Armed Metropolitan Police are responding to a stabbing at Central Middlesex Hospital in London.

For years, a primary metaphor for the internet has been the “town square,” an endless space for free expression where everyone can have their say. But as scaled digital platforms have grown to dominate most of modern life, metaphors centered solely on speech have failed to explain our current civic dysfunction.

Perhaps the better way to understand the internet is to compare it to a much older infrastructure problem: citywide sanitation systems. Posted content is akin to water; websites and other interfaces are analogous to pumps; and unintended feedback loops correspond to risk of infection. A public-health framework for understanding the internet would focus not on online information itself but on how it is generated, spread, and consumed via digital platforms.

This model’s genesis lies in the two-century-old story of early advocates for clean water in Victorian England. At the time, the life-threatening diseases that ravaged cities—cholera, typhus, tuberculosis, and scarlet fever—were not new. What were new were modern living conditions. Infections that might have taken weeks to spread through a village suddenly ravaged whole populations within days, and no one understood what was causing the massive outbreaks.

[Read: ‘The cloud’ and other dangerous metaphors]

The Victorian working classes knew whom to blame when disease broke out: doctors. Mobs assaulted members of the medical establishment, leaving government officials unsure how to weigh the safety of physicians against the public interest. Why the rage? The traditional response to disease—quarantines—had become ineffective in industrialized cities, prompting the public to distrust those who profited from treatment.

The first serious approach to the problem was taken by a coalition of doctors, liberal advocates, and social reformers starting in the 1830s. Known as miasmists, they pushed the idea that noxious air was the culprit in epidemics. If a neighborhood could not pass the smell test, the argument went, one immediately knew it was already too late to be saved.

Miasmists, including prominent ones such as Florence Nightingale, have an ambivalent legacy. They were among the first to emphasize that disease had not just biological but also social and economic causes, a crucial insight. But simultaneously, they were dead wrong about the role of air in the spread of the common diseases of the time, a reflection of an elitist worldview and overprescribed morality.

This tension revealed itself during two key events. One was the Great Stink of 1858, in which a combination of hot weather and poor waste disposal transformed the Thames into a cesspool. The stench was so bad that even the curtains of the houses of Parliament had to be caked with lime. No one was safe from the foul air, and by the miasmists’ assumptions, that meant that no one was safe from disease. But, in fact, no major outbreak followed the Great Stink.

Second was the groundbreaking work of a brilliant doctor, John Snow, who had suspected for years that water (not air) was the actual cause of urban epidemics. In a painstaking natural experiment, Snow demonstrated that the Broad Street pump was the source of the 1854 cholera epidemic in the Soho area of London. His data revealed that residents living across the city became sick if they happened to get water from the pump, even while a nearby brewery that drew its water from a different source had no recorded cases. There was no other reasonable explanation: Some as-yet-undiscovered mechanism, localized at the pump, was responsible for infection. Though Snow was careful to frame his results so as not to explicitly reject the miasma theory, the implications were obvious.

After much debate, over the next 20 years London implemented the world’s first modern sewer system. And from 1850 to 1900, urban illness was reframed from a problem of individual circumstance and negligence to one of economic dependency and social interconnectedness. Once it became clear that not only medical professionals but also effective water piping and safety valves were needed, public policy shifted from one-off treatments to longitudinal assessment of population health, fueled by new mechanisms for evaluation. The stakes of public health had shifted: If cholera epidemics continued, they did so only because cities refused to provide potable water to vulnerable populations.

Today we are living in an online version of the Great Stink, and have dire need of John Snow’s methods. Evidence is building rapidly that social media causes great harm at scale, especially in terms of declining mental health and societal trust. But because these effects are not directly measurable (except for what’s been revealed from whistleblowers and difficult natural experiments), like Snow, we are left to speculate about causes while trying to source better data.

[Read: The internet is rotting]

What would it take to build something similar to sanitation infrastructure for social media or generative AI? As we argue in detail in a recently released project, it would mean building assessment tools to connect design features—such as the feedback loops embedded in content-recommendation systems—with population outcomes such as mental-health effects.

To extend the metaphor, current technology interventions tend to focus on moderation strategies centered on specific users and individual pieces of content. This is akin to the role of nurses in public health, crucial and under-resourced providers of well-being. But just as nobody should think that good nursing is the best way to address unclean water, content moderation is insufficient to address dysfunctional platform architectures.

Modern platforms already operate as experimental laboratories, running randomized controlled trials over and over to improve outcomes based on companies’ goals. What we need are tools to assess the pump—the models and interfaces of platforms that determine how populations are exposed to content over time—to gauge whether restrictions need to be implemented to protect at-risk groups. For potential problems such as mental-health impacts or systemic reductions of trust, platform effects would be assessed alongside internal metrics for growth and revenue. And just as epidemiologists learned to focus on infants and children as particularly vulnerable subpopulations, today’s researchers must give special consideration to crucial risk vectors, such as chronic use of social media by adolescents.

Sanitation didn’t just make epidemics easier to control and mitigate; it made the diseases themselves easier to understand, leading eventually to germ theory. Beginning with the early experiments of Louis Pasteur, the new science of bacteriology confirmed the existence of microorganisms, as John Snow only suspected. Once the specific bacterium responsible for cholera was identified under a microscope, a new cornerstone of public health was established.

We are at a similar moment now: We have strong ideas about the causal mechanisms that may be mediating harms from products (such as interpersonal comparisons among teenagers leading to mental-health issues). But just as 1850s London did not need germ theory to start evaluating the effects of water and establish sanitation systems, the first step for mitigating harms in large-scale models is to establish baseline effects independent of explanation. The lesson of public health is that such baselines will be necessary in order to build consensus on what platforms and large language models need to measure and optimize for.

We can continue to treat technology platforms as a town square where the loudest, ugliest voice wins the day. But instead of metaphors that blame individuals, or encourage us to just sign out when things get noxious, we can embrace the standard of public health. The solution won’t come from more content moderators or ever-smarter chatbots but from new infrastructural commitments: pipes, valves, and pumps that would actually keep users safe.

Delegates representing businesses and dozens of governments have come together in London to promise billions of euros to help Ukraine rebuild

In 1883, Evert Nymanover, a Swedish scholar at the University of Minnesota, proposed a new invention that some thought would affect the future of humankind: a device that played recordings of books. Nymanover called the device a “whispering machine” and suggested that it could be placed inside of a hat so that someone walking down the street or reclining in bed “could be perpetually listening” to great works of literature.

Though mocked by some, Nymanover’s vision of a book recording in a hat wasn’t entirely far-fetched in 1883. After announcing the invention of the phonograph six years earlier, Thomas Edison turned almost immediately to the device’s implications for literature. He hoped to open a publishing house in New York that would sell novels recorded on six-inch circular plates. “The advantages of such books over those printed,” Edison wrote, “are too readily seen to need mention.” And Edison wasn’t the only one who thought listening to books would be obviously superior to reading. An 1885 essay in the influential British literary magazine The Nineteenth Century maintained that Nymanover’s whispering machine would be a “boon to our poor abused eyes,” and also that when we read print, “one half the power of literature is lost.”

It took a full century, but the technology finally did catch up to Nymanover’s vision of a world in which people could walk down the street listening to books. And yet, by the time portable cassette players became ubiquitous in the 1980s, the mood about listening to books had changed in a way that would have surprised 19th-century audio enthusiasts. Listening to novels no longer seemed like a utopian fantasy at all. To most, it seemed entirely unappealing. In a 1993 Wall Street Journal article on stagnating audiobook sales, one Random House executive lamented that “too many people still think audio books are only for the blind.”

Prominent literary figures tended to be particularly skeptical of listening to books. Strangely, the problem with the audio format was not that it made books less enjoyable. It was the opposite: Audio made books so relaxing and pleasurable that a listener couldn’t engage critically with the text in a way a serious reader should. Listening to literature, the essayist and critic Sven Birkerts argued in his 1994 book, The Gutenberg Elegies, was like “being seduced, or maybe drugged,” a very different experience from “deep reading,” which Birkerts characterized as “the slow and meditative possession of a book.”

According to Matthew Rubery, the author of The Untold Story of the Talking Book, a fascinating history of the audiobook, the notion that listening to a book is too absorbing to lend itself to deep reflection is the “most enduring critique” of the format. “It was striking to me when I began researching audiobooks how many people in Edison’s time welcomed efforts to make books more entertaining,” Rubery, a literature professor at Queen Mary University of London, told me. “The idea of books needing to be hard work, difficult, and read firsthand in order to be deemed valuable only took hold in the next century.”

That audiobooks have tended to produce anxiety in literary critics is perhaps not surprising. As film and television became the dominant modes of storytelling in the 20th century, book lovers were forced into a defensive crouch, left to argue that the very aspects of reading that made it more rigorous than watching a movie or a show were, in fact, precisely what made reading superior. Audiobooks were suspect because they turned reading into an easier, more passive experience. As the Irish novelist and critic Colm Tóibín once put it, the difference between reading a book and listening to a book was “like the difference between running a marathon and watching a marathon on TV.”

The stigma associated with audiobooks hasn’t gone away since The Wall Street Journal published its 1993 article on audiobooks’ failure to catch on. Daniel T. Willingham, a psychology professor at the University of Virginia who studies reading, says that the most common question he gets is whether listening to an audiobook for a book club is “cheating.” But if anxiety surrounding audiobooks lingers, it’s no longer stopping Americans from purchasing them. Audiobook sales have seen double-digit increases each year since 2012. Last year, the increase was 10 percent, amounting to $1.8 billion in sales. The trend is only likely to accelerate in the years ahead given that Spotify recently made a major push into the market, and Google and Apple are racing to produce AI-narrated books. (Even the dead can now narrate audiobooks.)

Still, if the audiobook moment has arrived, that doesn’t, of course, mean that all of the concerns about the format have been misplaced. I suspect that listening to a novel truly is less likely to elicit critical engagement. What I’m less sure about is whether that’s such a bad thing.

Like many fans of the format, I turned to audiobooks out of convenience. I was teaching a graduate course on contemporary American writers at Johns Hopkins, and it occurred to me that speeding through audio editions of the novels and memoirs I’d assigned could be a good way to refresh my memories of the books in the days before a class. But, along the way, something happened that surprised me: I started to fall in love with the audio novel. It took me a little while to admit it to myself—I had internalized the stigma so deeply that even entertaining the possibility felt heretical—but, in many cases, I was enjoying the books even more when listening to them.

The next surprise arrived when I began listening to audiobooks in bed. In recent years, I’d been reading much less at night. Exhausted from long days of parenting and emailing and Zooming, I would often end up watching a TV show I was not at all excited to watch rather than reading a book I was genuinely excited to read. Then, one night, I put in my earbuds and downloaded Maggie Gyllenhaal’s wonderful narration of Anna Karenina. Listening to a skilled actor read a literary masterpiece was every bit as blissful as the 19th-century utopians had imagined. “Netflix and chill” became “Tolstoy and chill,” and then “Jane Austen and chill,” “James Baldwin and chill,” “Kafka and chill!”

[Read: An ode to being read to]

Was I being seduced? Was I missing out on the wisdom these great authors had to offer by listening instead of reading? Maybe. There’s not a lot of science on the differences between reading and listening to books. The existing research suggests that adults score the same on reading-comprehension tests whether they read or listen to a passage. But it’s one thing to comprehend a book and another to think deeply about what you’ve comprehended. And Willingham, of the University of Virginia, told me there’s good reason to suspect that reading books does, indeed, lend itself to more intense critical engagement than listening to books does.

In one small study, college students were randomly assigned to either read a 3,330-word article or listen to a 22-minute podcast on a scientific topic. Two days later, when the researchers quizzed the students on the topic, those who had read the article did much better than the podcast listeners.

When you’re reading, Willingham explained, you’re in full control of the pace. You can stop and think before moving ahead. “Audiobooks,” he said, “make that harder to do.” Maryanne Wolf, a literacy scholar at UCLA’s school of education and information studies, likewise told me that although she sees advantages and disadvantages to various different book formats, reading—specifically reading on a printed page—is best for understanding something “at a deeper level.”

Audiobook skeptics are probably right. Listening to a novel will never be a substitute for reading, if the aim is to digest and analyze what we’re reading. Harold Bloom, the late critic and literary scholar, told The New York Times in 2005 that, for “deep reading,” you need the text in front of you in order to engage “the whole cognitive process.” And can we really argue with this? The harder question is whether we truly want to engage “the whole cognitive process” when we read novels or whether we want to be fully immersed in what we’re reading without the interruptions of our own thoughts, no matter how insightful. The harder question, put another way, is whether art should ultimately make us think deeply or feel deeply.

Fiction, which lies at the intersection of style and content, makes this question particularly tricky. There’s the music of the language, and also the concepts and ideas communicated through the music. There’s the story itself, and also all of the signs and symbols beneath it. As the critic James Wood says in How Fiction Works, when it comes to literature, “everything is at once a moral question and a formal one.”

The style and substance of a novel, of course, can never be fully disentangled. Someone who reads with more attention to a novel’s content doesn’t entirely miss out on its music, and someone who is drawn to a novel’s style is still fully capable of thinking about the scope of the book’s ideas. Reading, by allowing us to stop and ponder, might tilt the needle a little more toward content, but listening, by harnessing the emotional power of the human voice, might tilt the needle a little more toward style.

The content of a novel is typically what dominates the discussion, particularly in the classroom, but that might be only because it’s so much easier to talk about. We ask young readers to focus on a book’s themes, to write essays on what this or that image symbolizes, as though a literary work were merely a code containing hidden information. A novel, in the process, is often stripped for parts as opposed to appreciated as a form of entertainment.

Classroom lessons that focus more on style do little to solve this problem. Attention to how a writer makes use of foreshadowing or constructs a particularly brilliant metaphor can’t capture what the novelist and essayist Mary Gaitskill describes as a book’s “inner weave.” She notes that this aspect of the novel is “almost impossible to talk about,” and yet it determines “what the work is about as much as the plot or the theme or even the characters.” Gaitskill compares the inner weave of a novel to “a person’s unconscious.” I think of it as the rhythms of another mind, an animating intelligence that I want to spend time with less because of what it is thinking than the way it is thinking.

[Read: A podcast about the airport best sellers we can’t escape]

This anxiety about overanalysis is hardly new. Nearly 60 years ago, Susan Sontag described “interpretation” as “the revenge of the intellect upon art.” What’s new is the growing popularity of the audiobook and its potential to change the way we approach the novel. Some great filmmakers, Sontag pointed out, had avoided heavy-handed theoretical interpretations of their creations “by making works of art whose surface is so unified and clean, whose momentum is so rapid, whose address is so direct that the work can be … just what it is.” Though there is certainly plenty of theory about film, I suspect that a film may be somewhat less susceptible to interpretation than a printed novel not because film is a visual medium but because a film dictates its own pace. When you’re watching a movie, you have little time to stop and think. And though one could repeatedly hit “Pause” when watching at home, few would find it an enjoyable way to experience art.

The true promise of the audiobook, I’ve come to think, may be that it brings the momentum of television and film to literature. By propelling us forward and keeping the intellect a little bit at bay, the audiobook allows the novel, too, to be “just what it is.” Listening is a more passive experience than reading, yes, but, for many, it’s also a more relaxing and pleasurable experience. And the pleasure can’t be overlooked. As the literary critic Laura Miller put it to me, “Why would you even care about allusions or techniques if you don’t actually enjoy novels to begin with?”

Utopian visions don’t often come to fruition. But the 19th-century fantasy whispering machines that could narrate books arrived almost exactly as the futurists predicted—minus, fortunately, having to be placed under our hats. At a moment when fewer and fewer students are choosing to major in English, an unapologetic embrace of audiobooks may be exactly what the literary world needs. After all, the public, as sales figures show, is making its fondness for them clear. Those who love the novel and want our children to love it as well would be wise to listen.

The Guardians of the Galaxy star visits his sister Charlie in London for Thank a Teacher Day.

The World Bank says many years of financial support are needed as London hosts a major conference.