Cover Story

All of us have memories we'd prefer to forget. That foot-in-the-mouth moment at a party last summer, that embarrassing performance in our high-school talent show...

When memories like those come unwittingly to mind--prompted, say, by a run-in with someone who witnessed the moment--we might try to push the thought quickly away. But could such repeated suppressions actually make us less likely to remember the event years later?

Yes, quashing memories could impede our recall, suggests a recent series of studies by University of Oregon psychologist Michael Anderson, PhD, and his colleagues. Anderson says that his laboratory model--which taps the executive control processes that people use to concentrate and overcome interference during memory tasks--could in principle explain how people, over time, suppress distracting, unwanted or even traumatic memories.

His work could possibly help explain post-traumatic stress disorder and even, controversially, repressed memories of childhood trauma. But some researchers remain unconvinced and question whether his lab-based results will translate to real-world memory.

The experiments

Psychologists have been debating the existence of repressed memories--first suggested by Sigmund Freud--for years. And Anderson hasn't shied away from the controversial implications of his work. In his very first paper on the topic, published in 2001 in the journal Nature (Vol. 410, No. 6826, pages 366-369), he begins the abstract by acknowledging that "Freud proposed that unwanted memories can be forgotten by pushing them into the unconscious, a process called repression. The existence of repression has remained controversial for more than a century."

At the end of the paper, he writes that his study bolsters the evidence that suppression is real: "These findings thus support a suppression mechanism that pushes unwanted memories out of awareness, as posited by Freud."

In the study, which used what Anderson has termed the "think/no-think" paradigm, he asked 32 college student participants to memorize pairs of unrelated words, like "ordeal, roach." Then he showed the participants the first word in each pair and asked them to either think of the second word or to consciously try to avoid thinking of it.

Finally, in the recall phase of the study, Anderson showed the participants the first words again and asked them to recall the second words. He found that participants were nearly 20 percent more likely to remember words that they had been asked to think about than words they'd been asked to avoid thinking about.

"Obviously this research is proof of principle," Anderson says. "In the past, people have said that there's no mechanism for memory suppression...and here's a mechanism."

Anderson also wanted to make sure that the participants had actually forgotten the target words and were not simply continuing to come up with diverting thoughts when they saw the cue word. So in a second experiment, he showed them related clues (like "insect R___" for roach) and asked them to recall the target word that best fit that clue. Again, participants were less likely to remember the words that they had been instructed to avoid thinking about.

Curious about the neural underpinnings of the phenomenon, Anderson and his colleagues decided to repeat the experiment while examining participants using functional magnetic resonance imaging. In the resulting study, published in 2004 in Science (Vol. 303, No. 5655, pages 232-235), he found that the hippocampus--which is generally active when people retrieve memories--was not active when participants were trying to suppress thoughts of the target word. On the other hand, the dorsolateral prefrontal cortex--an area that helps inhibit motor activity--was more active than usual. This suggests that people may be using the prefrontal cortex to overcome memory processes in the hippocampus, Anderson says.

Of course, in the real world people rarely try to suppress a thought as simple as a single word. Given this, other researchers have picked up and are extending Anderson's work. University of Colorado at Boulder psychologist Marie Banich, PhD, for example, is investigating whether Anderson's think/no-think paradigm will work for nonverbal as well as verbal stimuli, and for emotional stimuli. In a study in press at Psychological Science, she and her colleagues used the same research design that Anderson did, but instead paired pictures of faces with pictures of different scenes--some neutral, like a hippo in a lake, and some emotional, like the aftermath of a car crash.

As in Anderson's study, Banich had her participants memorize the face/scene pairs, then showed them the faces and asked them to either think about or avoid thinking about the associated scene.

She found two things: First, the think/no-think paradigm worked--participants recalled the scenes they'd been asked to think about better than the scenes they'd been asked not to think about. Second, it actually worked better for scenes with emotional content than for scenes with nonemotional content.

This result makes sense, Banich says: "Emotional regulation requires us to have cognitive control over things that are difficult for us to think about."

The controversy

Despite these results, some researchers remain skeptical of Anderson's work. In an upcoming issue of the journal Memory & Cognition, Washington University psychologist Henry L. Roediger III, PhD, and graduate student John Bulevich will report that they haven't been able to replicate Anderson's results.

Bulevich says that he began the studies--part of his master's thesis--intending to replicate Anderson's study and then expand it to implicit memory tests. However, his project stalled when he failed to replicate the original results.

"Inhibitory paradigms are notoriously fragile," he says. "I'm still interested in this, but I have nothing planned right now, until Anderson or his colleagues can pin down what makes this paradigm difficult to deal with."

Anderson--who helped Bulevich and Roediger with their study--says he's unconcerned with the team's failure to replicate his results. There are, he says, many variables that can go awry if not carefully controlled. For example, in the real world, we don't need to be reminded not to think about the things we don't want to think about. But, Anderson says, in the experiment it's crucial to ensure that the participants really are trying their hardest to avoid thoughts of the suppressed target words. And, he says, he has conducted a meta-analysis of the more than 1,000 participants he's tested in all his think/no-think studies, and he's found a strongly significant effect.

"I think that Roediger's paper will be useful in the long run," he says, "and it does remind us that there are factors here that are yet to be understood."

Other psychologists question whether Anderson's results, even if replicable, really mean what he thinks they mean.

In a letter to the journal Trends in Cognitive Sciences (Vol. 6, No. 12, page 502) and in an upcoming book chapter, University of California, Berkeley, psychologist John Kihlstrom, PhD, argues that Anderson's mechanism involves conscious suppression, while Freud's theories posited unconscious repression.

And, of course, the debate about whether repressed memories of childhood trauma are credible has raged for more than a decade--and those who believe they aren't find Anderson's work hard to take.

"There is no evidence that traumatized people repress memories of traumatic events," Kihlstrom says.

Anderson acknowledges that his studies are only the beginning of what will need to be many more years of research--but he says he thinks that such work will be worthwhile. "Whether or not what I've found could be scaled up to explain intense emotional memories is an empirical question that should be researched," he says. "But it's premature to conclude that it's not relevant--we just don't know enough yet."