William Saletan Explores the Malleability of Memory


William Saletan

In 2010, Slate national correspondent William Saletan wrote an eight-part series about experimental psychologist Elizabeth Loftus and her work on false memories. He began the series by inviting readers to take part in an interactive online experiment designed to illustrate how easily memories can be manipulated. (Try Slate’s experiment yourself here.) Readers were presented with several images depicting recent political events and asked whether they remembered them. Unbeknownst to readers, one of the photos was doctored to show an event that hadn’t happened—President Obama shaking hands with Iranian President Mahmoud Ahmadinejad, for instance.

About half of the 5,000 readers who took part in Saletan’s online experiment later “remembered” the fake political stories as if they were true—until the end, when all was revealed. The experiment served as a powerful introduction to the concept of false memories and to Loftus, who has made a career of studying them. [Saletan’s series, The Memory Doctor, appeared in Slate beginning on May 24, 2010.]

Here, Saletan tells TON guest contributor Christie Aschwanden the story behind the project:

How did this series come about?

David Plotz, the editor at Slate, wants us to do long form journalism, so he’s got each of us doing one of these projects a year. You submit ideas to him, and he’ll choose among them.

I was familiar with Loftus from her testimony in trials. She’s very famous for her legal work, but I went and read her actual journal articles. She was a great resource for this because she keeps all of her stuff on her websites, first at the University of Washington, and now at University of California, Irvine.

What made Loftus a good profile subject?

The thing that really captivated me was the shift in her work as she became more interested in engineering memories. Looking back, the series ended up pivoting on this thing about her becoming a memory engineer, and—from my point of view—her literally forgetting some of the principles that she went in with. The whole gist of her career has been has been about human fallibility—the fallibility of our minds and the fallibility of memory. We think we accurately remember things. We’re so convinced of it that we’re willing to send somebody to jail for it, and yet we make mistakes. You’re pointing and saying yes, that’s the man. He did it. And you don’t have any idea.

So it was fascinating to me that she just didn’t realize that she had made this shift in her work. When I eventually put it to her and asked, “You’re messing with people’s recollections of their eating behavior, and you had specifically warned against this. What were you thinking at the time?” Her answer was: nothing. She hadn’t thought about it at all, and she didn’t see it, and she’s a very smart person and a person who thinks about a lot of things. So that really struck me.

As I read along in your series, I recognized this writerly obsession that you can develop with a subject. Did you become totally consumed with Loftus at some point?

Yeah. That was a little embarrassing, because there was this moment when she told me something to the effect of, you know me better than I know myself. I don’t think she literally meant that, but it was just like I was way too far inside.

One of the funniest things about the whole project—since it’s all about memory—is that we’d have these interviews, and by that time I had read practically everything she had written and I’d find myself correcting her memory about things. She thought it was kind of funny that I would say, “Actually no, that’s not what you said; you said this other thing.” She was sort of amused, because she’s very relaxed about memory—yeah, memory fails and you go back and you check things.

Your series is sympathetic to Loftus, but you also call her out for her shift from warning about the dangers of manipulating memories to conducting research where the explicit aim was to re-engineer memories. Did you find yourself feeling sympathetic to her in a way that made it difficult to write that part of the story?

You couldn’t have these interviews with her and not feel fundamentally sympathetic to her. I got the sense that she was passionately committed to her values. She has spent her whole career fighting the repressed memory people. Once I had read enough about that debate to decide that she was right, and she understood that I felt that way, I think that there was complete trust on her part.

We talked about all this stuff, and I wasn’t sure as we talked exactly how I was going to write about it. I knew I had to talk about the shift to memory engineering, and I knew that I had problems with it. That was hard, because I wasn’t sure how was she was going to see it, after it was written from another person’s point of view, saying there are real problems here.

So what was Loftus’s reaction to what you wrote?

The series ran in segments, and it wasn’t until I think the fifth of the eight parts that it was going to be a problem. After she saw the first one, she emailed me, “I really like this. Oh yeah, this is great.” Up to that point it was fairly consistent with profiles that had been written about her. At some point, I wrote to her and said, “It’s going to get a little bit unpleasant for you.”

She wrote back and said, “How bad is it going to be? Just give me the bad news.” And I said “Look, it’s not that bad. This does not take away from the bulk of your career and all the good stuff you’ve done, and this is just my take on it, but I’m going to raise some problems here, and you’ll see them over the next three or four sections.”

Do you think that warning her like that might have actually softened the blow? Our minds can create these awful scenarios.

I don’t know. I felt weird doing it. I don’t normally like tell a subject, here’s what it’s going to be like. But I had spent so long talking to her that—well, you get compromised. You’re with somebody, you start to identify with them. They start to become your friend. I felt like I was her friend.

I gave her a warning, and then I felt even worse because I figured now she’s imaging the worst, and this is not going to be that bad. I remember feeling like I was putting a friend through a kind of emotional torture that was not necessary. And once she saw it, she said, “Oh this isn’t bad.”

Your series ran in eight parts. How did you structure the narrative to work in these chunks?

I wanted to have a theme in each section and I wanted to do it roughly chronologically. The first four parts of it are roughly chronological and there was a theme in each era of her life and you can just sort of find good breaking points in there. But you have to repeat yourself sometimes just because somebody may come into the series at part four not having read the stuff before, and you’re going to have to summarize a little bit.

The same is true in a long print series. Any time I come across a piece in the Atlantic or the New Yorker that’s really, really long, I can’t sit through the whole thing. I’m going to read it in sections anyway, and I’m grateful that we break things up on the web. I think it’s helpful in terms of telling the reader “Okay, I’m going to ask for 1500 words of your time, maybe 2000 words, but it’s not the commitment of a lifetime. You’re going to be able to read this and go on and do something else, and you can come back that later.” And readers did come back.

What were you able to do with this story online that you couldn’t have done in print?

We have a mandate in these projects to try to include multimedia. So I went looking for all the audio and video I could. I basically Googled everything. And that was important to me because one of the things that you learn from working with a memory expert like Elizabeth Loftus is, don’t trust anyone’s memory. Don’t trust hers. Don’t trust yours. Look for contemporaneous records wherever possible.

I could say, here’s my account of what Elizabeth Loftus wrote in this journal article 12 years ago, but don’t take my word for it. Here’s the link to the whole article. Hopefully the whole series with all the links in it is a good resource for anybody who wants to check up on anything that’s in there.

We took also took video clips that illustrated various things, and we just interspersed them through the series. You could actually see [Loftus] in action, for instance giving a speech.

And then of course there was the experiment.

So let’s talk about the experiment, which essentially introduced readers to the concept of memory manipulation by subjecting them to it. How did this part of the project come together?

Credit goes to David Plotz, who really pushed us to think in terms of multimedia, and got me thinking about ways that we could involve the readers in this. And so I remember thinking about doing a memory experiment with readers. It started out as a small idea.

I had the idea to just rerun this experiment she did with a stop sign, and I thought I could show people what that’s like in a slide show. But then I thought well, we’re a political magazine, mostly. So can I do something with politics? What about running an experiment where we mess with political memories? I drew up a plan and gave it to David. He loved it and wanted me to pursue it.

At one point, we ended up in a conversation with the Washington Post Company lawyers, because we were going to run this experiment in which we’re going to lie to people about the past. We’re a journalistic enterprise, part of a media company whose flagship publication, the Washington Post, is about documenting history, and we are going to misrepresent history in this experiment.

If you’d run the experiment at a university, you would have a debriefing at the end where you tell people, “We lied to you.” We needed to tell people before they left the experiment what the truth was. We told [them] that Joe Lieberman voted for Clinton’s conviction, but he didn’t.

We were on the phone with the lawyers trying to get chapter and verse about exactly what we needed to do to be ethically okay. I basically took it from there and designed the whole experiment and chose the scenes that we were going mess with. We set it up so that people were told about the deception before their data were counted.

Was it technically difficult to execute?

Yeah. Our art director helped me design it. The art people were given the assignment to fake the photographs. The first version of the one of Obama shaking Ahmadinejad’s hand, I said, “No, this is just not going to cut it.” But the second attempt at Ahmadinejad was magnificent. I stood up out of my chair because that was such a good fake. And you really needed it to sell that one, because everything else had a little bit of truth in it, and that was a completely fabricated incident.

Rejected: The first version of the fake Obama-Ahmadinejad photo. Courtesy of Slate.

A better fake: This is the photo Slate ran in its memory “experiment.” Courtesy of Slate.

Finally, we had gotten to the point where everything was set up and you could do the experiment, but the data were not being properly recorded. The development team could not get it to work, and David said, “Just drop it.” I walked down the hall, and told Chris Wilson, who was an editorial assistant at the time, “I’m really, really sorry. We’re going to have to just drop the experiment.” And Chris just basically said, “No. I refuse to let it die.” He went off on his own after hours, and he studied programming so he could make it work and he did. He saved it.

So what did you learn from the experiment?

We had 5,000 people participate in the first three days. The thing that surprised me was, I was expecting more of a correlation between ideology and which photographs you remembered accurately or not. I was a little surprised that it hadn’t turned out that way. [Full results are here.] I thought you would be politically biased, and so you’d be more likely to remember what you wanted to remember. But the people who called themselves progressive were as likely as conservatives to remember Ahmadinejad’s handshake, which is something I totally expected more conservatives to think was true.

The absolute numbers were very high. I guess I didn’t expect them to be as high as they were. [One photo purported to show Dick Cheney rebuking Sen. John Edwards in their debate for mentioning Cheney’s lesbian daughter, when in fact he’d actually thanked him.] For the Cheney one, 65 percent said they thought it had really happened. I guess in retrospect, I had chosen them for plausibility. And so it may be that because they were related to true incidents, it could be a little bit unclear what people mean when they say yeah, I remember that thing with Edwards and Cheney. I mean, it did happen, just not the way that the caption describes.

Has this project changed the way you work as a reporter? Do you find yourself less willing to trust sources, or more apt to double-check remembered facts?

I came into this with some perspective on that, because I had written a book about abortion where there was this big debate about the origin of this famous phrase, “Who decides?” which became the pro-choice rallying cry. People had conflicting memories of it.

Reading Loftus’s work reinforced to me the importance of checking for corroboration. I know that I have a crappy memory. Her work drove home to me the danger we face as reporters. If you’re interviewing somebody, especially if you’re in person, they’re talking, they’re right to your face. They seem totally credible and their recollection is quite vivid. They’re describing quite clearly what happened and I have to admit that I am much more inclined to assign credibility to that person. What you’re actually doing is you’re buying into the source’s confidence about their memory. It’s contagious. And in some ways, you should be more wary of the memory of someone who is very confident in their memories, because they’re less likely to check. You can be led astray—I know from my experience with that abortion story that that’s true. You really get suckered into that way of thinking.

Christie Aschwanden

Christie Aschwanden

Guest contributor Christie Aschwanden is an award-winning writer whose work has appeared in dozens of publications, including The New York Times, The Washington Post, O, the Oprah Magazine, Health, Skiing, Runner’s World, New Scientist, and many others. She lives in western Colorado. Follow Christie on Twitter @cragcrest.

One Comment

  1. If anyone bought the fake photo, it says less about “fake memory” than sheer cultural ignorance (and this goes for its creators too). Iranians don’t wear ties.

Leave a Reply

Your email address will not be published. Required fields are marked *