Last post, we described Princeton’s experiment that showed how stories facilitate brain-synchronization between listeners and storytellers. This week, we’ll look at the University’s research into the relationship between stories and the human memory.

The experiment:

  • Use an fMRI machine to observe participants’ brains as they watched a 50-minute episode of BBC’s Sherlock
  • Ask each participant to recall what they saw without cues from the researchers.

The experiment resulted in three conclusions. The first two duplicated what we learned with mirror neurons. The third conclusion, however, was significant.

  1. The episode ignited similar brain areas in each participant
  2. Recalling the movie also ignited similar brain areas
  3. Although the details recalled by each participant differed, group’s collective recall exhibited high correlation

The researchers compared two types of recall: movie-to-recall and recall-to-recall. Movie-to-recall compares what a participant recalled to the actual movie details. Recall-to-recall compares the details of what each participant recalled.

“Strikingly…we found that recall-recall similarity (the details each participant recalled compared with the other recollections) was stronger than movie-recall similarity (the details each participant recalled compared to the actual movie), indicating that the neural representations were transformed between perception and recall in a systematic manner across individuals.”1


In other words, although the movie watchers recalled different things, the ones that they did remember were common. This result is counterintuitive, considering that we filter our perceptions through different experiences and vocabularies.

“A memory is not a perfect replica of the original experience; perceptually representations undergo modifications in the brain before reflection that may increase the usefulness of memory, for example, by emphasizing certain aspects of the precept and discarding others.”2

Therefore, if our memories are imperfect, shouldn’t we expect the imperfections to be random instead of common? The report thinks so.

“If recall patterns were merely a noisy version of the movie events, we would expect the opposite result: lower recall–recall similarity than movie–recall similarity. No regions in the brain showed this opposite pattern. Thus, the analysis revealed alterations of neural representations between movie viewing and recall that were systematic (shared) across subjects.”3

We alter our perceptions based on unique perspectives, worldview, and life experience. However, this experiment revealed that even though we each encode our memories differently, we tend to remember the same things. But why?

A clue is found in the data. The researchers divided the Sherlock episode into fifty scenes. By tallying the data and comparing what people remembered from each scene, a pattern emerged.

Figure C shows how one participant recalled the episode. Note that since the episode lasted 50 minutes (Y-axis) and the participant took over 20 minutes to recall it (X-axis), that some details were lost. Figure D illustrates how all 17 participants recalled the movie. Also, note the wide range of recall times. It took between 10 and 45 minutes to recall the details of the episode. 4

However, although the recall-to-movie details differed—the researchers found that collectively, the group recalled similar elements. This is even more amazing considering that the none of the participants were prompted for their recitations.

So, how did these participants remember the same parts of the story? I propose that the StoryHow PitchDeck™ has the answer.

According to StoryHow PitchDeck Card #50: Scenes are mini-stories.

  • Something changed creating a problem (Then one day)
  • Someone chose an action (And because of that)
  • Which resulted in a consequence (And because of that)

If the consequence doesn’t resolve the problem, it then becomes the initial impulse (then one day) for a subsequent scene. Thus, if a participant wants to remember the Sherlock episode, all they need to do is follow the breadcrumbs of “then one days” and “consequences.”

Supplementary data from the study5 backs up this “Scene Theory.” Consider the following table that summarizes the number of words that four participants used to describe two scenes #13 & #36.

Although the number of details that each participant recalled varied (participant S06 used 37 words to describe Scene #36 compared with participant S17 who used 321), all four participants got the salient points:

  • Scene #13: is a press conference lead by a police officer who’s speculating about the similarity of a rash of suicides. Everyone in the room gets a mysterious text that the theory is wrong. The police officer gets a text from Sherlock Holmes inviting him to talk.
  • Scene #36: John Watson is walking on the street. He ignores two telephone booths with ringing phones, but can’t ignore the third. The voice on the other end tells him to get into the arriving car.

By comparing the common things that each participant recalled, a pattern emerges. Each scene has some sort of mystery (like a rash of suicides or ringing telephone booths) followed by an action (Sherlock’s invitation and Watson getting into a strange car). Since these actions didn’t resolve anything, they become the impetus for subsequent scenes.

By segmenting their data by scene, the researchers found the connection between story and memory. People don’t remember every detail of a story. Instead, we remember the things that have meaning to us. Since we each have different interests, the specific things that we remember differ, as demonstrated by the fact that participant S17 remembered almost 10 times more details of a scene than S06. But, in the end, they both remembered the most important parts of the scene–how it connected to the rest of the story.

These two studies prove something that we’ve always known intuitively: people connect through stories and our memories depend upon them.



  1. Chen, Janice et al. Shared Memories Reveal Shared Structure In Neural Activity Across Individuals. Nature Neuroscience, 2016. p.2
  2. Chen, Janice et al. p9
  3. Chen, Janice et al. p7
  4. Chen, Janice et al. p2
  5. Shared experience, shared memory: a common structure for brain activity during naturalistic recall (Table S2 p. 32)


Photo Credit: Genthe, Arnold, photographer. Arnold Genthe seated outdoors with two friends. , None. Between 1911 and 1942. Photograph. Retrieved from the Library of Congress,