People with superhuman memories still mistake fantasy for reality

When I was young, the one superpower I craved above all was perfect memory. I’d picture my eyes as camcorder lenses, recording everything that I read, saw and experienced into the Kodak film that was my brain. Anytime I wanted to re-experience something, I’d simply hit a mental “play” button and BAM! The video of my life would play before my eyes, as clear and detailed as the day it was created.

Reality or fantasy? Source: http://www.raanetwork.org/

Little did I know that for those with Highly Superior Autobiographical Memory (HSAM), my fantasy was their reality. Researchers from University of California reported the first case of this phenomenal ability in 2007 when a lady dubbed AJ reached out to them – with a desperate plea for help – with the following email (experts):

Dear Dr. McGaugh,
As I sit here trying to figure out where to begin explaining why I am writing you and your colleague (LC) I just hope somehow you can help me.
I am thirty-four years old and since I was eleven I have had this un-believable ability to recall my past...I can take a date, between 1974 and today, and tell you what day it falls on, what I was doing thatday and if anything of great importance (i.e.: The Challenger Explosion, Tuesday, January 28, 1986) occurred on that day I can describe that to you as well...Whenever I see a date flash on the television (oranywhere else for that matter) I automatically go back to that day and remember where I was, what I was doing, what day it fell on and on and on and on and on.
... Most (people) have called it a gift but I call it a burden. I run my entire life through my head every day and it drives me crazy!!!…

AJ was different than any “memory experts” that the researchers had previously encountered. Her ability seemed completely innate; she did not use little tricks (“mnemonics”), such as mental imagery or story telling to help her remember. Her superior memory only pertained to specific aspects of her own life story and related events; she had trouble learning historical dates and reciting poems. Like your average Joe, she couldn’t remember what each of the 5 keys on her keychain was for. Although specific dates triggered recall in an automatic, “non-stop, uncontrollable and totally exhausting” manner, when asked if she’d talked about a particular date with the researchers in previous interviews, she said she couldn’t remember.

AJ, and the handful of people whom had come forward with HSAM since then, pose a troubling dilemma to memory researchers. Recalling an event is nothing like watching a video recording. Instead, it is an active reconstructive process prone to distortions from misinformation. For example, in labs, researchers have been able to manipulate people into remembering events they’d only previously imagined or trick them into recalling that they’d watched a news video that did not exist. In courts, “corrupt” memory plague eyewitness reports. Elizabeth Loftus, an expert in memory research, had even suggested that planting fake memories in children may be a successful way to modify delinquent behaviour.

Contamination seems like an inevitable part of our memory process, yet common sense suggests that those with HSAM would be spared. Their memories, instead of malleable and fluid, should be etched in stone. But is this really the case?

Researchers from UC Irvine recruited 20 HSAM participants and compared them to 38 age- and sex- matched controls. In the first test, they asked the volunteers to remember a string of related words, such as “light”, “shade” and ”table”. What was missing from the list was the highly similar lure word “lamp”. During the test, when volunteers were asked whether they had seen a particular word in the list, around 70% of the control group (light grey, left graph) falsely remembered seeing the lure. Incredibly, HSAM volunteers (dark grey, left graph) did not fare any better.

Verbal memory is hardly autobiographical. Next, researchers showed the volunteers photos of two separate crimes – a man stealing a wallet and putting it into his jacket pocket, and a man jacking a car with a credit card. Forty minutes after the picture show, volunteers read 50 sentences describing each crime. Unbeknownst to them, 3 sentences included misinformation about minute details, such as “he put the wallet in his pants pocket” or “he used a clothes hanger to break in”. When tested an hour later, the volunteers not only misremembered details but also attributed them to the incorrect source. Astonishingly, those with HSAM were far from immune, generating significantly more false memories than controls.

Again, you may say, the above events did not happen to HSAM individuals themselves. Nor were they particularly disturbing – perhaps volunteers simply didn’t care enough to pay attention to detail? Emotionally charged memories tend to demand attention and stick around for longer than neutral ones; like most people, I have no trouble recalling what I was doing when I first heard of the plane crashes on September 11, 2001 (whether my recall is accurate, however, is another question).

Researchers tapped into this powerful memory. Specifically, they asked what the volunteers remembered about the crash of United Airline Flight 93 with a questionnaire. Hidden in the questions were sprinkles of misinformation and flat-out lies -for example, “a witness had filmed the crash on the ground and the film was subsequently aired”, when no such footage actually existed.

Fig 3D and 3E from paper. Dark grey=HSAM, light grey=control. High PEQ=better memory on initial screening test.

As you can see in the graph above, later in an interview, roughly ~20% of HSAM participants (D left, dark grey) said they had seen the footage compared to 29% of controls (D left, light grey). The percentages were roughly the same if the fabricated video was mentioned off-hand by researchers during the interview (not shown). Even more mind-boggling is this: in the HSAM participants, those tested at top 50% for autobiographical memory accuracy performed worst; in fact, they were just as bad as the controls (D right, high PEQ=better autobiographical memory on initial test). The participants weren’t reporting vague feelings of “having seen the video” either; when probed further, they gave elaborate details:

Interviewer: OK, Can you tell me what you remember about the footage?
HSAM: Uh, I saw it going down. I didn’t see all of it. I saw a lot of it going down uh, on air.
Interviewer: Ok, do you remember how long the video is?
HSAM: Just a few seconds. It wasn’t long. It just seemed like some-  thing was falling out of the sky. It was probably was really fast,   but I was just, you know, kind of stunned by watching it you know, go down.
Interviewer: Ok, so now is the last question, I would like for you   tell me how well you can remember having seen the video on the scale from 1 to 10, where 1 means no memory at all and 10 means a very     clear memory?
HSAM: I’d say about 7.

So here’s the conundrum: we have a cohort of people with astonishingly accurate and detailed autobiographical memory, whom nevertheless are susceptible to misinformation and false memories. Perhaps, you may have already pointed out, it’s due to the small sample size. While certainly a problem in most studies, here researchers have consistently found false memories in HSAM individuals, thus proving the point that fallible memory is (somewhat) inescapable. In other words, increasing sample size would not reduce misremembering to zero percent and thus would not disprove their conclusion (it may help tease apart differences between low- and high- performing HSAMs though). The results would have been more difficult to interpret if they had not found false memories in those with HSAM.

The second problem is that the crash of United 93 is semiautobiographical in nature. The volunteers witnessed the event vicariously through multimedia rather than experienced them first-hand. Previous experiments with AJ, our first identified HSAM individual, showed that while “she could quickly, reliably and accurately tell what she was doing on a given date, she couldn’t recall specific events from a videotape the month before”. In other words, her –and presumably, other HSAMs- extraordinary memory are selective. Here the researchers show they could “implant” a fake factoid 11 years after the fact, and some HSAM individuals would incorporate the memory into their original set of memories of the crash. What I would love to know is whether researchers could distort the volunteers’ own experience of the event. That is, whether they can alter an existing autobiographical memory rather than introduce a new one.

If HSAM individual have these common flaws, how are they still capable of remembering trivial details from a decade ago down to a tee? Researchers think it’s because very little misinformation is generally introduced in their daily lives. “No one comes up to them and says March 2nd, 2001 was a Monday not a Friday,” said one researcher to National Geographic. I’m not so sure that’s true – I know every time I go down memory lane with my friends we recall things differently; misinformation abounds. Our personal accounts end up influencing each other’s memory of the event.

Regardless, HSAM is in and of itself a fascinating phenomenon. Along with many other studies that investigated memory distortion in people with typical memory, this paper suggest that fallible memory reconstruction (reactivation and reconsolidation) is a basic and widespread part of human memory, a global phenomenon deeply routed in our neurobiology. No one, not even those with superhuman memory, is spared. But who knows, maybe discoveries of future “memory experts” will prove this wrong.

For those interested in HSAM (how could you not?!), here’s a 60 Minute episode on the topic. Pleading to @radiolab to do a show on it.

Patihis L, Frenda SJ, Leport AK, Petersen N, Nichols RM, Stark CE, McGaugh JL, & Loftus EF (2013). False memories in highly superior autobiographical memory individuals. Proceedings of the National Academy of Sciences of the United States of America PMID: 24248358