the repositories of a long-term memory. [For more on the milder sorts of memory problems we all face, click here.]

SA: Doesn't some research indicate that every time a memory is recalled, there is a window of a few hours in which it can be erased?

JM: That's highly controversial. This issue was first studied in the 1970s, and the idea was that if you recall a memory it becomes vulnerable to destruction, just as it was during the original learning. Additional studies showed that the effect is temporary. When you apply some amnesic treatment after animal subjects recall something, the memory impairment goes away within a few days. So it's not as though the treatment destroyed the memory. That hypothesis, called the reconsolidation hypothesis, was revived a couple of years ago, and evidence supporting it is mixed at best. It is on very shaky grounds.

SA: You discovered that a drug called propranolol could reduce recall of emotionally charged memories. How did that come about?

JM: My colleague Larry Cahill [of the University of California at Irvine] and I discovered that, based on about 20 years of prior animal research. (My work is primarily with laboratory animals.) We learned that strong emotions make for strong memories. The people who were around at 9/11 are going to remember that forever. And I'm sure the people who win the lottery remember where they were and what they were doing when they won the lottery. Over the course of several decades we traced out how this happens. Adrenaline is released from the adrenal medulla and then activates the brain by what are called beta-receptors. These are protein receptors on neurons that receive adrenaline and its first cousin noradrenaline. Their activation enables strong emotions to make strong memories.

We reasoned then that we should be able to block the formation of strong emotionally evoked memories by giving human subjects a beta-blocker, which would prevent the action of adrenaline and noradrenaline. The experiment was as follows. Subjects were shown a series of 12 slides depicting one of two stories. The subjects believe they are just rating emotionality of the pictures. In the boring story, a mother and boy leave home, see a damaged car, and go to visit father, who works in a hospital. The other group gets the same slides, but as mother and boy cross the street, the boy is hit by the car, his legs are severed, and they rush him to the hospital where surgeons work frantically to reconnect his legs. Now the important thing is that only the middle part of the story is emotional and, as it turns out, the subjects that got the emotional story remembered more of the middle part.

The next experiment was very simple: we presented only the emotional story, but before it started we gave the subjects a beta-blocker, propranolol. We measured their memory three weeks later and it was as if we had given them the boring story. We had eliminated the effect of emotional arousal on memory. This paper was published in the mid-90s, and that has been replicated many times now.

SA: Is anyone using this therapeutically?

JM: Yes, two clinical trials have been published recently, one by Roger Pitman at Harvard University in which subjects are put on beta-blockers as soon as possible after they have had a traumatic injury such as an automobile accident. People are at risk of post-traumatic stress disorder when they have such injuries. The idea is to blunt the intensity of those memories so that they won't take over the lives of those people. And the answer in both of the clinical trials, one in the U.S. and the other in France, is yes: several months after the incident or injury there are significantly decreased signs of post-traumatic stress disorder.

SA: Aside from electric shock and propranolol are there any other examples of drugs or methods that have memory dulling or erasing effects?

JM: Benzodiazepines such as Valium create anterograde amnesia. As a matter of fact, one of the drugs, Halcion, was removed from the market in Great Britain because it has very potent anterograde amnesic affects. Human subjects can take these drugs to reduce anxiety and go through the day and think that everything is fine. But the next day they remember very little of the previous day.

SA: Do we know how that happens?

JM: Yes, to a large degree. I've done an awful lot of work investigating that in my own laboratory and we know that when receptors in the basolateral region of the amygdala are hit by benzodiazepines, memories are not formed.

SA: What is the amygdala doing?

JM: What is now widely accepted is that when the basolateral amygdala gets activated, it instructs other regions of the brain to make stronger memories of what has just happened. We've done countless experiments in which we've trained animals and then tickled the basolateral amygdala with various chemical agents either immediately afterwards or several hours afterwards. We found that if we tickled it immediately afterwards we could either enhance or impair the formation of memory of all kinds. So when it is put out of commission by benzodiazepine there can be no modulation of the storage of recent events. Now there are several studies using brain imaging pioneered, once again, by my colleague Larry Cahill that show the degree of activation of the amygdala at the time subjects learn emotionally arousing material is a highly reliable predictor of how well that material will be remembered several weeks later.

SA: Is it conceivable that we could ever have control of memory erasure or do you believe it will always be a global thing?

JM: Ever is a long time. It is conceivable that things might get more precise. You never know what technology is going to provide. I never thought 20 years ago that we would be able to peer into the brain using brain imaging and watch it at work. So it's conceivable that technology in the future will allow us to do some things to the brain from outside that we cannot now do, and there could be some greater selectivity.

Let me give a caveat to that. Let's suppose I said I am going to say a word to you and then I am going to use a new magical technique that technology has provided to eliminate that memory. So now I say "bicycle" and I go zap. Now think about bicycle. I said the word bicycle, but you know the general meaning of bicycle, you know what a bicycle looks like, you may remember a specific bicycle, you may remember having done something on a specific bicycle, and you remember how to ride a bicycle. Did I get rid of all of those? Not very likely, because those are all very deeply embedded in all kinds of memories that you have. Let's suppose that I did get rid of all of those; then you'd lose a huge portion of your life. Everything you knew about riding a bicycle would go, which would mean everything connected to that would also go. Your home, your family would be gone, and it can't work that way. There is an interconnectedness to the knowledge we have. Information does not exist in the abstract.

SA: Do movies ever treat memory problems accurately? Memento (in which a man who can't form new memories tries to avenge his wife's murder) is a personal favorite.

JM: I've seen a number of movies in which memories are not supposed to be working properly and I have not yet seen one that is faithful to what we know about how memories work. Memento was very unfaithful. An amnesic cannot learn new information and this guy was randomly learning sometimes and not others. I was making note of it while watching the movie. He was doing things that he couldn't possibly do if he were amnesic because he wouldn't be able to remember to. Then there were other times in which they showed him as a classic amnesic in which he would repeatedly introduce himself to someone even though he had introduced himself many times. That's what the amnesic would do, but also he has this plan to do something; well, amnesiacs don't have the plan to do anything, because they can't even remember if they had breakfast. A true amnesic's life wouldn't be interesting because they have to be taken places--because they don't know where they are supposed to go. We forget that memories are required in order to make plans. You have to have an idea of what you intend to do. And what you intend to do is a memory.

If you lose the ability to recall your old memories then you have no life. You might as well be a rutabaga or a cabbage. And you see that when you see memories disappear in people who have severe brain diseases. You see the memories wilting away. They can't remember words, they can't remember the names of their family, then they can't remember to do anything, and then they become comatose and they die. Memory--there can be no doubt about it--is our most important set of abilities.

JR Minkel is a freelance writer based in New York City.