When it comes to reading, most people can identify a book (often one they read in adolescence) that changed the way they think. Stephen King has cited William Golding’s Lord of the Flies, which he encountered when he was twelve years old. He described it as, “The first book with hands—strong ones that reached out of the pages and seized me by the throat.” We don’t even have to take his word for it. Elements of Lord of the Flies permeate King’s books, which reveal the innate potential in kids for both good and evil.
When we say that something changes the brain, there are two ways that a change might manifest. The first is a transient alteration. The majority of psychological experiments are designed around this phenomenon. These transient changes are relatively easy to detect. You define a control condition and then present the subject with a stimulus that is designed to evoke a particular response. The experimenter assumes that once the stimulus is gone, the response will return to its baseline.
The response could be anything measurable. It could be a keypress on a keyboard, or it could be a physiological response like a change in heart rate, skin conductance, or brain response as measured with fMRI. These types of experiments are efficient. Trials can be repeated over and over until the experimenter acquires enough data for analysis.
The second type of alteration is a long-lasting change, but these are more difficult to measure. When it comes to the brain, transient alterations represent ephemeral change, and most neuroscientists interpret them in terms of momentary information processing rather than lasting change. The visual cortex, for example, responds to changes in the visual field, but these are not thought to persist. Once a stimulus is gone, so is the brain response. When it comes to a cultural trace in the form of literature, we would really like to know whether there is some sort of permanent alteration to the structure of the brain. This requires a different type of measurement.
A general rule of biological systems is adaptation. For example, the visual system adapts to changes in overall brightness. You are hardly aware of the differences in luminosity between an outdoor scene under the noonday sun and one inside illuminated by a soft-white light bulb. Adaptation makes it far more difficult to detect changes in the brain. It was largely for that reason that not much work had been done on how a book changes the brain.
By 2011, new methods in neuroimaging had emerged that purported to measure stable patterns of activity in the brain. Prior to this, we could only measure transient changes with fMRI, on the timescale of seconds. The alternative, structural imaging, was more like a snapshot of brain anatomy, but even that was not sufficiently detailed to capture the changes that were thought to occur as a result of immersive experience. And, as I had discovered, even a murderer’s brain didn’t show obvious abnormalities.
The new technique was a variation on fMRI called resting-state fMRI, or rs-fMRI. The idea was to have a person lying in the scanner, awake but doing nothing. If you continuously scan a resting person’s brain with fMRI over a period of about ten minutes, patterns begin to emerge. The brain shows coordinated activity in disparate regions, where the measured signals fluctuate up and down in synchrony. These are called resting-state networks or, sometimes, default mode networks, because they represent the default mode of activity when a person isn’t doing anything.When it comes to a cultural trace in the form of literature, we would really like to know whether there is some sort of permanent alteration to the structure of the brain.
Neuroscientists continue to debate the functional significance of the resting-state networks. One possibility is that this activity is just the background hum of the brain, like the buzzing of a beehive. It doesn’t have any functional significance other than neurons, like worker bees, quietly going about their business keeping the brain and body alive. Another, intriguing possibility suggests that the resting-state networks represent the anatomy of daydreaming. Proponents of the beehive theory point out that resting-state networks are present even under light sedation, when spontaneous cognition is blunted. However, anyone who has had light sedation—dental work, colonoscopy—knows that light sedation is not the same as full anesthesia.
Resting state is a bit of a misnomer because the networks can be perturbed by other tasks. In one experiment, students were scanned before and ninety days after studying for the standardized Law School Admission Test (LSAT). Connections within the frontoparietal resting network were found to be stronger after studying, and the researchers concluded that intensive training for logic questions strengthened these patterns. Maybe the students were thinking about the LSAT during the scan. More likely, the act of studying, especially because it had occurred repeatedly over days and weeks, resulted in a physical alteration to the brain itself, and these changes persisted into quiet rest periods.
If studying for the LSAT can result in detectable changes in the resting state of the brain, what about reading a book? I wondered whether we could capture a transformative reading experience like the one Stephen King attributed to the Lord of the Flies. This was the question at the heart of an experiment I undertook in 2011.
The first decision we had to make was: Which book? Every day for weeks, my team members gathered around the central table in the lab. Undergrads, graduate students, research specialists, and other faculty threw out ideas for their favorite books, ones that changed their lives. One person loved poetry, but no one else did, nor would our likely pool of volunteers, whom we expected to be young undergrads.
Harry Potter, of course, was all the rage, and Harry Potter and the Deathly Hallows was the most rabidly anticipated movie that summer. But we had to assume most of our volunteers would already have Harry Potter on the brain. So that was out. We debated about the so-called classics, but here, too, we had to assume that any Emory undergrad would have been exposed to at least a few. That ruled out my favorites: The Odyssey, Crime and Punishment, Moby-Dick. And forget about Dune or Foundation. Those were hopelessly out of date.
Although everyone had a favorite novel, we could not agree on any single work of fiction that we thought would impact the brain of a college sophomore. Some argued for nonfiction because it portrayed real events, but the idea was dismissed when nobody could think of a nonfiction book that changed their worldview as a teen. In the end, we settled on historical fiction, a plot-driven genre grounded in true events but told with a narrative flair that made history dynamic.
We chose the 2003 book Pompeii, by Robert Harris. I remembered how much I enjoyed it when it came out. But because it had been published eight years earlier, nobody in the lab had heard of it. Everyone was familiar with the basic story: Mount Vesuvius erupted and rained hot ash down on the Roman city of Pompeii, burying its inhabitants. Harris brings the story to life by following the story of the fictional engineer, Marcus Attilius. There is love, sex, death, and tragedy. Would this page-turner make a lasting impression on the brain of a young adult? We hoped to find out.
For our experiment we chose to study the impact of Pompeii on a group of young adults. These would be teens between their freshman and sophomore years in college, around eighteen or nineteen years old. At this age most people are struggling to define their personal identity. Issues of race, gender, class, purpose, and, of course, relationships weigh upon the young adult’s mind. I didn’t expect that Pompeii would be life-altering for the participants, but I hoped that it would at least be powerful enough to cause lasting changes in their brains. Hopefully, they would identify with the hero of the story and his struggle to save the woman he loved from being buried under hot ash.
To make sure the participants actually read the material, we adopted a two-pronged strategy. First, they would have to read a physical copy of the book. Although e-readers were just becoming popular, we didn’t want people skipping ahead, so we purchased a paperback copy for each subject. We then tore each book into nine sections. Participants would receive one packet each day over the course of the experiment. Second, to make sure that our volunteers were actually reading the material, they would take a short quiz every day before receiving the next packet.
As for the brain scanning, the plan was to have each participant come to the MRI center every morning for nineteen consecutive days (including weekends), where they would receive a resting-state fMRI scan. The scan would last about seven and a half minutes. They wouldn’t have to do anything other than rest quietly with eyes closed. After that, they would take the quiz. To assess their engagement with the material, we would also ask them to rate how excited they became while reading. No material would be assigned for the first five days of scanning, which would serve as a baseline, including any daily reading they normally did. The next nine days would be reading days, and this would be followed by another five days with no reading so that we could see if any changes persisted beyond the active reading period.
This was the most complicated experiment I had ever designed. The logistics of convincing twenty volunteers to come in every day at the same time, without fail, for nearly twenty days, and to read an entire book in addition, were daunting. As an incentive, we offered a payment of $400. But there was a catch: we would deduct $100 for every missed session. We also prescreened volunteers for their commitment to reading with the simple question, “What was the last book you read for fun?” As students, everyone would have been reading class material throughout the academic year, but we wanted only people who also found time to read for fun. To be accepted into the study, a student had to have read at least one book for fun during the preceding school year.
In the end, nineteen participants (eleven female, eight male) made it through the whole experiment. Before getting to the imaging results, we had to address the question of whether the book had had any impact at all. Did the students identify with Attilius or Corelia—the woman he loved? Did they find the story interesting or was it just historical information without any emotional resonance? One clue came from the question about excitement we posed each day. The question varied slightly based on the day’s material, but went like this, “On a scale from 1 to 4, overall how excited were you by this reading?”
Participants underwent resting-state fMRI scans on nineteen consecutive days. On evenings before the middle nine days of scanning, participants also read a portion of the novel, Pompeii. The excitement rating of each excerpt showed a rising trend toward the climax of the novel.
The average excitement rating began a bit above the midpoint of the scale, but started to take off on day 5, after the big Roman feast and orgy. The excitement continued to build until the eruption of Mount Vesuvius on day 7. There was a brief respite, before peaking on the final day, when Attilius and his lover are buried in the ashes, along with everyone else trapped in Pompeii. The curve looks remarkably like the canonical rags-to-riches template. But because everyone dies, Pompeii is really a story of riches-to-rags. Although these curves are related, our question did not ask whether their excitement was good or bad. At a minimum, the excitement curve told us that the material had a measurable impact. If it had been flat, then I would have worried that the students hadn’t paid attention or didn’t relate to the material.
Confident that the book had caused at least a transient change in subjective feelings, we analyzed each person’s resting-state activity across the nineteen days to determine where in the brain these changes occurred. Given how excited the students became, I expected changes in brain regions associated with emotions. But that is not what we found. Instead, we found a network of regions organized in a hub-and-spoke pattern, with the hub centered on an area of the left temporal lobe called the angular gyrus. This region is known to be critically involved in language comprehension. This change in connectivity represented a carryover effect of the actual reading, similar to how a muscle feels the day after exercising.
This interpretation becomes clearer in the context of the predictive functions of the brain. As I mentioned earlier, the brain is never really at rest, and the default networks can be considered a suite of modes that dynamically shift between one another as the mind turns inward, processing recent events. In the case of our experiment, the recent events occurred within the novel, and the change in brain connectivity reflected the incorporation of those events into each person’s brain and personal narrative. The changes in the temporal lobe were present only during the reading days. After the novel was finished, the connectivity pattern returned to its previous configuration.
To see whether the novel caused longer-lasting changes, we also looked for a different pattern: one that occurred with the on-set of reading and then persisted after the novel was over. Only one network showed this pattern: the sensorimotor strip. This was a surprise, because it wasn’t the emotional regions I had expected. The sensorimotor strip occupies the bank of folds along the central sulcus, where tactile impulses enter the cortex and motor impulses leave. So, why should a novel change the pattern of activity there?
One possibility is that reading the novel invoked neural activity associated with bodily sensations and that these activity traces carried over to the resting-state scan. Pompeii is, in fact, a visceral book. The descriptions of Roman feasts and orgies and the showering of molten ash on the inhabitants of the city can make one’s skin crawl. This explanation fits with the theory of embodied semantics, which says that the brain regions responsible for producing an action are also used to represent the action in your mind.
In other words, when you read about someone hitting a home run, your brain unfolds a compressed representation of hitting a homer and then uses your sensorimotor cortex to simulate it. Similar results have been observed for the sensory side. In another imaging study, participants read tactile metaphors, including expressions like “hot-headed,” “unbending attitude,” “weight matters,” and “coarse language.” The simple act of reading these phrases was associated with activity in the sensory strip, suggesting that tactile concepts repurpose the same regions we use for physically feeling things.
Literature immerses the reader in a world created by the writer. In many novels, the reader feels like they are in the body of the protagonist. This would explain the changes we observed in the sensorimotor network during the reading days. It is telling, though, that these changes persisted after the novel was finished. The experiment ended five days after the completion of the reading, so unfortunately we don’t know just how long these changes might have lasted.
Even so, the sensorimotor changes resonate with the subjective feeling of how a book changes a person. Thinking back to the books that have changed my life—Foundation, Dune, and later, The Teachings of Don Juan and The Monkey Wrench Gang—I cannot really recall much detail about the plots. But I can vividly remember the characters—Hari Seldon and the Mule, Paul Atreides, Don Juan, Doc Sarvis. In hindsight, I must have been drawn to their iconoclasm as they thumbed their noses at conventional social norms and forged their own journeys. They have all been incorporated into my identity or, at least, who I think I am.
Adapted from The Self Delusion: The New Neuroscience of How We Invent—and Reinvent—Our Identities, by Gregory Berns. Copyright © 2022. Available from Basic Books, an imprint of Hachette Book Group, Inc.