The Invisible Gorilla: And Other Ways Our Intuition Deceives Us. Christopher Chabris

than we do in vague or generic recollections, especially when the detail has such a neat relationship to the rest of the story. Had Chris not been present, Ken Norman would have gotten away with his Captain Picard story in part because of the distinctiveness of the Baked Alaska, autograph-seeking cooks, and embarrassed manager. But as we have seen, these deceptively vivid details can be telltale traces of the processes of distortion and reconstruction that operate on memories after they are formed. Could the detail about the taxi driver be accurate? Certainly. Might Chris’s father have fabricated the Arab driver out of whole cloth? Possibly. Could he have inadvertently combined two separate memories, one of going home by taxi on 9/11 and another of having an Arab taxi driver (a common experience for someone living in the New York area)? Absolutely. The ironic final twist does make for a more compelling story—which is exactly what our memory systems are constantly, unbeknownst to us, striving to do.

      Let’s revisit the story of Leslie and Tyce, the couple who witnessed a stabbing and were put on hold by 911. Within a minute of the event, they realized that they already disagreed about what they had seen. Despite recounting this story many times over the six years between the incident itself and their interviews with Chris, their memories have only diverged further: Leslie reported honking their horn to draw attention to the crime scene; when told of this, Tyce said “Really?” Leslie remembered being several lanes away from the sidewalk; Tyce recalled just a row of parked cars between them and the assault. Leslie thinks the attack happened in front of a dark, boarded-up building; Tyce recalls “a convenience store or takeout chicken store, a place with big neon lights in front.” Leslie says the attacker was bigger than the victim; Tyce says the opposite. Leslie thinks it took about thirty seconds for 911 to pick up, and that the conversation lasted three or four minutes; Tyce remembers a five-minute wait followed by a one-minute conversation. And while we told you that Leslie placed the call from the passenger seat while Tyce was driving, Tyce remembers himself calling 911 while Leslie was driving. It seems that our memory systems do like to place us in the center of the action.⁴¹

      Think back one last time to your own memory of how you learned about the attacks of September 11, 2001. Now that you know about the illusion of memory, you know that you should doubt the veracity of your own recollections. But if you still have trouble overcoming the convincing impression that your memory is right, you aren’t alone. In a more recent flashbulb memory study, psychologists Jennifer Talarico and David Rubin examined people’s accounts of how they heard about the 9/11 attacks.⁴² Unlike all previous studies of flashbulb memories, theirs compared how well people remembered the flashbulb event with how well they remembered another event from about the same time. Thinking creatively and quickly at an emotional time, on September 12, 2001, Talarico and Rubin had a group of Duke University undergraduates come into the lab and complete a detailed questionnaire about how they first heard about the attacks. They also had the undergraduates recall another personal memory of their choosing that was still fresh in their minds from the few days just prior to the attacks. Then either 1, 6, or 32 weeks later, they asked their subjects to recall each event again. All of the memories, whether of 9/11 or of the more ordinary event, became more inaccurate as more time passed. The longer the gap between the original recollection and the later test, the less consistent the memories, and the more false details they included.

      Talarico and Rubin did one more clever thing. They asked the students to rate how strongly they believed in the accuracy of their own memories. For the everyday memory, people had a good sense of how accurate they were: As their memories got worse, they were less confident in them. That is, they did not suffer from the illusion of memory for everyday events. Just as people know that memory for arbitrary facts is fallible, they know that they forget otherwise trivial details about their experiences. When they cannot recollect the details well, they become less trusting of their memories.

      The flashbulb memories showed an entirely different pattern, though. Subjects continued to believe strongly in the accuracy of their memories even though their memories became less accurate over time. The illusion of memory—the difference between how accurate our memories are and how accurate we think they are—operates at maximum strength for flashbulb memories. Early writing on flashbulb memories suggested that they were created by the activation of a special “print now” mechanism in the brain. In light of Talarico and Rubin’s findings, it may be better to think of this mechanism as saying instead “believe now.”

      Can We Ever Trust Our Memories?

      In many cases, memory distortions and embellishments are minor matters, but in some contexts they have tremendous consequences, precisely because of the illusion of memory. When people are subject to the illusion of memory, they impugn the intentions and motivations of those who are innocently misremembering. The power of this illusion was revealed in a crucial incident in the 2008 presidential campaign. Running against Barack Obama for the Democratic nomination, Hillary Clinton repeatedly emphasized her greater experience in international affairs. In a speech at George Washington University, she described a particularly harrowing March 1996 mission to Tuzla, Bosnia: “I remember landing under sniper fire. There was supposed to be some kind of a greeting ceremony at the airport, but instead we just ran with our heads down to get into the vehicles to get to our base.” Unfortunately for Clinton, the Washington Post