The Fallibility of Memory

by Craig Good
December 23, 2014

Take a moment to think of a cherished childhood memory. Try to recall it in detail. Think of where you were, who you were with, the sights, the smells, the tastes. Recall the sounds, like the wind in the trees, and how you felt. Were you happy? Anxious? Laughing? Crying?

We would all like to think that our memory is like a camera that records a scene, tucks it away in a corner of our brain, and retrieves it for playback when we want to relive that birthday ice cream or feel a long lost summer breeze on our cheeks. In a large sense we are what we remember, so memories are an integral part of who we are.

Unfortunately memory isn't even remotely like a record/playback device. As neurologist and renowned skeptic Dr. Steven Novella puts it,

When someone looks at me and earnestly says, "I know what I saw," I am fond of replying, "No you don't." You have a distorted and constructed memory of a distorted and constructed perception, both of which are subservient to whatever narrative your brain is operating under.

As I like to say, we are a story our brain tells itself, and our brains are motivated, skilled, pathological liars.

Lets take a look at memory, get a rough idea of how it works, and learn when and why we need to be cautious about trusting it. Functionally, the three parts of memory are encoding, storage, and retrieval.

As Dr. Novella points out, the problems begin with encoding, even before a memory has been stored. Our brain is constantly filtering information, and constructing its own reality. We are surrounded by detail. Take a moment right now to be aware of every distant sound around you, of all the leaves on the trees, fibers in the carpet, your breathing, and the sensations on your skin — all of it. Imagine dealing with all of that all of the time! Our brains evolved to construct a narrative of what's going on, lending attention to what matters most. That thing over there that might be a predator is a more pressing matter than the sensation of every individual blade of grass you're standing on. But just as things get lost, distorted, or added when your favorite book becomes a movie, the running story your brain puts together isn't a faithful rendition. In fact, sometimes the circuitry in your brain that distinguishes what's currently happening from a memory gets confused. This is the most likely explanation for déja vu. It's a glitch in your own brain's matrix.

And the first thing your brain does with most information is forget it. In 1968 Atkinson and Shiffrin proposed their multi-store model of human memory. While it's come under some criticism, it has stimulated a lot of memory research and makes a handy top-level guide that suits our purposes here. This is where we got the terms short-term memory and long-term memory. The lesser known mode is called sensory memory. This is where information from our senses about the environment is briefly stored. Auditory information may last three or four seconds, and visuals less than a second.

Once we've attended to that brief burst of information, some moves on to our short-term memory. Sometimes called "active memory", this is the information we're currently dealing with and thinking about. Most of the information here is kept for maybe twenty to thirty seconds, and then forgotten. Some information will then move on to long-term memory. That's information that's not in your consciousness, but is available for recall. You could almost think of it as the data on your computer's hard drive.

Except that your memory is less like data on a hard drive than like a cryptic shorthand on a chalk board. Memories can be smudged, overwritten, and just plain fade away. Imagine using a computer that changed the contents of its files every time you opened them. You can already see how what the brain has stored is compressed, distorted data. Even more errors come into play when you try to retrieve it. Put very simply, we have four basic kinds of retrieval:

Recall — This is accessing information without being cued. For example, answering a fill-in-the-blank test.

Recollection — This is when your brain reconstructs a memory using logical structures, partial memories, narratives, and other clues.

Recognition — When you scan the options in a multiple-choice exam and trigger a memory of what the right answer would be, that's an example of recognition.

Relearning — This is one way you can strengthen a memory, namely by learning the information again and storing it.

Many memory errors happen upon retrieval. We've all experienced that "tip of the tongue" retrieval error when we know we know something, but just can't recall it. I know it happens to me with names all the time.

Now, keep in mind that a comprehensive and in-depth review of the science about how memory works is outside the scope of a single Skeptoid episode. What I've presented so far is clearly a simplification. Science is still working out how memory, and other brain functions, actually work. What is well understood, though, are the kinds of errors our memory system makes, and the consequences of those errors.

Remember how I had you recall that cherished childhood memory? Personally, and on behalf of Skeptoid, I hereby formally apologize for forever altering it. If it now has wind involved somehow and didn't before, that's probably my fault.

Psychologist Elizabeth Loftus wrote,

Misinformation has the potential for invading our memories when we talk to other people, when we are suggestively interrogated or when we read or view media coverage about some event that we may have experienced ourselves. After more than two decades of exploring the power of misinformation, researchers have learned a great deal about the conditions that make people susceptible to memory modification. Memories are more easily modified, for instance, when the passage of time allows the original memory to fade.

Every time we access a long term memory, it gets rewritten, with new errors, back into memory. The mere act of remembering something from long ago actually changes that memory. This is a common source of one of the most interesting errors, the false memory.

Skeptoid listener Patrick Moore is an instructor in Dallas, Texas. It was actually his email which prompted this episode. He shared a favorite childhood memory of being taken by his dad to an Arkansas Razorbacks football game.

Before the game a local radio station played several minutes of the sounds of actual hogs snorting. Every Arkansas fan in the stadium tuned their transistor radios to that station at top volume. It was deafening: 60,000 radios tuned to hogs snorting and grunting.

I must have told that story at least a hundred times since then. Then one day at work I told it to a Razorback fan I met. He was a little puzzled. Turns out he was a season ticket holder during the years before and after I would have been there, and never missed a game. The problem was that he'd never heard all the radios tuned to the station playing hog sounds.

Since then I haven't found anybody who remembers the hog sounds. Not even my father. Now, I still have that memory. I can hear it, see it, smell it, feel it. It's a memory I can, and do, still enjoy even though I now recognize it as "self-constructed."

Patrick's experience is hardly unique. Everybody likely has false memories about something, and the more we remember them the falser they get. It's important to note that having a false memory doesn't make you dishonest. People who "remember" paranormal sightings, visits from extraterrestrials, and such are likely being perfectly sincere. That really is what they remember.

Stressful situations are very likely to produce false and distorted memories. I have a very clear one of watching the 9/11 attacks on television, and seeing the top of one tower sit at a precarious angle for several seconds before the building collapsed. Watching the videos since then I know it never happened that way, but I would have sworn in court that's what I witnessed. This is an example of what's called flashbulb memory which has also been tested and found to be "dramatically inconsistent".

More troubling are implanted memories. Elizabeth Loftus famously got people to believe they saw Bugs Bunny at Disneyland, got lost in a mall as children, or became sick on hard boiled eggs. Remember how our brain continually constructs the version of reality we experience? Recalled memories get the same kind of "enhancements". Loftus planted false memories of wounded animals outside a terrorist attack, and witnesses later embellished the memories with all kinds of details. Loftus calls these "rich false memories because they are so detailed and so big."

When pseudoscience meets planted memories, the results can be tragic. So-called "recovered memories" have been called "the most dangerous idea in mental health". Fathers have been separated from their children and placed on public registers as child abusers, all based on false, planted memories. In the 1980's through early 1990's there was a panic about day care child abuse, even Satanic ritual abuse, which led to many innocent lives being ruined. The real abuse turned out to be that of the children by the authorities and putative therapists who implanted horrifying, false memories.

Our malleable memories, combined with confirmation bias, are a key factor in the Dunning-Kruger effect, the inability to perceive one's own incompetence in a given area. Eyewitness testimony, so convincing to juries, can be fairly accurate but very often is not.

Nobody is immune from memory errors. Recently, no less than Neil deGrasse Tyson conflated two different statements by President George W. Bush and ended up criticizing him for something he never said. At least most of our memory conflations don't happen that publicly.

Well. This is all a bit disheartening, isn't it? I mean, if we can't trust our own memories, what can we trust? First, we can take a clear-eyed look at our own shortcomings and come up with ways to compensate. If something important has just happened, write it down while the memory is fresh. Take a photograph or video. Stay humble. When you catch yourself saying, "I know what I saw!" step back and realize that maybe you don't. See if you can corroborate it with other witnesses or records. And, as always, be skeptical — especially about things you are just sure you remember.

By Craig Good