Yesterday, Re-written: The Science Behind Memory

Imagine your freedom depended on someone's memory. 

Would you trust it?

"Memory is the diary we all carry about with us." - Oscar Wilde

Except, unlike a diary, your memories don't always stay the same each time you open them.

We like to think of memories as faithful records of the past. But research tells us something far more interesting:

Every time you recall a memory, you risk subtly rewriting it.

Memory isn't a record of the past; it's your brain's latest interpretation of what happened, shaped by who you are today.

When we recall a memory, it can briefly become unstable before being stored again, a process called memory reconsolidation. During this window, new information can be added, details can fade, and memories can change. 

Think of it like opening a saved Word document. We don't simply read it. Once it's open, you can accidentally delete a sentence, add a paragraph or change a few words before saving again.

To be clear, this does not mean that all memories are susceptible to change each time they are recalled. Instead, whether a memory enters reconsolidation depends on both the characteristics of the memory itself (e.g., its strength and age) and the conditions under which it is retrieved (e.g., the duration of reactivation, retrieval context, stimulus relevance, and whether retrieval involves prediction error or novel information). 

Memories are more likely to become temporarily malleable when recall includes new or unexpected information that conflicts with prior expectations, whereas older or well-established memories are often more resistant to change (Haubrich & Nader, 2016).

The stop sign that became a yield sign

One of the most famous demonstrations of this involved participants watching a slideshow of a car driving through an intersection where there was a stop sign (Loftus, Miller & Burns, 1978). 

Later, researchers asked some participants: 

"How fast was the car going when it passed the yield sign?"

Just one misleading word. 

That single misleading word was enough for many participants to later "remember" seeing a yield sign instead. 

This is known as the misinformation effect: information introduced after an event can become woven into the memory itself. One proposed mechanism that enables this change is memory reconsolidation, during which retrieved memories can become temporarily malleable before being stored again.

It gets worse when we remember together

Now imagine witnessing a crime alongside someone else.

Surely discussing what happened would improve both your memories?

Surprisingly, the opposite happens.

In one experiment, participants were paired and together watched what they believed was the same video. Unbeknownst to them, differently polarised glasses meant each person actually saw a slightly different version containing eight altered details (Garry, French, Kinzett, & Mori, 2008).

After discussing what they had seen, they completed an individual memory test. 

The results were striking. 

For details they hadn't discussed, participants were correct 79% of the time.

For details they had discussed, accuracy dropped to 34%.

Rather than reinforcing the truth, they unknowingly contaminated each other's memories.

Why eyewitness testimony isn't as reliable as we think

These findings have profound consequences in the courtroom.

One of the most concerning aspects of memory distortion is that confidence and accuracy do not always go hand in hand. In fact, repeated questioning, confirming feedback, or exposure to supporting information can increase a person's confidence in a memory, even when that memory has become less accurate - a phenomenon known as false confidence.

Another challenge is source confusion: our tendency to forget where a memory came from. Did we witness it ourselves? Hear it from another witness? See it on the news?

Over time, those sources blur together.

As witnesses are interviewed repeatedly, discuss events with others, or encounter media coverage, each recall creates another opportunity for memories to be subtly reshaped.

The consequences can be devastating.

According to the Innocence Project, as of the date of this article, mistaken eyewitness identification has contributed to around 69% (252 of 367) of DNA exoneration cases in the United States, making it the leading contributing factor of wrongful conviction. 

One of the most well known examples is Ronald Cotton, who spent more than a decade in prison after being confidently, but mistakenly, identified by a victim. DNA evidence later proved his innocence and identified the real perpetrator.

Practical implications for eyewitness testimony

If memory can be changed by what happens after an event, then protecting eyewitness evidence begins long before a courtroom.

Investigators should collect witness accounts as soon as possible, before conversations with other witnesses, media coverage, or repeated questioning introduce new information. 

Interviews should use open-ended questions, and identification procedures should minimize suggestion and bias. For example, suspect lineups should follow scientifically supported practices, such as blind administration, where the person conducting the lineup does not know who the suspect is, reducing the risk of unintentionally influencing the witness. 

A witness's confidence should also be recorded at the time of the first identification, because confidence can increase over time - even when accuracy does not.

For courts, the key lesson is simple: a confident memory is not always an accurate memory. Eyewitness testimony should be evaluated alongside the conditions in which the memory was formed, retrieved, and potentially altered.

Understanding memory as a reconstructive process does not mean eyewitness accounts are useless; it means they must be collected carefully and interpreted with the same caution we apply to any evidence that can be changed.

The takeaway

The science of memory reveals something deeply human: our minds are not passive recorders of events. They are active meaning-makers, constantly reconstructing our experiences.

By allowing memories to evolve, our brains can integrate new information and learn from experience.

So the next time you're absolutely certain you remember something exactly as it happened, pause for a moment. 

The memory you're trusting may not be the original version. 

It may simply be the latest edit.