The Tease of Memory
Psychologists are dusting off 19th-century explanations of déjà vu. Have we been here before?


"In the summer of 1856, Nathaniel Hawthorne visited a decaying English manor house known as Stanton Harcourt, not far from Oxford. He was struck by the vast kitchen, which occupied the bottom of a 70-foot tower. "Here, no doubt, they were accustomed to roast oxen whole, with as little fuss and ado as a modern cook would roast a fowl," he wrote in an 1863 travelogue, Our Old Home.

Hawthorne wrote that as he stood in that kitchen, he was seized by an uncanny feeling: "I was haunted and perplexed by an idea that somewhere or other I had seen just this strange spectacle before. The height, the blackness, the dismal void, before my eyes, seemed as familiar as the decorous neatness of my grandmother's kitchen." He was certain that he had never actually seen this room or anything like it. And yet for a moment he was caught in what he described as "that odd state of mind wherein we fitfully and teasingly remember some previous scene or incident, of which the one now passing appears to be but the echo and reduplication."

When Hawthorne wrote that passage there was no common term for such an experience. But by the end of the 19th century, after discarding "false recognition," "paramnesia," and "promnesia," scholars had settled on a French candidate: "déjà vu," or "already seen."

The fleeting melancholy and euphoria associated with déjà vu have attracted the interest of poets, novelists, and occultists of many stripes. St. Augustine, Sir Walter Scott, Dickens, and Tolstoy all wrote detailed accounts of such experiences. (We will politely leave aside a certain woozy song by Crosby, Stills, Nash & Young.)

Most academic psychologists, however, have ignored the topic since around 1890, when there was a brief flurry of interest. The phenomenon seems at once too rare and too ephemeral to capture in a laboratory. And even if it were as common as sneezing, déjà vu would still be difficult to study because it produces no measurable external behaviors. Researchers must trust their subjects' personal descriptions of what is going on inside their minds, and few people are as eloquent as Hawthorne. Psychology has generally filed déjà vu away in a drawer marked "Interesting but Insoluble."

During the past two decades, however, a few hardy souls have reopened the scientific study of déjà vu. They hope to nail down a persuasive explanation of the phenomenon, as well as shed light on some fundamental elements of memory and cognition. In the new book The Déjà Vu Experience: Essays in Cognitive Psychology (Psychology Press), Alan S. Brown, a professor of psychology at Southern Methodist University, surveys the fledgling subfield. "What we can try to do is zero in on it from a variety of different angles," he says. "It won't be something like, 'Boom! The explanation is there.' But we can get gradual clarity through some hard work."

Fatigue and Freud

In their brief late-19th-century flirtation with déjà vu, academic psychologists developed remarkably sophisticated hypotheses, some of which survive today. An article in a German psychology journal in 1878 suggested that déjà vu happens when the processes of "sensation" and "perception," which normally occur simultaneously, somehow move out of sync. Fatigue, it said, may be a cause.

Eleven years later, William H. Burnham, a psychologist at Clark University, in Worcester, Mass., offered the opposite suggestion: that déjà vu occurs when the nervous system is unusually well rested. "When we see a strange object," he wrote, "its unfamiliar aspect is largely due to the difficulty we find in apperceiving its characteristics. ... [But] when the brain centers are over-rested, the apperception of a strange scene may be so easy that the aspect of the scene will be familiar."

That idea may sound peculiar: Could our minds really be thrown out of kilter by unusually speedy and well-greased visual signals? But a large body of modern research strongly suggests that brains do use speed as a tool to assess whether an image or situation is familiar or not. If we can process an image fluently and quickly, our brains unconsciously interpret that as a cue that we have seen it before. Both the "fatigue" and the "well rested" theories of déjà vu remain on the table today.

In 1896 Arthur Allin, a professor of psychology at the University of Colorado at Boulder, wrote a long essay that covered many potential explanations. Among other possibilities, he suggested that déjà vu situations feel familiar because they remind us of elements of forgotten dreams; that our emotional reactions to a new image can conjure a false feeling of familiarity; and that déjà vu is generated when our attention is very briefly interrupted during our introduction to a new image.

Such inquiries nearly ground to a halt in the early-20th century, in part because of the shadow of Freud. A new generation of scholars arose for whom déjà vu was unmistakable evidence of the ego's struggle to defend itself against id and superego. In 1945 the British psychologist Oliver L. Zangwill wrote a 15-page essay explaining that Hawthorne's episode at Stanton Harcourt stemmed from an unresolved erotic yearning for his mother. (This despite Hawthorne's own plausible conclusion that his déjà vu was sparked by a dimly remembered Alexander Pope poem about the building.) As late as 1975 the prominent psychologist Bernard L. Pacella proposed that déjà vu occurs when the ego goes into a regressive panic, "scanning the phases of life in a descent historically to the composite primal-preobject-early libidinal object-representations of mother."

4 Modern Approaches

Most of today's déjà vu scholars have chucked primal-preobject-libidinal representations in favor of brain scans and neuroimaging. Taking advantage of a recent explosion of experimental research on memory errors, Mr. Brown and a few like-minded colleagues have dusted off the theories of déjà vu proposed during the late Victorian era. At last, he hopes, such hypotheses can be subject to rigorous experimental tests. He warns, however, not to expect quick results: "A lot of science is geared at, How can I get tenure? How can I crank out a study in a year? The luxury of being able to attack difficult problems is often more risky. There's a little more investment of your personal resources, a little bit of gambling."

In Mr. Brown's account, scientific theories of déjà vu fall into four broad families. The first are theories of "dual processing." The late neuropsychiatrist Pierre Gloor conducted experiments in the 1990s strongly suggesting that memory involves distinct systems of "retrieval" and "familiarity." In a 1997 paper, he speculated that déjà vu occurs at rare moments when our familiarity system is activated but our retrieval system is not. Other scholars argue that the retrieval system is not shut off entirely but simply fires out of sync, evoking the fatigue theory of a century earlier.

In the second category are more purely neurological explanations. One such theory holds that déjà vu experiences are caused by small, brief seizures, akin to those caused by epilepsy. That idea is buttressed by the fact that people with epilepsy often report having déjà vu just before going into full-blown seizures. Researchers have also found that déjà vu can be elicited by electrically stimulating certain regions of the brain. In a 2002 paper, the Austrian physician Josef Spatt, who works with epilepsy patients, argued that déjà vu is caused by brief, inappropriate firing in the parahippocampal cortex, which is known to be associated with the ability to detect familiarity.

Mr. Brown's third category consists of memory theories. These propose that déjà vu is triggered by something we have actually seen or imagined before, either in waking life, in literature or film, or in a dream. Some of these theories hold that a single element, perhaps familiar from some other context, is enough to spark a déjà vu experience. (Suppose, for example, that the chairs in Stanton Harcourt's kitchen were identical in color and shape to Hawthorne's decorously neat grandmother's, but that he didn't recognize them in this new context.) At the other end of the scale are gestalt theories, which suggest that we sometimes falsely recognize a general visual or audio pattern. (Suppose that the Stanton Harcourt kitchen looked similar, in broad visual outline, to a long-forgotten church that Hawthorne had once attended.)

In the final box are "double perception" theories of déjà vu, which descend from Allin's 1896 suggestion that a brief interruption in our normal process of perception might make something appear falsely familiar. In 1989, in one of the first laboratory studies that tried to induce something like déjà vu, the cognitive psychologists Larry L. Jacoby and Kevin Whitehouse, of Washington University in St. Louis, showed their subjects a long list of words on a screen. The subjects then returned a day or a week later and were shown another long list of words, half of which had also been on the first list. They were asked to identify which words they had seen during the first round.

The experimenters found that if they flashed a word at extremely quick, subliminal speeds (20 milliseconds) shortly before its "official" appearance on the screen during the second round, their subjects were very likely to incorrectly say that it had appeared on the first list. Those results lent at least indirect support to the notion that if we attend to something half-consciously and then give it our full attention, it can appear falsely familiar.

The study is one of many that demonstrate the potential pitfalls of everyday memory and cognition, says Mr. Jacoby. "At our core, I think all of us are naïve realists. We believe the world is as it presents itself," he says. "All of these experiments are a little unsettling if you're a naïve realist." He hopes that this line of research will point toward new ways to repair the mental abilities of elderly people with impaired memories. "If we highlight the distinction between memory as expressed in performance and memory as we subjectively experience it," he says, we might be able to train elderly people to avoid common errors.

Speak, Memory

Having published his survey of the déjà vu world, Southern Methodist's Mr. Brown is embarking on a research program of his own. Together with Elizabeth J. Marsh, an assist-ant professor of psychology at Duke University, he is developing an experiment that may extend the findings of Mr. Jacoby and Mr. Whitehouse. In the new studies, subjects are asked to quickly locate a small red cross that is superimposed on photographs of various campus landscapes. The researchers' expectation is that the subjects will concentrate on the crosses and not pay much attention to the backgrounds. A week later, when the subjects return, they are shown the campus photographs again -- along with many photographs not used in the first round -- and are asked, "Have you seen this place before?" and "Have you been to this place before?" (After all the slides have been shown, the participants are asked about which campuses they have actually visited.) Mr. Brown and Ms. Marsh wonder if the experiment will produce incorrect "yes" answers -- or even déjà vu experiences -- when the subjects look at the images they have half-consciously seen the week before.

Ms. Marsh, who specializes in more-orthodox studies of memory, had no particular interest in déjà vu before last year, when she was asked to review Mr. Brown's book manuscript. "I came at this as a student of basic memory and memory errors," she says. "But I became fascinated by what Alan had to say about the déjà vu literature. He described all of these funky little findings -- that people who travel frequently, for example, are more likely to experience déjà vu."

Further down the road, Mr. Brown would like to see studies that shed light on some of those odd findings. Why does déjà vu become less common as people grow older? Why do political liberals report more frequent déjà vu experiences than conservatives do? And why do the majority of déjà vu experiences seem to occur when people are in mundane settings? (Arthur T. Funkhouser, a Jungian analyst in Switzerland who is considering writing a book about the phenomenon, believes that it offers a window into the self -- but concedes that the raw material of déjà vu experiences are often oddly dull. "Why does the unconscious pick such banal elements for us to think about?" he asks.)

Mr. Brown would also like to work with people with epilepsy, and with people who have the rare condition of suffering déjà vu pretty much every day. "I'm in contact with someone by e-mail who has almost constant déjà vu," he says. "Someone like that would be very fruitful to work with in the lab."

But he does not expect to see any clear conceptual or experimental breakthroughs anytime soon. It is possible, he says, that what we call déjà vu is actually five or six phenomena, with separate causes. "This will be very slow progress toward a very abstract phenomenon," he says. "It's kind of like space exploration. You're not sure exactly what you'll find."