By Yana Weinstein

Early on, before cognitive psychologists started researching the processes involved, memory was often described with a “library” analogy. This is the idea that memories are put down in our minds as though they were written down in books, and stored away neatly in organized locations. If we wanted to retrieve a memory, we would go down the relevant aisle and select the appropriate book. If we can’t quite retrieve the memory, the words printed in the books may have faded with time, and if we can’t find the memory at the specified location, perhaps it was like a library book getting misplaced. But is this analogy a good one?

I am going to tell you right now to forget this seemingly innocent metaphor. Many studies have shown that this is not at all the way memory functions. We don’t lay down objective, definitive memory traces that are later retrieved verbatim. Instead, memory is reconstructive (1). This is a key concept in long-term memory: the idea that every time you retrieve a memory, you are actually changing it. Every time you tell the same story, it comes out a little more polished, with a few embellishing details added, or a few boring ones removed. The memory itself – not just the story – is changing, so that the next time you retrieve the memory of that event, it will be more like the story you last told, rather than the way it really was.

Here’s a concrete example of memory reconstruction, first demonstrated almost 100 years ago (2).

In this demonstration, one person was shown an ambiguous drawing (in the top left of the picture). This person was then asked to try to reproduce the drawing from memory. Because we like to classify things into categories rather than dealing with unknown objects (3), the person who was trying to draw the ambiguous original drawing from memory drew it to look kind of like an owl. Their drawing was then shown to another person, who again was asked to reproduce it from memory, and this cycle continued. As you can see, the drawing gradually evolved from an owl to a cat! Each time someone re-drew the picture from memory, they changed it, and eventually even the animal itself changed.

The fact that memory is reconstructive necessarily means that memory is not objective. Our memories are a lot more approximate and less accurate than you might like to believe. In particular, we are prone to having “false” memories – these are memories of things that never happened, or happened quite differently to the way we remember them (4). In addition, since we see the world through our own unique filter – our world view – we tend to remember things in a way that fits our “schema”, or pre-determined categorizations of the world and how objects and people behave (5).

The idea that we can have memories that are “false” gained credibility starting in the 1970s and was a topic of hot debate for decades after that. The leader of this field is Elizabeth Loftus, who demonstrated that eyewitness testimony can be inadvertently affected by information encountered after the event (“misleading post-event information”). Imagine you are a witness to a crime. It was dark, but you think you saw a tall man in a mask, holding something in his hand. After this event, you are questioned repeatedly by the police, and you also discuss the event over and over with your friend who was standing next to you when this happened.

Elizabeth Loftus revolutionized our understanding of how eyewitness testimony works. She demonstrated in numerous experiments that by the time you’ve had all those conversations with the police and your friend, your memory of the crime in the above story will have become a mixture of (a) what you actually saw, (b) what you told people you saw, and (c) what other people told you they saw or think you should have seen. So, for instance, if you were asked over and over about a weapon, you might come to imagine one and believe that the suspect really was carrying a weapon, even if your original memory of the event did not include a weapon. The details from your imagination will become part of the new memory for the event.

Back when I started my PhD in 2005, I was idealistic. I thought that I could figure out a way to distinguish between true and false memories using an implicit test of memory (see this glossary for an explanation of that term).

Many disappointing experiments later (6), I gave up and moved on to the more promising task of improving education through cognitive psychology research. Meanwhile, the perfect test to distinguish between true and false memories is yet to be found (7). What do you think: Will we one day have a tool to determine whether a memory is true or false? 

References:

(1)  Schacter, D.L. (2015). Memory: An adaptive constructive process. In D.Nikulin (Ed.). Memory in recollection of itself. New York: Oxford University Press.

(2)  Bartlett, Frederic Charles (1995 [1932]). Remembering: A study in experimental and social psychology. Cambridge: Cambridge University Press.

(3)  Harnad, S. (2005). To cognize is to categorize: Cognition is categorization. Handbook of Categorization in Cognitive Science, 20-45.

(4)  Loftus, E. F., & Pickrell, J. E. (1995). The formation of false memories. Psychiatric Annals, 25, 720-725.

(5)  Tversky, B., & Marsh, E. J. (2000). Biased retellings of events yield biased memories. Cognitive Psychology, 40, 1-38.

(6)  Weinstein, Y., & Shanks, D. R. (2008). Perceptual representations in false recognition and priming of pictures. Memory & Cognition, 36, 1415-1428.

(7)  Bernstein, D. M., & Loftus, E. F. (2009). How to tell if a particular memory is true or false. Perspectives on Psychological Science, 4, 370-374.