the poetry of vertical video

sonnets and tiktok

I didn’t really like memes until I was about nineteen. Before memes, I was most interested in poetry.

Sometimes people are surprised by this when I tell them, but I always felt the transition was perfectly natural. One day, walking out of a class on John Donne, I looked down at an Instagram meme page I followed and realized it was kind of doing the same thing, and was equally interesting and moving to me. This was in the late 2010s, that golden age of Instagram meme pages which ushered in the -cellectuals moment around the time of the pandemic, and still carries on in some form today as strands of DNA distributed scattershot across the whole internet.

I loved old and elaborate poetic forms the most because they exposed something happening underneath everyday language. In structures like meters, stanzas, and caesurae, I saw the stuffness of language. I traced the cadence of breath in punctuation and meter, the unspooling of ideas and images across lines. I learned my feelings and thoughts were not just some inclement interior weather I had to sit under, but a substance made of something called language that I could learn to pick apart and rearrange. I felt a need to do that, and still do — it gave me some sense of self-mastery, and of understanding.

But I perceived that one thing which made the life-world of John Donne so different from my own was the presence of digital textuality. Most of the reading I did was and remains on a screen. I wanted to look at the stuffness of that particular form of language, because that’s what surrounded me. I started at the beginning, studying early internet hypertexts from the 1990s. As a kind of meditative practice, I started writing my own poems on three-dimensional shapes and translating those shapes into hypertext structures of HTML files interlinked with Javascript code.¹

the kind of meme I was looking at in 2017

When I first started making TikToks, years later in 2023, I studied direct-to-camera vertical video like it was a kind of poem. I kept a document where I observed videos made by creators I liked and sketched out diagrams of how I imagined their editing timelines must look. I even copied out verbatim what they said, typing alongside the video. I also conducted research, asking some creators I’d met vaguely through my work at Know Your Meme for their takes on the form.

I knew I’d never been a particularly gifted poet, but I could competently write to a meter and a rhyme structure once I came to understand it. I figured the same would be true of vertical video. The poetic form which direct-to-camera vertical video reminds me most of is the sonnet.

John Donne

A sonnet is fourteen lines arranged in geometrical proportion. There are two general types. The Petrarchan sonnet, which is older, sets eight lines against six — elaborating an idea in those opening eight, and then refining it in those closing six. The Shakespearean sonnet sets twelve lines against two — accelerating across three four-line stanzas to a punch-line. The point where that switch happens (after eight lines in the Petrarchan, and twelve in the Shakespearean) is called the volta.

Typically, the sonnet has a rhyme scheme which maps onto that geometric scheme. Shakespeare sonnets rhyme like this: the quatrains go ABAB, with interlocked rhymes, then CDCD, EFEF, and finally a couplet GG on the punch-line. Typically, an English sonnet like the ones Shakespeare wrote also is made of iambic pentameter lines — ten-syllables long, which follow a particular rising-falling vocal rhythm, a meter.

One of the reasons why these poetic forms like the sonnet were so popular in the past was the specificity of writing to these rules: The sonnet enforced a discipline that made it so even untalented people, if they followed the template, could produce something compelling — and I’d argue we see the same thing with direct-to-camera vertical video forms today. Anyone can be a yapper, with reasonable success, so long as they approach a difficult form with patience and curiosity. Like sonnets, which travelled across commonplace books transcribed by hand, vertical videos invite their consumers to become creators.

Vertical videos do not have fourteen lines. But they do have very strict requirements about length. You want to hit under three minutes for Instagram Reels. TikTok caps you at ten, but going over about four and a half doesn’t work. Like with the sonnet, a lot of meaning needs to be contained in a small package, and so density is the way it happens — and part of the charm of vertical videos is their play with that density, just like part of a sonnet’s charm.

Every word in a sonnet is doing several jobs. It is participating in a punctuated sentence, it is a piece of line, it is setting up for arrival at a rhyme, it illustrates the character of a speaker, it is fitting into the meter, it is advancing a story or argument, and it is participating in side hustles such as alliteration and metaphor.

Every word in a vertical video is also doing several jobs. It advances a story or argument, it participates in a sentence, it illustrates the character of a speaker, it is a piece of a breath, it is pairs with gestures by the face and hand, it fits into a meter (in the case of vertical video, the irregular prosody creators use for retention that Adam Aleksic has brilliantly described in his work on influencer accents). The word also becomes part of a cut, which I perceive as the units which make a vertical video in the same way lines make a poem.

The audience learns to appreciate the way each of these balls is juggled, and experiences each game language plays in the poem or video.

The resulting structures are small and portable. What allows so much meaning to get contained in such a tiny container is the total saturation of that space — each second and each word means in several ways, across several dimensions.

Part of the difficulty of poetry for a lot of people is this density which we no longer have the cultural IQ to untangle — but we still have that IQ for vertical videos. The bafflement some Boomers may feel looking at a TikTok video is just like the bafflement a young student has confronting a Shakespeare poem for the first time. And in both cases, once the initial hump is overcome, the beauty of the text offers itself.

Another reason is the cultural alienness of both forms — a Shakespeare sonnet’s archaic language, or the TikToker’s use of words like “skibidi” and “sigma” confuse people. But the density, to me, is the main thing.

Just as the sonnet works within the geometric structure of the volta, the vertical video works within the structure of the hook. The starts of most videos are more important than their endings, since most people arrive at an impression almost immediately and decide whether to scroll or not. But the hook, to me, is less like a structure that “grabs” attention, and more like one that suggests a path of least resistance. It offers a clear, direct line through the content, a point to anchor on and — as you do with an anchor — drift outwards from. You need micro-hooks to reiterate, recenter, drum down the theme. Around the midway point of a video, I’ll usually throw out something that could be the first line of its own video — and then work towards an ending.

As a creator, one learns to juggle in different ways, mix up the act a bit, or more tightly and finely execute the form. By a process of evolution, coordinated somehow between audiences, makers, and algorithms, the traits of the form emerge and evolve — a video that does well today might not have done well in 2021, and the same is likely true of 2027. The artistic form changes as new people come in, innovate, and shift expectations.

There is a lot of bad poetry in the world, just as there are a lot of bad videos on TikTok. But the best ones take the limits of this form and push it.

My suspicion about sonnets (which I’m not sure if scholars of Early Modern textuality would back me up on) is that fourteen lines is about the natural amount a printer would want to put on a single page. It’s also the amount it takes for you to get to the end and have forgotten exactly what the first line says. It is the longest kind of short poem and the shortest kind of long poem, dwelling at a kind of attentive Goldilocks zone.

There is also something about the sort of attention a sonnet demands — a meditative, scanning sort of attention. And this is something that folks tend to miss about short-form video: it calls for the same attention. We may be scrolling through dozens and dozens of them, giving two seconds of our time here and ten seconds there, attention spans frayed, flitting, frazzled. But when we find a video we want to watch, we choose it and we watch it. We go so far as to press down in the corner and fix our finger there so it runs at 2x speed, making an effort to override our wandering mind. We hunt for the videos we want to allocate more time to, and the best vertical videos lock you in. I see a video from someone I love and I’m 100% there with them. There is deep intention and care to the practice, once the video is chosen, because it is so easy to not choose it.

I’m not offended when fifty percent of people stop watching me after ten seconds. They’re on the hunt. It’s not me, and at least I’m part of helping them, and the algorithm, know what they don’t want so they can find what they do want.

At some point, I want to time the watching of videos on my scroll or the scrolls of others — how much of an hour on TikTok is spent “hunting” for videos, and how much is spent “lingering” on them? Is it like twenty minutes of hunting through a hundred or so videos, and forty minutes of lingering on fifteen two-to-four-minute videos? Does it differ across niches in which people are involved, age groups, and so on? Someone who can do numbers help me conduct this study.

A comparison’s entropy might be its most charming quality. Make any kind of analogy gluing two things together and it holds firm at certain points — but the longer you knead it around in your head, the looser and thinner it becomes. The places where it begins to fall apart, and entropy interjects, are the loveliest.

The comparison of vertical videos to sonnets falls apart when you consider the ephemerality and quantity of vertical videos, which differs from the heaviness and slow process of sonnets. There’s also the role of the other forms which enframe and neighbor each — the play the sonnet-reader would watch after is different from the slop reel the DTC-watcher will see next; just as the political-economic structure a sonnet participates in (the clout game of a royal court, or the nascent market dynamics of early bookselling) is different from the algorithmic casino a vertical video plays in. But maybe the comparison falls apart most when you take into account the vertical video’s immediate pliancy to an audience’s touch: A sonnet is a thornier, tougher kind of material.

The charm in a comparison’s falling-apart lies in the confession it entails: as speaker and audience y’all were always playing pretend, holding contraries together to manufacture a moment of insight that, like two magnets pinched close at repelling poles, was bound to snap apart. The precarity of metaphor, in particular, is what makes its pretendness visible, and luminous. It is the invitation a kid will offer to another kid that turns the stick found on the ground into a sword which slays the dragon that has been conjured from the air.

And that precarity of the imaginative act — its dependence on a kind of contract between people who agree to spend a moment thinking as if the world weren’t the way we know it is — is maybe what most makes lyric poetry most like direct-to-camera video. It is a voice finding a mind and saying let’s sit together here, within a moment, and be otherwise than expected.

1. Someday, I want to do a Live on TikTok or here where I read through these poems and have people vote on which link to click. Writing it in this footnote to lock in the aspiration.