• Choose a category
• All
• Classic-Fiction

By Root 914 0

more accurately than we can recall their more mundane counterparts—a theory that accords with how it feels to remember them. Such recollections are called “flashbulb memories,” since they seem to have the perfect fidelity of photography. Psychologists speculated that these memories stemmed from unique evolutionary imperatives and were formed through different neurological processes than we use to recall everyday life. But while the unusual vividness and specificity of such memories were well established (thanks largely to a 1977 study of people’s recollections of the assassination of John F. Kennedy), no one had ever put their accuracy to the test.

National tragedy is good to memory researchers. In 1986, when the space shuttle Challenger exploded, Neisser saw an opportunity to remedy this gap in the memory literature, and to find out whether his own mistaken Pearl Harbor recollection was an anomaly. He surveyed his students about their memories of the disaster the day after it happened, and then again three years later. The results spelled the end of conventional flashbulb memory theory. Less than 7 percent of the second reports matched the initial ones, 50 percent were wrong in two-thirds of their assertions, and 25 percent were wrong in every major detail. Subsequent work by other researchers only confirmed the conclusion. Our flashbulb memories might remain stunningly vivid, but research suggests that their accuracy erodes over time at the same rate as our everyday recollections—a decline so precise and predictable that it can be plotted on a graph in what is known, evocatively, as the Ebbinghaus curve of forgetting. (For the record, a group of cognitive scientists and psychologists working together as the 9/11 Memory Consortium repeated and expanded on Neisser’s study after September 11, with roughly the same results.)

There is a vast body of literature, most of it in neuroscience and psychology, about how our memories come to be riddled with errors. But what interests me is why these wrong memories continue to feel so right—or, put differently, why they produce such a strong feeling of knowing. The subjects in the 1977 study of the Kennedy assassination described their recollections of that event as “burned on their brain,” and as vivid “as if it happened yesterday.” More strikingly, when Neisser showed one of his subjects her initial report of the Challenger disaster—a report that didn’t match her memory of it—she responded by saying, “I know that’s my handwriting, but I couldn’t possibly have written that.” Likewise, despite everything you just read, you probably remain powerfully confident in your memories of September 11.

You might be wrong, but you are not alone. None of us capture our memories in perfect, strobe-like detail, but almost all of us believe in them with blinding conviction. This conviction is most pronounced with respect to flashbulb memories, but it isn’t limited to them.* Even with comparatively trivial matters, we believe in our recollections with touching sincerity and defend them with astounding tenacity. We squabble with our sister over who shrank the sweater back in 1984, we disagree with our lover of fifteen years about the location of our third date, and we simply can’t let it go. We might drop the subject, but—barring clear-cut evidence against us—we retain the deep inner certainty that we are right.

How can we square this feeling of rightness with the very real possibility that we are wrong? This is a question that haunts all of wrongology, not just errors of memory. The problem is suggested by the very phrase “the feeling of knowing.” In life, as in language, we begin with a psychological state (the “feeling” part) and end up with a claim about the truth (the “knowing” part). In other words, we feel that we are right because we feel that we are right: we take our own certainty as an indicator of accuracy. This isn’t completely foolish of us, since studies show that there is some correlation between confidence and correctness. But it isn’t completely foolproof, either. As the case of