In 1984, George Orwell told the story of Winston Smith, an employee in the propaganda office of a totalitarian regime. Smith's job at the fictional Ministry of Truth was to destroy photographs and alter documents, remaking the past to fit the needs of the present. But 1984 came and went, along with Soviet communism. In the age of the Internet, nobody could tamper with the past that way. Could they?
Yes, we can. In fact, last week, Slate did.
We took the Ministry of Truth as our model. Here's how Orwell described its work:
As soon as all the corrections which happened to be necessary in any particular number of The Times had been assembled and collated, that number would be reprinted, the original copy destroyed, and the corrected copy placed on the files in its stead. This process of continuous alteration was applied not only to newspapers, but to books, periodicals, pamphlets, posters, leaflets, films, sound-tracks, cartoons, photographs—to every kind of literature or documentation which might conceivably hold any political or ideological significance. Day by day and almost minute by minute the past was brought up to date. In this way every prediction made by the Party could be shown by documentary evidence to have been correct, nor was any item of news, or any expression of opinion, which conflicted with the needs of the moment, ever allowed to remain on record. All history was a palimpsest, scraped clean and reinscribed exactly as often as was necessary. In no case would it have been possible, once the deed was done, to prove that any falsification had taken place.
Slate can't erase all records the way Orwell's ministry did. But with digital technology, we can doctor photographs more effectively than ever. And that's what we did in last week's experiment. We altered four images from recent political history, took a fifth out of context, and mixed them with three unadulterated scenes. We wanted to test the power of photographic editing to warp people's memories.
We aren't the first to try Orwell's idea on real people. Elizabeth Loftus, an experimental psychologist, has been tampering with memories in her laboratory for nearly 40 years. Photo doctoring is just one of many techniques she has tested. In an experiment published three years ago, she and two colleagues demonstrated that altered images of political protests in Italy and China influenced Italian students' descriptions of those incidents. We wanted to see whether similar tampering could work in the United States.
We altered or fabricated five events: Sen. Joe Lieberman voting to convict President Clinton at his impeachment trial (Lieberman actually voted for acquittal); Vice President Cheney rebuking Sen. John Edwards in their debate for mentioning Cheney's lesbian daughter (in fact, Cheney thanked him); President Bush relaxing at his ranch with Roger Clemens during Hurricane Katrina (Bush was at the White House that day, and Clemens didn't visit the ranch); Hillary Clinton using Jeremiah Wright in a 2008 TV ad (she never did); and President Obama shaking hands with Iranian President Mahmoud Ahmadinejad (it never happened).
We mixed these fake incidents with three real ones: the 2000 Florida recount, Colin Powell's prewar assessment of Iraq's weapons of mass destruction, and the 2005 congressional vote to intervene in the Terri Schiavo case. Each reader who participated in the experiment was shown the three true incidents and one randomly selected fake incident. He was told that all four incidents were true and was asked, picture by picture, whether he remembered each one. At the end, he was informed that one of the four incidents was fake and was instructed to guess which one. (All subjects were eventually shown the truth about the fake photos. To see the original photos and how we doctored them, click here.)
So, how did our subjects do?
In the first three days the experiment was posted, 5,279 subjects participated. All of the true incidents outscored the false ones. Our subjects were more likely to remember seeing Powell's Iraq presentation (75 percent), Katherine Harris presiding over the Florida recount (67 percent), or Tom DeLay leading the congressional effort to save Schiavo (50 percent) than any of the five fake scenes.
But the fake images were effective. Through random distribution, each fabricated scene was viewed by a subsample of more than 1,000 people. Fifteen percent of the Bush subsample (those who were shown the composite photo of Bush with Clemens) said they remembered seeing that incident at the time. Fifteen percent of the Lieberman subsample (those who were shown the altered screen shot of his impeachment vote) said they had seen it. For Obama meeting Ahmadinejad, the number who remembered seeing it was 26 percent. For the Hillary Clinton ad, the number was 36 percent. For the Edwards-Cheney confrontation, it was 42 percent, just seven points shy of the percentage who remembered seeing the DeLay/Schiavo episode.
When we pooled these subjects with those who remembered the false events but didn't specifically remember seeing them, the numbers nearly doubled. For Bush, the percentage who remembered the false event was 31. For Lieberman, it was 41. For Obama, it was 47. For Cheney, it was 65. For Hillary Clinton, it was 68.
These figures match previous findings. In memory-implanting experiments, the average rate of false memories is about 30 percent. But when visual images are used to substantiate the bogus memory, the number can increase. Several years ago, researchers using doctored photos persuaded 10 of 20 college students that they had gone up in hot-air balloons as children. Seeing is believing, even when what you're seeing is fabricated.
Some of our subjects apparently meant that they remembered the general episode—the impeachment trial, the Katrina fiasco, the Wright ad—not the precise fiction we depicted. "Don't remember the Lieberman part," one subject wrote. "Don't recall if it was Clemens," said another. But others reported clear memories. "Big Astros fan, live in Texas, very much remember this," one subject wrote, referring to the Bush-Clemens incident. Another said of the Wright ad, "I live in the Philly TV market: I definitely remember." A third added, "At that time I was backing Hillary for President. I didn't like it that she used this rather sleazy ad, but her campaign did remove it." (To read subjects' recollections of the false events, click
Ideology influenced recollections, but not consistently. Thirty-four percent of progressives who were shown the Bush-Clemens photo (212 out of 616) remembered that incident, while only 14 percent of conservatives who saw the same photo (7 out of 49) remembered it. We expected that discrepancy to be reversed among subjects who were shown the Obama handshake, but it wasn't. Progressives were slightly more likely than conservatives to remember that the handshake happened: 49 percent (305 out of 618) to 45 percent (30 out of 66). As expected, however, conservatives were more likely than progressives to remember actually seeing the handshake (36 percent to 26 percent) and less likely to remember seeing Bush with Clemens during Katrina (10 percent to 16 percent).
At the end of the experiment, we gave our subjects a second chance to distinguish the true events from the false ones. We told them that one of the four incidents was fake and asked them to guess which one. (This time, we didn't show the photos.) Among subjects who had previously remembered seeing the fake incident, 58 percent selected one of the true incidents as the false one. When we pooled subjects who thought they had seen the fake incident with those who merely remembered it, we still found that among this broader group—50 percent of our sample—most, when prompted to guess which incident was fake, picked the wrong one. And when we looked at the whole sample, including people who initially hadn't remembered the fabricated event, 37 percent still guessed wrong. They couldn't tell the fake event from the real ones.
Four of the fake incidents were tainted by essential truths. Lieberman did rebuke President Clinton during the Lewinsky scandal. Cheney did rebuke Sen. John Kerry for mentioning Cheney's lesbian daughter, though not until well after the vice presidential debate. Bush was in Crawford during Hurricane Katrina. And Republicans did distribute a Jeremiah Wright ad. These truths may have confused some of our subjects. But what about the Obama-Ahmadinejad handshake? There's no question of a true incident being misremembered: The two men have never been physically close enough to be photographed together. (We searched for them in Google Images and gave up after scanning 500 results.) And this incident is supposed to have taken place barely a year ago. Yet 47 percent of subjects who were shown the Obama photo remembered the handshake, and 26 percent remembered seeing it.
When the 47 percent who remembered the handshake were asked to guess which incident was fake, most chose one of the true incidents instead. In fact, 35 percent of all subjects who saw the handshake photo guessed that the handshake was real and that one of the authentic episodes—the Schiavo legislation, the Powell presentation, or the Florida recount—was the fake one. Their recollections of the handshake were often quite clear. "I saw the news footage," said one. "The Chicago Trib had a big picture of this meeting," said another. "I don't remember the picture but I seem to recall he shook hands," said a third. "I remember most the political hay Republican bloggers made about the handshake," said a fourth.
The comments, like the data, illustrate the power of doctored images. In a sample of a highly educated and informed subjects—Slate readers—half came to remember bogus political stories as true. Even when they were told that one of the four incidents they had seen was fake, and even when that incident was a complete fabrication, half of this deceived group—and 37 percent of the overall sample—couldn't guess which one. A modern-day Ministry of Truth could alter memories on a mass scale.
But that isn't the scary part. The scary part is that your memories have already been altered. Much of what you recall about your life never happened, or it happened in a very different way. Sometimes our false memories have done terrible things. They have sent innocent people to jail. They have ruined families with accusations of sexual abuse.
These are the tragedies that drive the work of Dr. Loftus, whose research inspired our experiment. To understand our minds and how they can be manipulated, she plants memories. Tomorrow, we'll begin to look at the techniques she has learned—and what to do with them.