Why Remembering is Inextricably Linked With Imagination

Memory is imagined; it is not real.

Don’t be ashamed of its need to create.

Nick Cave, The Sick Bag Song

One of the most expansive memories ever documented belonged to a Russian newspaper reporter named Solomon Shereshevsky. For much of his life, he was oblivious of the peculiar nature of his memory. Then, in his late twenties, the young reporter’s habit of never taking notes during morning staff meetings caught the attention of the editor of his Moscow newspaper. Shereshevsky told the editor he never wrote anything down because he didn’t need to, then repeated verbatim the long list of instructions and addresses for that day’s assignment. The editor was impressed, but even more interesting to him was that Shereshevsky seemed to think there was nothing unusual about this. Wasn’t this how everyone’s mind worked?

The editor had never seen anything like it, so he sent Shereshevsky to have his memory tested. Shereshevsky then crossed paths with a young researcher, Alexander Luria, at a psychology laboratory at the local university. For thirty years, Luria, who would go on to become one of the founding fathers of neuropsychology, tested, studied, and meticulously recorded Shereshevsky’s remarkable ability to quickly memorize made-up words, complex mathematical formulas, even poems and texts in languages he didn’t speak. Even more astonishing than his ability to recall much of this information with the same accuracy many years later was that Shereshevsky could remember what Luria was wearing on the day he had administered a particular memory test. In his classic 1968 monograph, The Mind of a Mnemonist: A Little Book About a Vast Memory, Luria wrote, “I simply had to admit that the capacity of his memory had no distinct limits.”

Luria linked Shereshevsky’s remarkable capabilities to an extremely rare condition called synesthesia—meaning that every stimulus, regardless of which sense it came through, triggered every other sense. Shereshevsky could taste words, see music, and smell colors—even the sounds of words could impact his perception. He described asking an ice cream vendor what flavors she had. Something about her tone of voice made him see a stream of black cinders pouring from her mouth as she spoke, “Fruit ice cream”—which promptly ruined his appetite.

The connection between the worlds he created in his mind and the world he lived in was so visceral that Shereshevsky could elevate his heart rate by simply imagining he was running after a train. He could raise the temperature of one hand and lower the other by picturing one hand on a stove and the other resting on a block of ice. The distinctiveness of Shereshevsky’s sensory world extended to his imagination, giving him the ability to form distinctive memories that were resistant to interference.

New Yorker writer Reed Johnson, who spent years researching Shereshevsky, described how he was able to attach any memoranda, no matter how bland, to stories he conjured in his imagination, which he could follow like a trail of bread crumbs to find his way back to that information later on. The strength and durability of his memories seemed to be tied up in his ability to create elaborate multisensory mental representations and insert them in imagined story scenes or places; the more vivid this imagery and story, the more deeply rooted it would become in his memory.

In his later years, when Shereshevsky began performing his incredible memory feats for a paid audience, he amplified this ability with a technique familiar to modern memory athletes such as Scott Hagwood and Yänjaa Wintersoul. Though it appeared to have been self-discovered rather than learned, Shereshevsky used a memory device similar to the method of loci. When he wanted to remember a sequence of words or numbers, he would visualize them as characters within the familiar schema of, say, a street in Moscow and take a “mental walk” through the vast worlds of his interior landscapes.

Although he is often discussed as an example of someone with an extraordinary memory, the key to Solomon Shereshevsky’s mnemonic capabilities was his vibrant imagination. Much of Luria’s decades-long study reveals a fundamental truth about the connection between memory and imagination, one that lies at the center of how we all remember. In this chapter, we’ll explore how the peculiar way in which we form memories can lead us to stray far from reality, yet gives us the fuel to imagine a world with endless possibilities.

WHAT CAN HAPPEN

The simplest way to see the machinery of episodic memory at work is to scan people’s brains while they describe an event from their lives. For instance, if you showed me the word photograph while I was lying in an MRI scanner and asked me to use that word to help me recount an event from my life, I might pull up the memory of my first live rock concert.

At fourteen, I was obsessed with the album Pyromania by British heavy metal band Def Leppard. If you examined my brain activity while I recalled seeing Steve Clark play the signature riff during the band’s performance of “Photograph,” you’d see activation in the hippocampus, as I pulled up the contextual information that mentally transported me back to 1985, and in the DMN, as I brought up knowledge about concerts that enabled me to elaborate on how the event had unfolded.

Now, let’s try something a little different. Suppose you were lying in an MRI scanner and I showed you words such as pasta or skydiving and asked you to use those words to imagine something that hasn’t happened, or even something that would be unlikely ever to happen. You might conjure up a mental image of cooking spaghetti with Motown legend Marvin Gaye or jumping out of a plane with pioneering physicist Marie Curie.

In 2007, three research labs published experiments along these lines, and here’s the twist: the brain activity changes that occur when people imagine these kinds of scenarios are remarkably similar to those that occur when people recall events that they actually experienced. This odd parallel between imagination and memory came as a surprise to many in the scientific community, and it captured the attention of the media - Science magazine declared it one of the top ten breakthroughs of the year - but it did not come out of the blue.

It was anticipated nearly a century earlier by English psychologist Sir Frederic Bartlett, whose work would become the foundation for the idea that we use mental frameworks (i.e., schemas) to organize and process the world around us. Bartlett began his research on human memory as part of his dissertation at the University of Cambridge in 1913. After receiving his PhD, he focused not on memory but on cultural anthropology and then on applications of psychology for the military. Fortunately, Bartlett eventually circled back to the topic of memory and, in 1932, published his most important work, Remembering: A Study in Experimental and Social Psychology.

Bartlett’s book was a dramatic departure from the tradition of memory research established by Hermann Ebbinghaus back in 1885. Ebbinghaus quantified memories for strange, meaningless information under strictly controlled conditions. In contrast, Bartlett drew on his experiences in applied psychology and anthropology, observing and describing how we use memory in our everyday lives. Put more succinctly, Bartlett was interested in understanding how we remember, rather than simply quantifying how much.

In his most famous experiment, Bartlett introduced a group of volunteers at the University of Cambridge to a Native American folktale called “War of the Ghosts”—specifically chosen because the cultural context was entirely foreign to these British students. Bartlett’s subjects could recall the gist of the story, but they made some characteristic errors. It was not simply a case of failing to remember some details, but rather that they adapted the details to match up with their own cultural expectations and norms. Words such as canoe and paddle were replaced by boat and oar; seal hunting became fishing.

Poring over these results, Bartlett observed that, although people do recall some details from the past, their recollections are approximate at best. He concluded, “Remembering is not the re-excitation of innumerable fixed, lifeless and fragmentary traces. It is an imaginative reconstruction.” We do not simply replay a past event, but use a small amount of context and retrieved information as a starting point to imagine how the past could have been. We put together a story on the fly, based on our personal and cultural experiences, and tack on those retrieved details to flesh out the story. Bartlett’s insight is key to understanding why the brain’s machinery for imagination and its machinery for memory aren’t completely independent—they are both based on pulling up knowledge about what can happen, though not necessarily what did happen.

FABLES OF THE RECONSTRUCTION

The reconstructive nature of memory means that our recollections can sometimes take on a life of their own. Consider the case of former NBC news anchor Brian Williams. Speaking at a New York Rangers hockey game in 2015, Williams recounted how, in 2003, he and his crew were in a helicopter in Iraq that was forced down by a rocket-propelled grenade. His account was promptly debunked by several veterans who had been there. Williams had never encountered enemy fire during his visit to Iraq—though he certainly found himself at the center of a firestorm of controversy after what seemed like a blatant lie.

In reality, Williams and his crew were flying about an hour behind a convoy of three military helicopters, one of which was hit by a rocket-propelled grenade; all three were to make an emergency landing in the desert. Williams’s helicopter eventually caught up to the other three, and a sandstorm stranded them in the desert together for several days. Although elements of Williams’s 2015 account overlapped with what really happened, the story he told twelve years later didn’t belong to him but to the soldiers in the helicopter that had gone down.

Williams apologized, chalking it up to the “fog of memory,” but the damage was done. Widely suspected of lying to enhance his reputation, Williams’s integrity as a journalist was tarnished. He was suspended without pay for six months and ultimately stepped down from his role at NBC Nightly News. I can’t say whether Brian Williams intentionally embellished his story, but if we give him the benefit of the doubt, it appears that he recollected many of the right pieces, yet reconstructed a narrative of the event that was fundamentally wrong. His dramatic account was an imaginative reconstruction gone awry.

Most of the time, our memories do not stray as far off the mark as Williams’s Iraq anecdote, but ample scientific evidence suggests that we can confidently remember stuff that didn’t happen. In 1995 Henry Roediger III and Kathleen McDermott of Washington University, St. Louis, demonstrated this phenomenon in an experiment that is now taught in almost every introduction to psychology class. They had volunteers study lists of words such as this: FEAR TEMPER HATRED FURY HAPPY ENRAGE EMOTION RAGE HATE MEAN IRE MAD WRATH CALM FIGHT.

Now, indulge me for a moment. Without looking back at the list, what words do you recall seeing? Do you remember reading the words fear and wrath? What about the word anger? If you remembered seeing the latter, you would be mistaken—but you wouldn’t be the only one. In fact, the people who participated in Roediger and McDermott’s study were just as likely to remember seeing anger as they were to remember words that they had actually studied, such as fear and wrath. This finding is often used by scientists to describe how people are susceptible to “false memories,” and the potential implications are staggering.

It’s easy to read about studies like this and go down a rabbit hole of self-doubt, wondering whether our most cherished memories might be entirely false. But that’s not quite the right way to think about it.

As former Sex Pistols / Public Image Ltd singer John Lydon put it, “I don’t believe in false memories, like I don’t believe in false songs.” In the real world, memory can’t easily be reduced to simple, black-and-white dichotomies such as strong or weak, true or false. When people with a healthy brain get a little feeling of mental time travel and say, “I definitely remember that,” they are probably pulling up something from their past. However, even when the elements of what we remember are true, the entirety of the story can be false.

Roediger and McDermott’s experiment was explicitly designed to encourage people to think about anger, or maybe even think about a moment in their lives when they felt angry. So, when their subjects (or you) recalled that the word anger was among those on the list, it was, in some sense, a real memory, but one that was reconstructed incorrectly. Perhaps this is also what happened to Brian Williams.

Not everyone is equally susceptible to generating false memories. Individuals with amnesia generally don’t fall for the trick of recalling anger because they aren’t able to recall enough information to make a reconstructive error. Some research suggests that people with autism or other neurodevelopmental disorders are also more resistant to false memories because they sometimes remember events in a concrete manner, focusing on the details at the expense of the meaning. These studies indicate that complete resistance to false memories might come at the expense of the capability to generate meaningful reconstructions of the past.

All this work points to a deeper realization about memory. When we talk about memories as being “true” or “false,” we are fundamentally mischaracterizing how memory works. I like to think of memory as less like a photograph and more like a painting. Most paintings typically include some mixture of details that are faithful to the subject, details that are distorted or embellished, and inferences and interpretations that are neither absolutely true nor entirely false, but rather a reflection of the artist’s perspective. The same is true of memory. Memories, we will see, are neither false nor true—they are constructed in the moment, reflecting both fragments of what actually transpired in the past and the biases, motivations, and cues that we have around us in the present.

FILL IN THE BLANKS

We like to think of memories as though they are tangible. As if somewhere in the vault of our brain there is a complete record of every event we have experienced. As I mentioned in chapter 2, the hippocampus does appear to record a kind of index, enabling us to find the various cell assemblies across the brain that were active during a past experience, so we can reboot our brain to the state it was in during that event. Neuroscientists often focus on this aspect of memory, casting remembering as the turning on of the exact collection of neurons that were active during an event.

Frederic Bartlett, with his conviction that memory is an act of “imaginative reconstruction,” took a radically different view. He explicitly eschewed the idea of memory traces, arguing instead that memories are born in the moment of reconstruction. In other words, it doesn’t make sense to talk about a single memory for an event when innumerable possible memories can be constructed to describe the same experience. I definitely believe that the hippocampus enables us to get into a past mindset and pull up some details from a past event. But I also agree with Bartlett’s contention that, once we get back to the past, we don’t simply replay things as they happened. If that were the case, while recalling a ten-minute phone conversation, you would spend ten minutes reliving everything you experienced during that conversation. That’s not what happens. Instead, we typically compress that experience into a shorter narrative that captures its gist.

So, the hippocampus might get us back to some of the cell assemblies that were active during some moments in that conversation, but we still need to use schemas in the default network to make sense of what we are pulling up. This reconstruction is prone to error, however, because schemas capture what typically happens, not what did happen.

When we remember, we’re like detectives, trying to solve a mystery by piecing together a narrative from a limited set of clues. A detective can build a case based on an understanding of the killer’s motive, which can be helpful, but it can also lead to biases. In a similar vein, when we remember events, motive can exert a powerful explanatory role, helping us make sense of what occurred. It infuses action with meaning, allowing us to gather up threads of information and weave them into a memorable narrative.

But assumptions about people’s motivations can also fuel our imagination, leading us to fill in the blanks about events in ways that distort our narratives of what happened. Another factor that can bias our memories is that our own goals and motivations affect how we reconstruct an event.

I’m often asked, “How is it that two people can experience the same event together and yet recall it so differently?” To quote Ben Kenobi from Star Wars, “Many of the truths we cling to depend greatly on our own point of view.” People’s different goals, emotions, and beliefs lead them to interpret an event from particular perspectives, and those perspectives will also shape how they reconstruct that event later on.

For example, two people might watch the same World Cup match together yet remember it quite differently. In a 2017 study, rival fans of two German soccer teams watched the same Champions League final game, yet their memories were so biased that each side recalled their own team gaining more ball possessions and the opposite team gaining fewer.

Fortunately, soccer fans notwithstanding, we are not doomed to live in siloed realities. We can break through the walls of our perceptions and find common ground when we assume the other person’s perspective while reconstructing the event. Shifting perspective can also help you recall pieces of information that you couldn’t pull up before.

Distortions in memory can also be driven by external factors, and it only takes a little nudge to influence our reconstructions. In the early 1970s, Elizabeth Loftus, then an assistant professor at the University of Washington, became interested in how witnesses recall events in the courtroom, and whether their testimony can be biased by leading questions from an attorney. To answer this question, she had a group of volunteers view short films of seven traffic collisions and then asked them to estimate the speed of the cars. Without access to the actual speedometer readings, they could only make guesses based on what they remembered. Loftus found that it was remarkably easy to bias those guesses. One group of volunteers was asked to estimate how fast the cars were going when they “contacted” each other. Their average estimate was about thirty-one miles per hour. Another group was asked how fast the cars were going when they “smashed into” each other. The average estimate for this group was about forty-one miles per hour. By simply changing the verb she used in her question, Loftus increased the estimated speed by about 33 percent.

Loftus’s study highlights the ways eyewitness testimony can be corrupted in the courtroom, but the ramifications are even greater. If even subtle hints and cues can affect the stories we construct about our past experiences, then our memories are reflections of both the past and the present. Our thoughts and motivations at the moment of recollection can influence our memories of the past. Given the evidence that we reexperience the past by imagining it, and that the process is vulnerable to influence, we are faced with a difficult question: How can we we know the difference between fact and fantasy? How do our brains navigate the borderless landscape of our imagination while simultaneously keeping us anchored in reality?

IS THIS REAL LIFE, OR JUST FANTASY?

As anyone who has tried to meditate for more than a few minutes can tell you, our minds are constantly churning with what-ifs. We conjure up scenarios for what could happen in the future, and we wonder what our present would be like if past events had turned out differently.

My former mentor and collaborator Marcia Johnson, director of the Memory and Cognition Lab at Yale University, was among the first in the field of psychology to realize that this creates a huge problem. All the scenarios we imagine leave us with memories of events we have never experienced, and our memories don’t come with labels certifying them as imagined or real. So, it’s not a question of whether we get confused between memory and imagination, but rather, what factors keep us from making such mistakes all the time?

Over the course of her career, Marcia delved into the processes by which we distinguish between internally generated information (thoughts, feelings, imagination) and the information we take in from the outside world. Her work has shown that we can keep some of the inaccuracies of memory in check by consciously applying some critical thinking to our imaginative reconstructions, a process she called reality monitoring.

Two factors can help us when it comes to reality monitoring. One is related to differences in the features generated by real versus imagined events. On average, imagined events are more focused on our thoughts and feelings and are less detailed and vivid than events that we actually experienced. As a result, we are more likely to believe real events than imagined elements as we are remembering.

Here’s an example of how reality monitoring might play out in real life. If I asked you to think of your last visit to the doctor, you might call to mind your irritation at the stack of medical forms you were asked to fill out before being called into the examination room, your anxiety as you waited for your doctor to come in with the results of your lab tests, or your reluctance to confess that you haven’t been taking a prescribed medicine. Now, you might have had those thoughts and feelings even if you just imagined going to the doctor’s office.

On the other hand, perhaps as the memory unfolds, you hear the annoyingly cheerful Muzak in the waiting room, the antiseptic smell permeating the examination room as you changed into a scratchy paper gown, or the cold of the stethoscope on your chest. The more sensory details that come to mind when you remember an event, the more likely it is that it really happened, because on average what we imagine is not as detailed as what we have experienced.

The other factor that can help with reality monitoring has to do with the way we go about evaluating our mental experiences. A particularly vivid memory constructed with imagined elements (especially elements that fit our schemas or motives) can lead us astray, but you can stay on track if you’re willing to put in the work to gauge the accuracy of your memories.

Take a moment to ask yourself, “Am I pulling up the actual sights and sounds from a past event or just thoughts and emotions I could have experienced when I anticipated an event or thought about it after the fact?” We can counteract the fallibility of memory by considering not only the quality of the details that seem to put us back in a specific place and time but also the likelihood that those details could have been constructed to create an alternative reality.

As with all critical thinking, it helps to remain skeptical until presented with further evidence. Anything that affects the prefrontal cortex is going to affect your ability to remember with intention, and that also applies to reality monitoring. Marcia Johnson’s ideas inspired my dissertation research in graduate school, which identified the prefrontal cortex as a key region in using intention to monitor the accuracy of our memories, and a few years later Marcia and I collaborated on an fMRI study showing that monitoring the details of our memories engages the most evolutionarily advanced areas of the human prefrontal cortex.

Our findings coincided with similar results from a number of other labs. Later, my friend and collaborator Professor Jon Simons at the University of Cambridge discovered that people who have more gray matter in this area are better at reality monitoring. At the opposite end of the spectrum, some people who have extensive damage to the prefrontal cortex can confidently recall things that never happened, a phenomenon called confabulation.

Neurologist and author Jules Montague described the case of a young seamstress from Dublin, Ireland, named Maggie who was convinced she had visited Madonna’s house the week before and advised her on what outfits to wear on tour. Although Maggie had never met the pop singer, she was not psychotic nor was she lying; she had encephalitis, swelling of the brain, which was interfering with her ability to monitor the sources of information that popped into her head.

We all are guilty of minor confabulations. When we’re tired or stressed-out, or when our attention is divided by multitasking, reality monitoring goes out the window. As we get older, prefrontal function gets worse, and we find it harder to tell the difference between imagination and experience. This happens even to memory researchers. Every morning I am greeted with a flood of emails in my in-box. Often, I will see a message that requires me to do some task, and I make a mental note to get to it later when I have time to respond. After going through all the messages in my in-box, however, I will frequently confabulate that I responded to the email, only to be surprised a week later when I get an exasperated follow-up message from the same sender with URGENT in the subject line.

Reality monitoring can also be challenging for people with vivid imaginations. If we visualize something in extraordinary detail, we can drive the sensory areas of the brain in a way that’s similar to what happens when seeing something in the outside world. Accordingly, it can be extraordinarily difficult to tell the difference between experienced events and events that were vividly imagined.

Solomon Shereshevsky, whose remarkable memory was inextricably tied to his extraordinarily vivid imagination, struggled to walk this line. He noted “no great difference between the things I imagine and what exists in reality.” After decades studying Shereshevsky, Alexander Luria concluded, “One would be hard put to say which was more real for him: the world of imagination in which he lived, or the world of reality in which he was but a temporary guest.”

Ultimately, the same magical thinking that allowed Shereshevsky to control his heart rate or body temperature, even to imagine away pain or illness, alienated him from the world around him. Although for a time he had some success performing as a traveling mnemonist, his mental powers never led to the greatness he had imagined for himself. Eventually, he married and had a son and, in his later years, made a living as a taxi driver in Moscow, but his reality never quite measured up to the vast worlds he had created in his mind. He is said to have died in 1958 from complications of alcoholism.

SPARK OF CREATION

Close your eyes and imagine yourself on a white sandy beach by a beautiful tropical bay. Perhaps you visualized lying in a hammock under the shade of a palm tree, sipping a mai tai and watching waves of turquoise water gently lapping against the shore. If you’re like me, you probably got so caught up in this tropical fantasy that, for a moment, you lost track of the outside world.

Now, consider how one individual with amnesia imagined the same scenario: “Really all I can see is the color of the blue sky and the white sand.” Even with repeated prompts, patients in this study struggled to come up with anything that would conjure up a coherent and detailed image of the scene.

Frederic Bartlett not only believed that remembering is an imaginative reconstruction, he also argued that imagination is a product of memory. In a 1928 paper that was far ahead of its time, Bartlett speculated that creative works are constructed by essentially doing the opposite of reality monitoring—that is, by pulling up fragments of memories and then, much like Maggie’s Madonna confabulation or Shereshevsky’s mental walks, assembling those bits and pieces into a cohesive product of imagination.

Bartlett’s paper consisted mostly of descriptive observations and speculation, but there is more compelling data for the idea that memories serve as the raw materials for creative inspiration. Results from fMRI studies and work done in people with amnesia suggest that the mental processes we use to vividly imagine ourselves in a situation, such as sipping a cocktail in a hammock on a tropical beach, are similar to those that we draw upon when recollecting the past. We pull up details from a few different events in our past (via the hippocampus) and rely on information from schemas (via the default network) to assemble a story about how it all could fit together.

The takeaway from this research is not that we generate distorted, inaccurate memories because of our wonky brains. Instead, research by neuroscientists Daniel Schacter and Donna Addis suggests that the hippocampus and the DMN might function at the crossroads between memory and imagination by allowing us to extract the ingredients from past experiences and recombine them into new creations.

For instance, fMRI studies have linked activity in the hippocampus and DMN to performance on laboratory tasks designed to elicit creative thinking, and conversely, dysfunction in these areas seems to impair performance on such tasks. And, consistent with Bartlett’s ideas about imagination and memory reconstruction, people who show higher performance on tests of creative thinking also are more susceptible to so-called false memories.

There seems to be an inextricable link between reconstruction and creation. These findings have given me, as a songwriter and musician, a new perspective on creativity. I sometimes feel stuck in a creative rut, adrift waiting for a new idea to pop into my head. But a new work of art, music, or literature doesn’t arise out of thin air—it emerges from the process of integrating elements from different past experiences. As Steal Like an Artist author Austin Kleon put it, “The idea that the artist should sit down and create something ‘new’ is a paralyzing delusion. We can only create a collage of our influences, our memories—filtered through our imagination.”

The schemas that form the backbone of our episodic memories also serve as raw materials for storytelling. In The Hero with a Thousand Faces, literature professor Joseph Campbell proposed that every culture has a story that follows a universal narrative arc, in which an ordinary person (almost always an orphan or outcast) is called to leave familiar circumstances, navigate an unfamiliar world, do something extraordinary, and eventually save the day.

Virtually every blockbuster film of the late twentieth century, from Star Wars and Spider-Man to The Lion King and The Lord of the Rings, follows this narrative blueprint. These movies layer enough content variation onto a familiar structure to allow audiences to feel the excitement of experiencing something new within the comfort of a predictable story arc.

The link between imagination and memory also shows up in the field of generative artificial intelligence, where companies are creating increasingly sophisticated programs, trained with massive numbers of examples of a particular genre, that can generate new outputs that seem to capture the essence of what they were trained with.

One of my favorite examples is a project called Relentless Doppelganger, a nonstop livestream of AI-generated death metal that is capable of producing songs that could have been recorded by any of a hundred black metal bands from Scandinavia. Another program, Stable Diffusion, uses AI to respond to user prompts such as “monkey eating a banana split” to create new works of visual art. Other AI-based products, such as controversial chatbot ChatGPT, can make entire poems or stories based on simple user inputs. The results can be amazing, but it’s not necessarily surprising. AI art is not about generating something new, it’s about taking elements from preexisting human art and recombining them (based on human direction and curation).

Should artists, authors, and musicians be worried about the rise of the machines? No. If anything, AI art is just a reflection of the cultural milieu that spawns human-generated artwork. As I described in chapter 2, the neural networks powering modern generative AI are inspired by neocortical networks in the human brain that are optimized to learn the general structure of what we have experienced. So, it makes sense that a computer program with no understanding of the artistic process can capture the essence of human-generated art that fits within an established genre.

Unlike a computer program that finds common elements among examples of numerous artworks from a particular genre, more innovative artists draw from an eclectic range of influences, often surrounding themselves with people from different backgrounds and exposing themselves to a broad range of creative works from different genres, eras, and cultures. Innovative artists discover connections between movements and ideas that would traditionally be labeled as distinct (by conventional humans and generative AI).

We see this tapestry of eclectic influence reflected in creative genius across all forms of artistic expression. Pablo Picasso’s cubist paintings were preceded by a period during which he was profoundly influenced by traditional African sculpture and masks. Japanese auteur Akira Kurosawa, who spent his formative years studying Western painting and cinema, used light and shadow to compose shots in his films in a way that is reminiscent of the brushstrokes of expressionist painters. Members of the rap group Wu-Tang Clan grew up on Staten Island in New York City, but their groundbreaking music and lyrics are a mash-up of cultural influences ranging from comic books and science fiction, to Chinese martial arts films, to the philosophy of the Nation of Islam.

By exposing ourselves to a diverse range of people and ideas, we can discover new connections and recombine our experiences into new artistic constructions that transcend the sum of their parts. What makes great art both singular and universal is that it offers a version of reality marked by the idiosyncrasies of its creator, rather than a perfect recording. The same can be said of memory, and for much the same reason; our memories reflect both what we experienced and our interpretations of what happened.

The science of memory is a still-unfolding frontier. Over the last century we have broadened our understanding of how the neural networks that have evolved to remember the past intersect with those allowing us to imagine the future. We now know that it is in that space where memory meets imagination that we interpret reality and create our greatest treasures.

by Charan Ranganath from Chapter 4 of Why We Remember: Unlocking Memory's Power to Hold on to What Matters