If you’re reading this on a screen, fuck off. I’ll only talk if I’m gripped with both hands.
– Joshua Cohen, Book of Numbers (2015)
In 1961, Philip Roth laid bare his assessment of literature’s losing battle to the twenty-four-hour news cycle in the seminal essay “Writing American Fiction.” “American reality stupefies, it sickens, it infuriates,” he writes, “and finally it is even a kind of embarrassment to one’s own meager imagination. The actuality is continually outdoing our talents, and the culture tosses up figures almost daily that are the envy of any novelist.”
Yet, in the face of this slanted challenge, Roth continued to write with focus and intensity. He erred away from the purely imaginative that came to vise postmodernists like John Barth and Robert Coover, and conceded to the same sensational realism that had come to dominate news television. For Roth, this meant writing fiction from an autobiographical paradigm that drew on his growing celebrity, and compelled readers to investigate the enigmatic borderlands of his various textual personae: Roth, the author (his pseudonymous Zuckerman); Roth, the character (there are two Philip’s Roth in Operation Shylock: A Confession [1993]); Roth, the narrator; and Roth, himself. He paired this approach with imagistic prose that adopted the clarity and perspective of a fixed film camera; in other words, a televisual style.
Twenty-six novels later, and in the twilight of his career, Roth adopted a mournful tone, telling a reporter in 2009 that the novel, as a form, wouldn’t be anything more than a cultic obsession within twenty-five years. In his estimation, if television had threatened the novel, the internet had given it a fatal blow: “The book can’t compete with the screen.”
A decade away from the novel’s anointed date of demise, the question remains: How can the novel respond to the internet in a way that’s constructive to the form?
Many editors of legacy publications, including Will Blythe of Esquire and Adrienne LaFrance of The Atlantic, have taken the position that the internet is in direct opposition to the novel. Their argument goes that the internet, at its core, offers distraction as an entertainment unto itself: to softly scroll and survey, but not to see; to scan, but not to read.
Critics of “internet novels” roughly characterize them by their amateurism and a sterile, observational prose that mimics the flattened affect of social media; what Becca Rothfeld describes as the internet’s tendency to “scoop out the mind and mash it into a wet pulp” (whether the categorization of novels as “internet novels” is constructive is an altogether separate conversation). In her recent essay “Extremely Online and Incredibly Tedious,” Rhian Sasseen rightly contends that the amateurism and flat style of the social media scroll has disseminated so widely because it is “easy to imitate, which is why it’s popular; all you have to do is listlessly describe a series of actions and throw in a few references to masturbation, drugs, or, ideally, some combination thereof. The books written in this mode tend to blur together.”
Even beyond social media, the internet seems to flatten prose. This is likely due to the distinct ways our brains interpret text – or, how they have been user-engineered to do so –when reading online. In The Shallows: What the Internet is Doing to our Brains (2010), journalist Nicholas Carr homes in on the stultifying effect of reading digital texts cascaded by advertisements, terms of service, and articles, backgrounded by whatever else might be open on your computer: tabs, texts, apps, email, et al. Additionally, user researchers have found that most users scan web pages in an F-pattern while reading content, anchoring their attention on the first few lines of text, left to right, before descending along the leftmost words of subsequent lines. Both bodies of research conclude that internet reading promotes scanning for quick answers, rather than searching for meaning, encouraging shorter texts and those that favor direct speech, simple sentence structure, and, most importantly, whitespace.
Not only, then, are the “experimentations” of “internet novels” not exactly new; they are tired, even contrived, and, more often than not, feel dated by the time they are published.
The result is today’s prevailing internet aesthetic: dispassionate, even deadening, and offering little formal innovation that might reflect the user reciprocity that sustains the internet and makes its architecture unique. For instance, the declarative, unpunctuated, and associative sentegraphs in Darcie Wilder’s literally show me a healthy person (2017) were celebrated upon the book’s release for mimicking the continuity of her highly popular Twitter account. However, the book’s jumping fluidity more closely mirrors cinematic montage; Wilder herself stated that assembling the book was “more like when I used to edit film.” To wit, her form, which mimics one small corner of the internet, was preempted long before in the methods of appropriation and expressionism pioneered by John Dos Passos in the “U.S.A. Trilogy” (1930-36) and Alfred Döblin in Berlin Alexanderplatz (1929). Both authors imitated the aphoristic polyphony of the radio – the scattershot symphony of scrolling across different frequencies – through the excision and erasure of actual broadcasts.
Not only, then, are the “experimentations” of “internet novels” not exactly new; they are tired, even contrived, and, more often than not, feel dated by the time they are published. Generally, the public sentiment about internet writing reflects those of Blythe and LaFrance: Not only is the internet killing literary reading, it is killing the quality of prose.
But are there ways the novel can respond to the internet, apart from mimicking its speech patterns? For the writer, the challenge at the core of creating fiction after the internet, of capturing this always-updating interconnectedness, extends the contention of David Foster Wallace in his television-concerned essay “E Unibus Plurum” (1993): that is, to enduringly capture the way we live now by representing facts of our current conditions, in this instance our current technology, without being reduced to a period piece, satire, or nostalgia. The literary critic Harold Bloom once wrote that reading a novel is “the most healing of pleasures.” How, then, can a novel in conversation with the internet rehabilitate us from the fundamentally distracting experience of the internet, without being eroded by that very distraction?
*
Years before the internet big-banged into its own universe of thirty trillion hyper-textually linked web pages, philosopher Gilles Deleuze and psychoanalyst Felix Guattari anticipated the defining elements of its structure. In A Thousand Plateaus (1980), they describe two organizational networks: the arborescent and the rhizomatic. The former mimics the structure of a tree, with one central, authoritative pillar that hosts a network of fascicular roots and branches in one direction – outward – in a series of bifurcations. Arborescent networks draw clear lines of causality between a cause and its two effects, and grow along a chronological axis.
The rhizome, by contrast, has no center; instead, it has many sources of growth, which reach out and connect to each other. It is an open, reciprocal system with no beginning or end; where causality runs in both directions, at times in parallel, with no beginning or end.
Per Deleuze and Guattari, a rhizomatic novel would consist of a “system of systems” whose subsystems expanded and contracted, and were all equally fertile for the reader’s entry or exit. By virtue of its nonlinearity, and would always be “between things, intermezzo.”
The essence of the autofictional form does not necessarily say anything about the internet’s superstructure beyond social media, nor does it explain why autofictional works have become of greater interest to readers.
In the latter half of the 1980s, an unexpected group of academics became materially curious about how we might realize the rhizomatic novel: computer scientists. From their research, the hypertext novel was born: an early form of electronic literature – run on hardware – composed of hyperlinks that users click to actively navigate and order the blocks of narrative. What began as an effort to digitize the footnotes of Vladimir Nabokov’s Pale Fire (1962) and the indeterminacy of Julio Cortázar’s Hopscotch (1963) soon inspired a new generation of writers, including Michael Joyce, whose revolutionary novella afternoon, a story (1987) comprised 539 novel fragments connected to one another by links hidden in words, letters, and even blank space.
The hypertext movement, however, dissipated as quickly as it had come. Steven Johnson, a publisher of the pioneering online magazine FEED, cites two principal reasons for the genre’s demise in the late 90s: (1) the technical challenge of writing stories in discrete segments, unordered segments; and (2) the economic challenge of distributing the necessary hardware to run hypertext programs.
Johnson, however, views the internet as the hypertext novel’s first descendant. He says: “a whole different set of new forms arose in [the hypertext novel’s] place: blogs, social networks, crowd-edited encyclopedias. Readers did end up exploring an idea or news event by following links between small blocks of text; it’s just that the blocks of text turned out to be written by different authors, publishing on different sites.”
In most cases, the autofictional novel concedes to a relational paradigm that positions the novel itself in the shadow of the authorial persona.
But what of the internet novel today? It’s tempting to read A Thousand Plateaus as anticipating the structure of the internet. (Wikipedia, with its rhizomatic hypertext, might be seen as a scale model of the internet as a whole.) However, it’s less clear how their formulation maps onto contemporary fiction, as was their intent, and, more broadly, why the culture writ-large has so inexorably linked autofiction and the internet novel.
Following Deleuze and Guattari, we might define the internet novel as a physical, rhizomatic text with two instantiations: the autofictional novel and the database novel. In the former, more common case, the author is interpreted, willingly or not, as its hermeneutic center; whereas the latter, less common case employs some combination of innovative structuring, excision, and appropriation that highlights (1) the factuality of elements that are presented as such; and/or (2) the order of its content. As a result, most database novels tend to be indeterminate. Olga Tokarczuk’s Flights (trans. 2018) is written in various, rhizomatic set pieces – with different protagonists spanning centuries and locales – that relate to each other by loose association, and can be read in any order. Conversely, autofiction’s scrollable, anesthetized prose tends to be linearly ordered, winking and nudging the reader’s attention to the authorial persona outside of the text, as is the case in Gabriel Smith’s Brat (2024), which is narrated as it is being written by one Gabriel Smith.
These classifications leave us with two clean brackets of categorization, but also two questions: Why has so much of recent literary fiction deferred to the autofictional camp? and why should we look beyond it?
*
The philosopher Byung-Chul Hun writes in The Burnout Society (trans. 2015) that neoliberal market forces have created a culture of “voluntary disclosure bordering on the pornographic” in an age when social media has become an unpaid requirement for career growth, or, at least, for its possibility. Autofiction, some argue, is a product of a hustle economy that rewards people who make products of their confessionalism, a cultural narcissism described by the likes of philosopher J. David Velleman and historian Christopher Lasch.
In the lynchpin essay of her recent collection No Judgment (2024), “I Am the One Who is Sitting Here, for Hours and Hours and Hours,” novelist and critic Lauren Oyler reflects on the concurrence of the American autofiction boom, spanning Sheila Heti to Ben Lerner, and the rise of social media at the end of the aughts. Autofiction, she argues, isn’t a genre so much as a situation: What happens within the novel is just as important as what happens around it, with both activities influencing each other. Reading a Ben Lerner autofictional novel will irrevocably be influenced by how its author lives his life as a public literary figure, just as our responses will inform how he formulates his novels – an anxiety of personal reception that mimics the feedback loops of social media. (In her Bookforum review of Sean Thor Conroe’s Fuccboi [2022], Hermione Hoby quotes a friend who avers that Ben Lerner “was better when he was bad,” preferring the “guilty, horny, callow narrator” of his earlier work to the self-conscious “man of marches” in The Topeka School [2019].)
For a New Historicist, this tidy sociological analysis might provide adequate explanation for the proliferation of autofiction and its close association with “internet novels.” However, the essence of the autofictional form does not necessarily say anything about the internet’s superstructure beyond social media, nor does it explain why autofictional works have become of greater interest to readers.
*
One of the earlier American “internet stories” to become infamous among the junior literati was Adrien Brody (2012). In the 15,000-word autofictional piece, author Marie Calloway recounts visiting New York to meet up with a middle-aged editor that she’d first encountered online. She quickly learns that he has a partner and is spiritually, if not physically, impotent; nevertheless, they manage underwhelming sex. First published on Calloway’s Tumblr in 2011 as an essay that included the editor’s legal name, the story was later taken down, anonymized, and re-published as fiction by Tao Lin’s Muumuu House, then collected in her what purpose did i serve in your life (2013). Reading Adrien Brody is, in many ways, like reading an exposé in New York magazine: The narrative’s propulsion is charted by a voyeuristic curiosity to learn more about Adrien Brody’s identity and Calloway’s persona, a pursuit that draws toward the reaches of the story’s rhizome on Calloway’s blog, far outside the text itself.
Calloway, now reclusive, was a precursor to a faction of “internet writers” controversially dubbed “Literary It Girls” by Sophia June. June writes that “what really makes [Literary It Girls] influential is the creative ways they stage and elevate their work – both on the page and in [their] persona.” Delia Cai, author of Central Places (2023), attests by offering that “People understand personalities and personas so much automatically. I think literary It Girls are very good at … promoting and understanding that the persona is kind of part of the work, too.”
The problem with this paradigm is that, in practice, it displaces the object of literary critique from the novel to the persona; a parasitic arrangement that erodes the reader-writer relationship by pre-supposing one. Novelist Jess Row concretizes this point from the reader’s point of view in a recent piece for LitHub: “The paradigmatic example in my mind is Brandon Taylor, who became well-known for his argumentative, self-dramatizing voice on Twitter years before publishing his first – highly autofictional – novel Real Life (2020), which I found stiff and almost impersonal by comparison. While reading it I experienced a strange sense that something was missing: the sensibility of the ‘real,’ online Brandon.”
In most cases, the autofictional novel concedes to a relational paradigm that positions the novel itself in the shadow of the authorial persona, meaning that what the author performs becomes the primary subject of critical evaluation rather than what they produce. This is the hazard of Cai’s proposition that people readily understand personae relative to novels, when we only think personae are easier to understand because we’ve become accustomed to consuming them en masse on social media, however shallowly. When an authorial persona becomes fundamental to a novel, literacy becomes conflated with social media literacy, and critical evaluation becomes reduced to ad hominem critique.
More fundamentally, while the style of the autofictional novel’s disaffected prose might sound like the internet, it doesn’t embody the totality of its underlying structure. Any appearance of bi-directionality is illusory; persona is always its central cause, in which case, we aren’t looking at a rhizome; we’re looking at a tree. The arborescent submits the reader to a hierarchical system that “preexists the individual, who is integrated into it at an allotted place.”
In the arborescent, then, there is no free movement for the reader to transcend the limits of the author’s persona, nor can the writer break free from this tangled root of narcissism. The effect of the tyranny of persona is a broader literary culture that substitutes the difficult work of understanding a text and, as a consequence, the person described in a text, by deferring to first judgments of their persona; first judgments from which none of us in life are otherwise spared, and first judgments that literature ought to have the unique capacity to challenge.
*
We’ve arrived at a fork in the road. After a long tenure as the internet novel’s representative form, it is clear that the autofictional novel, as we’ve typically conceived of it, is not up to the task. But what of the other path?
The sine qua non of the database novel is its focus on the superstructure that hosts the persona, rather than the persona unto itself. Database novels are organized like a database: They comprise a repository of independent units – their sentences – and a seemingly infinite number of coherent wholes. Though their sequencing does have a bearing on rhythm, it has no effect on apparent causality or the hypnotic effect of the novels on the whole.
One recent example is Benjamín Labatut’s When We Cease to Understand the World (trans. 2020), a self-described “work of fiction based on real events” that spans the lives of notorious scientific figures, like Fritz Haber, who invented a process that synthesizes half of the nitrogen fertilizer used globally, while also contributing to the development of mustard gas and Zyklon B; or Alexandre Grothendieck, whose research illuminated a number of mathematical fields with unprecedented aperture, before withdrawing from society. Each independent entry represents a record in the repository of the database novel, creating a tableau that requires neither a prescribed sequence nor an understanding of the scientists therein. What it does require, however, is a skeptical eye trained to spot where Labatut has diverged from fact in his historical re-tellings.
In the book’s coda, Labatut slyly underlines, then undermines his tenuous relationship with facticity. He confesses that while the first entry, “Prussian Blue,” is effectively a work of nonfiction, there is a diffuse strain of invention that percolates in the stories that follow. However, he does not admit where he has selectively placed or misplaced fact and fiction, and the bibliography he provides is noticeably incomplete. Labatut’s half-truths intentionally destabilize his narrational authority.
Whereas the autofictional narrator establishes an autocracy over the world they relate, like an influencer who won’t respond to their followers, the author of the database novel calls their own authority into question.
In her aforementioned essay, Sasseen writes that “there is an immersive quality to the internet, the Wikipedia rabbit holes and the endless link trees, that the affectless writing that has become the house style of online life fails to capture… It’s in the sorting of the information, the understanding of how this information gets filtered into the very structures of our language, that the art lies.” This is precisely what Labatut does when he enlists his readers to dialectically explore the content of his novel alongside extra-textual sources that describe the “real” events fictionalized in his book, a process that informs the book’s internal relationships and meanings. Ruth Franklin’s write-up in the New Yorker was critical of the text; yet she acknowledges that “tantalized, I found myself Googling anecdotes and details, each more preposterous than the last – those cyanide capsules passed out by the Hitler Youth, or a Nazi drive to plant mulberry trees in order to cultivate silkworms – and discovering them to be true.”
Interestingly, like W.G. Sebald or Tokarczuk, Labatut’s books are not without some elements of autofiction. The novel’s last entry, “The Night Gardener,” is the only one written in the first person. Labatut’s account, which includes apparent autobiographical details, purports a series of interactions in a small, isolated mountain town between the narrator and a meek gardener, a former mathematician who briefs Labatut on Haber and Grothendieck.
What this intrusive tendency suggests, beyond a narrative convenience, is an impulse among database novelists to acknowledge their function as arbiters of the narratives they construct. Whereas the autofictional narrator establishes an autocracy over the world they relate, like an influencer who won’t respond to their followers, the author of the database novel calls their own authority into question. This reduces the distance between reader and writer, the latter encouraging the former to pursue the rhizome that mutually links them – reenacting the process of, say, surfing Wikipedia – and fostering a deeper engagement with the world and history, rather than with a celebrity persona.
*
A question remains: Why does the autofictional novel still dominate the database novel? One clear answer is that the former has the aura of celebrity, as well as its associated false intimacy. However, as we’ve learned, the latter has its own rhizomatic logic, one that scoffs at false binaries; of fascicular tyranny. It’s clear that the database novel can function with autofictional elements. In a more sustained case, how might that look?
Sheila Heti’s How Should a Person Be? (2012) – an ur-text of North American autofiction – opens with a discussion about Andy Warhol’s belief that anyone (even writers) could become a celebrity. Having been deemed “hideously narcissistic,” condemned to the business of “solipsistic existentialism,” and falsely assumed to map one-to-one with her narrators, Heti is deeply familiar with the mechanism by which contemporary criticism trains its crosshairs on an authorial persona.
Fourteen years later, Heti’s new novel, Alphabetical Diaries (2024), has it both ways. By diffusing her persona across the database novel’s design, Alphabetical Diaries is by turns more “personal” and less liable to accusations of “solipsistic existentialism” than the books that came before it. True to rhizomatic form, its first incarnation was published in n+1 under the title “From My Diaries (2006–10) in Alphabetical Order” in 2014. Increments were then serialized for a ten-part series for the New York Times in 2022, before being released as a novel this year. Introducing the New York Times run, Heti writes: “A little more than 10 years ago, I began looking back at the diaries I had kept over the previous decade. I wondered if I’d changed. So I loaded all 500,000 words of my journals into Excel to order the sentences alphabetically. Perhaps this would help me identify patterns and repetitions.”
Whether in the novel, gripped with two hands, or in the purlieu of a computer screen, the synthesis of both autofictional and database elements is what holds the reader’s focus on the broader world.
Alphabetical Diaries’ organizational logic presents a rendering of contemporary life that appears asynchronous, atemporal, and acausal because of its departure from the paradigmatic engine that tends to drive narrative movement: time. This is reflective not only of the ways in which information is stored and sequenced on social media, but on the internet more generally. Entries can be dated – whether by a timestamp, or by the references that Heti makes – but her declarations are presented to the reader as though they are all happening simultaneously and in the present moment. The experience of reading these sentences is not unlike looking at an old photograph on Google images, which is no less clear in its resolution now than when it was first uploaded, much less first taken, however many years ago.
Reading Alphabetical Diaries induces a trance-like experience; the implicit anaphora of her Oulipian-constraint, and the Lishian “swerve” (each sentence undoing the last) of the associative narrative, create a hypnotic rhythm paired with poetic, pantoum-like sequences of recontextualization. For instance:
I wonder if I will get pregnant this year. I wonder if that’s what life is going to be – a gradual process of moving away from all the things which seemed to promise a center. I wonder if the beginning of a friendship is the optimistic waiting for something bad to happen, which can be overcome together, at which point the real friendship can begin. I wondered if this was what I patterned my love on, this feeling of scarcity which I have always had. I wore my new scarf and stockings, my boots from Paris, my navy blue dress, my spring coat, and I bought a set of hair bands which I would have liked to have been wearing earlier at the shoot, or at least I could have thought to put a bit of lipstick on. I worry that he is too much like my father, and that things will grow stale as they did with Pavel and I.
This sequence captures how Heti’s construction of a rhizomatic world dissolves the primacy of her “self.” We are uncertain how Heti wants to get pregnant; like the rhizome, she views her life as a movement between nodes which, despite appearances, are not a center. A friendship is ending, or perhaps it has ended, and she has patterned her love on it, or has she patterned her love on the man with whom she wonders if she will have a child? Is there even a man? And is it Pavel, or the other, unnamed man? And which man, or woman, is she going on a date with when she puts on her Parisian boots, if she is going on a date to begin with?
The ambiguity of Heti’s fragmentation diverges from the ethos of the blogosphere, the therapizing style of the personal-essay boom, the scattershot aphorisms in literally show me a healthy person, and even the early canon of American autofiction that Heti belonged to, which all tended toward an absolute judgment. Because of her “Lishian swerve,” just as soon as Heti makes a judgment in Alphabetical Diaries, it is undone, and then undone again, invoking the reader to interpretively assemble her sentences into a conceptual structure of their own from which they can derive meaning. This relationality between sentence fragments is too poetic to be reduced to the “flattened prose” characteristic of contemporary autofiction. The effect of Heti’s prose is, in part, heated up by its apparent earnestness, a trait that is above all balanced between a sense of intimacy and discussion, rather than declaration. Not only is Heti looking for “patterns and repetitions” throughout a novel with no inborn endpoint; she is asking us to do so as well.
What Alphabetical Diaries achieves is maintain the (para)social function that satisfies readers’ desires to connect with the author, without lapsing into a critical voyeurism of the authorial persona; instead, the reader is enlisted to assemble the novel’s fragmented structure. Whether in the novel, gripped with two hands, or in the purlieu of a computer screen, the synthesis of both autofictional and database elements is what holds the reader’s focus on the broader world as related by a single narrator, rather than the persona related through an ever-distant world. In turn, the world is brought closer to the solitary reader, and the solitary reader closer to life.
___
