i reviewed sheila heti's alphabetical diaries!

plus tips on pitching and researching a book review ✦ and how to write artistically and intellectually ambitious work

Sheila Heti’s eleventh book, Alphabetical Diaries, is out today—along with my review of it: “Sheila Heti, In Conversation with Herself”, online at ArtReview and in the February print magazine.

I’m tremendously excited that this review is finally being published and I can share it with you all! Please do give it a read and share your thoughts—you can reply to this email, comment on Substack, or reply to my tweet about it.

The American cover of Sheila Heti’s Alphabetical Diaries, published by FSG (it’s published by Fitzcarraldo in the UK)

I’d also like to take some time to reflect on how the review came about. I’m extremely new to literary criticism and freelance writing—this is the second “real” book review I’ve ever done!—so I’m always looking for advice on how to navigate this world. How do people get opportunities? How does a byline happen?

I am regularly amazed by the generosity of other writers who describe how pitching, researching, drafting, and editing work. So this is my attempt to give back—and also reflect a bit on what it means to write literary criticism, and how to do so well.

Pitching the review

I pitched this review ages ago—on October 18 last year. I’d seen a tweet where Louise Benson, ArtReview’s director of digital, asked for pitches of “smart, original, funny, incisive criticism on the visual arts, film, books, TV, the internet, and culture in general”.

I first sent my pitch to Benson and received an auto-reply that she was out on leave and Orit Gat would be filling in her role. I really admire Orit Gat’s writing—especially her 2016 essay “Scroll, Skim, Stare”, on artists’ websites versus artists’ books—and I was frankly a bit intimidated about pitching her.

I didn’t need to feel that way. Orit was incredibly kind and encouraging, in her initial emails and throughout the editing process. I initially pitched a review of a book that had been published 2 months earlier, which she politely turned down—and asked if I had other ideas for forthcoming books.

I replied with this pitch:

I'd love to review Sheila Heti's Alphabetical Diaries (FSG/Fitzcarraldo, Feb 2024). For several years, Heti has put her journal entries into Excel, arranged the sentences in alphabetical order, and edited down the resulting text. Parts of this project were published in the New York Times in 2022, and reveal small slices of Heti's life and concerns. The lack of overt narrative from sentence to sentence ends up creating, instead, a really interesting portrait of Heti's repetitive, patterned, reoccurring anxieties over time.

I could possibly also place this in relation to other writers' diaries—Kafka's, Sontag's—and Edouard Levé's Autoportrait, which similarly creates a portrait of Levé's concerns through disconnected sentences and I statements.

Another possible expansion—thinking about how, in an age of large language models, writers can interact with their own, self-made "small" language corpuses, and use data/sorting techniques (such as Heti's alphabetization) to find new ways of prose writing.

I’m including the pitch in full because it’s one of the most important—and mysterious—parts of freelance writing. I was (and still am, frankly) desperate to know what makes a pitch good/bad, too long/short, interesting/boring. Seeing example pitches is so helpful! I don’t know if this is a good, bad, or average pitch—but it worked.

Orit wrote back and said she’d be happy to commission me. She also mentioned that she was much more interested in the ideas I mentioned in the last paragraph—how writers use data/sorting techniques—because it felt more original and compelling.

There’s a lesson here about what makes an editor interested in a pitch: a new angle on an existing topic (writers’ diaries), ideas that feel more urgent and relevant (data, sorting, and Excel in literature), and topics that the person pitching has particular expertise or interest in.

Researching the review

As I mentioned earlier, I’m quite new to writing literary criticism. But I really love reading it, and genuinely think of it as an art form. The best critical writing is both artistically and intellectually ambitious:

• An artistically ambitious review, to me, has style and voice. It’s not utilitarian writing, but writing that, in Ryan Ruby’s words, involves a “heightening of the prose style to levels expected of literary artwork”.

• An intellectually ambitious review is not just about the book in question, and doesn’t content itself with simply answering: Is this book good or not? Did I like it or not? Instead, it discusses literature and literary culture more broadly—using a particular book to do so.

I wanted to accomplish that in my review of Alphabetical Diaries—and if I failed, at least I would fail in an interesting way.

I also wanted to use the review as a container to explore particular intellectual interests, deepen my understanding of them, and refine my particular takes on them. Some of the interests that motivated this review:

• An interest in process: literary process, design process, computational processes

• Interesting uses of technology in art and writing—not in a very banal way (think Look at this generic science fiction story I created using ChatGPT!) but in a deeply considered way. Specifically, I care about uses of:

  • Naïve technology: Simple sorting and randomization techniques, “low-tech” approaches. This is something that

    Kristoffer
    writes beautifully about in his Substack,
    Naive Weekly
    , when it comes to websites and web-based artistic or literary works. I also focused on naïve technologies and the aesthetic of low-tech web design during my MA research at the Royal College of Art/V&A Museum’s History of Design program.

  • AI and machine learning: off-the-shelf or artist-trained machine learning models and artificial intelligence systems; generating text with an LLM (like ChatGPT) or images with Midjourney and DALL-E; whether AI can produce art; whether artists should trust or distrust AI.

• Autofiction: arguments against autofiction, defenses of autofiction, and how autofiction relates to contemporary life and how we construct ourselves through narratives (whether it’s a novelist’s first-person book or an influencer’s Instagram presence)

These ambitions and interests led to a fairly involved research process. Early on, I received the advice to not just review this book—but rather situate this book in the context of Heti’s trajectory as a writer, her overall artistic development and approach, and how it’s developed over time. So to prepare for the review, I went back and read a lot of Heti:

• I’d already read 3 of Heti’s other novels: How Should a Person Be? (2010), Motherhood (2018), Pure Colour (2022). For this review, I went back and reviewed my Kindle highlights for each (I do 90% of my reading via Kindle library books!). My final review includes a paragraph where I (try to) efficiently summarize these 3 books and the processes she uses to construct each one.

• I found a copy of a true Heti deep cut: her first novel, Ticknor, which isn’t at all autofiction-y but clearly fits in with her later concerns. It’s a very interior work of historical fiction narrative about a man who feels envious of a more famous friend. Heti writes a lot about the desire/distrust of fame and wanting to do great artistic/literary work, especially in Alphabetical Diaries, so it was interesting to read passages from Ticknor, even though I ended up not referencing it in my review!

In addition to these longer works, I also read:

• Heti’s interviews, essays, reviews of other books, and shorter prose pieces

• Literary criticism about Heti, autofiction, and literary process

• Books about potentially similar artists/writers

Reading Heti’s writing

There were 3 pieces that had a major influence on my review:

1. A 2014 review Heti did of another autofiction writer, Karl Ove Knausgaard

2. A 2012 interview where Heti describes the process of writing How Should a Person Be?, her breakout novel

3. A 2022 series of conversations between Heti and various AI chatbots

I’ll elaborate a bit on what I got from each—

The first piece is Sheila Heti’s 2014 review of second volume of Knausgaard’s autobiography, My Struggle, for the LRB. In the opening paragraphs, Heti describes how impressed she was at Knausgaard’s seemingly “photographic…novelistic recall of his own life”—and the disillusionment of meeting him and learning that he’d made up certain scenes.

It’s a strange and fascinating anecdote: why was Heti disappointed? What does this say about Knausgaard’s approach to fiction versus Heti’s approach? I wanted to put this detail in my review somehow, especially since—when I first read my review copy of Alphabetical Diaries—I’d written down themes like:

• suspicion of narrative/story

• obsession with truth

• tired of autofiction

So it seemed like there was something here—something about Heti’s relationship to truth and fiction—that was worth drawing out in my review. This anecdote is included in the second paragraph of my review, which begins with:

But if a work of autofiction appears to be an unmediated reflection of the writer’s actual life, the reality is anything but.

The second was a 2012 interview with Heti in The Millions, shortly after How Should a Person Be? was published. This seems to be the novel that made her famous—it kicked up an enormous critical furor, with James Wood panning it in the New Yorker; men and women describing it as raw, unpolished, self-absorbed; and other writers, largely women, defending it. From the vantage point of 2024, it feels revelatory: like it or not, How Should a Person Be? had an enormous influence on contemporary fiction and autofiction.

A key part of the novel’s origin story—referenced in many, many articles about Heti—is that Heti’s previous editor at FSG read the novel and passed on it. It’s a salient detail because the novel ended up becoming a hit, and everyone loves a spurned-writer-shows-up-her-skeptics plotline.

This story is referenced in her Millions interview, but with a new, intriguing detail I hadn’t seen before:

The book was much, much more fractured in its earlier form…I tried to show [Lorin Stein, Heti’s former editor] that it was better than he understood by explaining to him this complicated process I had used to write many of the sections, with three decks of cards. It didn’t seem to change his feelings about it!

Decks of cards! How did they work and what did they do? As Heti explains:

I was trying to write a book that came from the world. I had note cards with all these sentences, which I carried around with me. I started recording myself narrating. I wanted to find a new way to write that would take things from the world.

I went through a crazy period where I had all these cards. Each card had symbols on them. It was a way of making scenes. All the symbols related to something real that happened in my life. I reduced those anecdotes down to a word and then I put those words on cards, and put those cards together randomly with a few other cards and then I’d try to come up with a scene from that. So it was a way to try and write about life, but not write about my life.

I can’t describe how excited I was to come across this detail. I’m fascinated by artists and writers who use generative processes in their work, who create works by first creating a system—with deterministic or random processes involved—that then generates an artwork or a text.

I studied computer science in undergrad, so this obsession is likely explained by the fact that I spent my formative intellectual years working on data structures and algorithms problem sets. And perhaps it’s this background that gives me a very specific critical sensibility.

The whole piece was oriented around my belief that Heti is frequently and readily understood as a confessional writer, an autofiction writer, a woman writing about her little romantic entanglements and little artistic anxieties—and not as someone who is constantly developing new and interesting processes for writing fiction and narrative.

You see this very clearly in Heti’s Paris Review series titled “Hello, World!” and published in 2022. The phrase hello, world is often used when teaching people to program, a tradition that dates from a 1974 tutorial written by the Canadian computer scientist Brian Kernigham. Here, Heti deploys the phrase to describe her conversations with various AI chatbots—and her experiments with creating her own. In the final installment, Heti even mentions enrolling in a programming and computer science class to try and “understand the bots from the inside”.

The conversations themselves are signature Heti: frank about sex (but cautious about indulging in it too much), invested in spirituality and philosophy, deeply searching, helplessly sincere. And the “Hello, World!” series is further evidence of Heti’s interest in trying out all kinds of technological tools—the emails and tape-recorded conversations of How Should a Person Be?, the Excel alphabetization of Alphabetical Diaries—in order to generate new and interesting ways to write about life.

Reading about Heti

I also found two review essays by the peerless Christian Lorentzen incredibly helpful in my review. The first was a 2018 essay for Vulture, titled “Sheila Heti, Ben Lerner, Tao Lin: How ‘Auto’ Is ‘Autofiction’?”, where Lorentzen writes:

Heti’s and Lerner’s books don’t lack artifice — they are novels, however their readers receive them — but the artifice is in service of creating the sensation that there’s no artifice, which is the whole point.

This is a fascinating and, to me, extremely precise analysis of how we read and receive autofiction. It seems like real life—and indeed readers often believe it is real life, that the I in the text is exactly who the author is, instead of an I that the author is intentionally constructing.

The other Lorentzen essay I relied on was his Bookforum essay, published one year later, titled “Tell Me Everything: Fiction in the age of radical transparency”. In my review, I summarized his argument as:

The critic Christian Lorentzen has argued that autofiction succeeds not because of its confessional qualities, but its constructed nature.

I then go on to describe the specific constructed nature of Heti’s works, emphasizing her use of random processes (coin flips, drawing from decks of cards) and technological tools.

Reading about other artists/writers/musicians

Shortly after I was commissioned to write this review, I had a conversation with my friend Chandler about Brian Eno and his use of chance and randomness in music composition, inspired by the I Ching. I’ve written a bit about my fascination with Eno before—

oblique strategies for starting a new project

—and part of the reason I became so obsessed with him is that his approach reminded me of what Sheila Heti had done in her 2018 book Motherhood. My ArtReview piece explicitly compares them:

Heti’s use of alphabetisation, along with her interest in mystical, mathematical and technological processes, might place her in a different lineage of artists and writers than autofiction: less Knausgaard, more Oulipo and Brian Eno (whose Oblique Strategies cards were also inspired by the I Ching).

I’m quite pleased with the gentle rhyming—or is it assonance? poets, please help!—of the phrase “more Oulipo and Brian Eno”.

Was it ridiculous to read a whole book just to pull off this one sentence? Maybe. But all this research isn’t really just for the review; it’s for my overall development as a writer and critic.¹

---

The online review, by the way, is about 1,000 words long. A shorter version, around 650 words, is in the February print issue of ArtReview.

All of that research ended up compressed into a fairly short piece—shorter than most of my Substack posts, actually. I’m relaxed here, at ease; and so my writing runs longer, repeats ideas, slowly approaches some point. But in my review I had to be crisp, contained, and exceptionally rigorous.

Working with Orit helped tremendously; she did a masterful edit of my initial 1,200 word draft to get it to the 650-word print limit. (I honestly don’t think I would have been able to do this myself!)

And I’m generally quite pleased with the review! My goal was to write something intellectually ambitious: to say something about this specific book, and Sheila Heti’s overall approach as a writer, and uses of technology in art and writing, and autofiction as a genre, and how autofiction is discussed in contemporary literature culture.

I think I’ve managed to do that. It’s nice to take this moment to acknowledge that, to feel proud—and now on to the next project, the next review, the next opportunity to grow as a writer and critic. ✦

1. Similarly, for my first-ever piece of literary criticism—on Mieko Kanai’s Mild Vertigo for the Cleveland Review of Books, I read Roland Barthes’s Camera Lucida, which had been on my to-read list for years and years, in order to write these sentences:

This nameless affect was articulated by the French literary theorist and critic Barthes, who wrote in Camera Lucida that a photograph has the power to produce “an internal agitation, an excitement, a certain labor too, the pressure of the unspeakable which wants to be spoken.” Kanai is a devotee of Barthes (an untranslated work of hers is named after Camera Lucida), and Natsumi’s encounters with photography depict this agitation beautifully.

Writing projects are such a nice opportunity to research the things I’ve always wanted to learn more about—in this case, photography theory.