According to Colin Gorrie’s 2025 rhetorical analysis on Dead Language Society, the writing patterns readers now brand as AI slop are classical rhetorical devices deployed without restraint. Parallel structure. Balanced contrast. The form Aristotle called antithesis.
Aristotle laid the devices out in Book 3 of the Rhetoric. Here is the example he used, rendered for the ear:
“They helped both groups. They helped the ones who stayed home and the ones who went with them. The travelers gained more land than they had before. The ones who stayed kept enough land to live on.”
Aristotle explained why the pattern works. Opposite ideas become clear when you set them next to each other. The arrangement feels like proof. Place two opposing claims side by side and one of them looks weaker.
He named the related tools. Parisosis makes the two halves of a sentence the same length. Paromoiosis makes the endings sound alike. A single sentence can carry all three at once.
The Rhetorica ad Herennium and Quintilian’s Institutio Oratoria taught the same toolkit. Roman students practiced these constructions the way modern students practice five-paragraph essays. The devices produced clarity, force, and memory.
Writers now circulate a prompt that forbids the exact tools. No antithesis. No negative parallelisms. No contrasting pairs. No parallel sentence structures inside a paragraph. No rule of three. The list circulates on X as the fix for AI-sounding prose.
The ban treats the classical handbook as a defect list. LLMs trained on the Western canon absorbed the patterns. Most living authors never opened the books that taught them.
Why it Matters for Authors
The machine has read the material you skipped. When it leans on the old devices, readers hear formula. The reaction often carries status threat. The system that studied the great books produces the rhythms that once marked elite education. Authors who never studied those books hear contamination instead of recovery.
True slop still exists. Hollow consensus language. Missing specifics. Zero lived judgment. The classical ornaments become noise only when the model piles them on without purpose.
Read the short stretch of Aristotle’s Rhetoric that covers style. Book 3, chapter 9 runs a few pages. Learn the names of the devices so you can choose them. Deploy contrast when the idea needs it. Balance clauses when the ear wants weight. Refuse the automatic version the model reaches for first.
The human edge remains judgment and specificity. The machine owns the average pattern. You own the decision to use the pattern or leave it on the shelf.
Sources:
- Colin Gorrie, “Why ChatGPT writes like that,” Dead Language Society, July 2025
- Aristotle, Rhetoric 3.9 (standard English translations of the antithesis example and the explanation of why the form satisfies)
- Rhetorica ad Herennium Book 4 on contentio
- Quintilian, Institutio Oratoria 9.3 on antitheton
Commentary
Thomas: Those techniques became foundational, not just for Greek and Latin, but for every Western language, because they've been trained into Western writers for the last 2,200 years. Until five minutes ago, everybody considered educated had studied Aristotle. If you hadn't, you weren't considered educated, for almost all of Western history. It was almost a prompt injection of Aristotle into the corpus of Western language.
Jonathan: Which means AI has read all the books you skipped or hated reading in school.
Thomas: AI hasn't just read Aristotle. It's read everything Aristotle influenced over the next 2,200 years. I discovered this while working on a project called Daily Aristotle, which now has a website at DailyAirstotle.com.
The vision is a podcast you listen to for a year and come out the other side having gone through all of Aristotle. It's still a work in progress. I'm using AI to translate him, but I'm trying to keep it true to the original without watering anything down.
I created a character called Grammaticus who does the intros for each book. I wanted to separate Aristotle from the modern scholarship that has, in a lot of ways, ruined him.
Too many people treat Aristotle the way Vizzini treats Plato and Socrates in The Princess Bride, "Morons." It's not that Aristotle was always right. But you can't just disregard him, which is basically what scholarship has been doing since the modernists and postmodernists went to war with him.
I didn't want the commentary on Aristotle to be modern slop because the whole point of the project is to deprogram people from modern slop and reconnect them with the classics.
A lot of people listen to scripture every day. Daily Bible podcasts run five, ten, twenty, thirty minutes and slowly deprogram listeners from the secular world. Aristotle can do the same thing, but from the modern world. If you're trying to get back to the greatness of Western civilization, Bible every day and Aristotle every day is like skating on two skis instead of one. It lets you go faster.
So, Grammaticus explains Aristotle to a young patrician's son named Aurelius in the first century BC, using only the ancient commentaries on Aristotle, not the modern ones. Every great thinker throughout history has written commentary that helps Aristotle make sense, so we don't actually need modern scholarship to understand him. We already have the ancient sources.
Is "AI slop" actually just Aristotle?
Thomas: While I was building Grammaticus's voice, I ran a "stop slop" skill I'd shared on authormedia.social, a common list of overused AI phrases and how to cut them. When I ran it against the Aristotle material, the AI told me I couldn't use it, because half of those techniques aren't just things Aristotle would say. They're things Aristotle invented.
Half of what we call slop is just cliche. AI is now making some of Aristotle's own techniques feel overused, but they're still effective. You have to be careful not to strip out the greatness of writing along with the slop.
Take putting things in threes, giving three examples back-to-back in a rhythm. That's a technique of good writing. Or putting things in twos, like Hansel and Gretel, or peanut butter and jelly. There's nothing magic about the number three specifically, but the pattern and the sequencing matter, and the fact that AI overuses them doesn't make them bad.
What does it take to judge an AI's writing?
Thomas: To judge AI output, you have to know what greatness looks like yourself. You need to know what a platonic ideal is. You need to read Aristotle. I'm building this project as much for my own benefit as anyone else's. It was my escape during a stressful season, I was literally translating Aristotle while my wife was in labor. I'd walk around with her through contractions, and when she needed to be alone, I'd go work out how to translate Aristotle accurately without watering it down. That tension is still something I'm navigating.
There's currently one episode up at DailyAristotle.com, and I keep replacing it.
The Categories is the hardest book of Aristotle's to translate, since he's essentially defining categories of things. If I can get The Categories right, the rest of Aristotle will follow. Eventually I'll release more episodes.
What happens when we stop reading the greats?
Thomas: I want to encourage everyone to study Aristotle, because studying great literature is the only way you earn the authority to judge AI. I keep noticing a parallel to what happened in the South during Reconstruction, when white Southerners suddenly had to compete with freed slaves, some of whom turned out to be smarter and better educated. They didn't like that, because they wanted to see themselves as superior.
I see the same bias now between humans and AI. There's an assumption that because you're human, you're automatically smarter, and the AI is automatically dumber. But the AI has read Aristotle, and you haven't. It's hardworking. You can be smarter than the AI, but not just by virtue of being human. You get there by getting an education, by reading Aristotle, Socrates, Plato, Thomas Aquinas, the greats of the Western canon. A lot of authors haven't read a single classic. Nothing older than 50 years. For them, The Lord of the Rings is as far back as it goes. They haven't even read the original Conan books by Robert Howard.
If you haven't been reading great literature, you'll be the first person AI outcompetes. Go buy some old books. They're out there, they're cheap, and the oldest ones are out of copyright, so AI can't touch them. Great literature will persist, but only if we actually read it. What's the point of preserving Aristotle for 2,500 years if nobody reads him? Countless scribes gave their lives copying his words from worn manuscripts onto new ones, each one deciding Aristotle was worth the effort. His work nearly got lost several times, then was preserved, then rediscovered. We owe it to our ancestors to read what they thought was worth sacrificing to preserve.
Jonathan: The way people put information together is the same way AI does it. You're building a network you can draw on, a library in your mind, so you have material to pull from when you're problem-solving in your writing. What word do I choose? What direction do I take the plot? How do I build this character's arc? If you don't have an adequate library to draw from, from the greats specifically, you're stuck.
As a fighter, I learn from the greats because they solved problems differently than I would. If you're fighting someone who's 6-foot-4 and 280 pounds, and you're a 5-foot-9, 40-year-old dad, you take lessons from Muay Thai masters. You learn leg kicks, because something has to hold up all that weight. You learn when to shoot for a leg, when to grapple, when to strike. When the Gracies came onto the fighting scene, nobody had a counter to them, because nobody else was doing what they did. They studied the greats and used it to dominate mixed martial arts. The UFC is largely built on what they brought.
Thomas: Attitude matters here. You have to approach the greats with humility, not with "this is just a bunch of old white guys, what do they know." That doesn't mean you have to agree with everything they said. Aristotle said plenty of things he was wrong about. But the same is true of martial arts traditions. You can't dismiss a fighting style just because it's old. You have to respect it, ask what you can learn, and stay curious about why two thousand years of scholars invested their lives preserving it.
Jonathan: Why did this book send people to war? Why did people risk their lives over words on a page? What idea was powerful enough to start that?
Thomas: Aristotle mentored someone who turned out to be mildly influential in world history. His name was Alexander. Philip of Macedon hired Aristotle to pause his writing and tutor his son, who became Alexander the Great.
Alexander conquered most of the known world, then Hellenized it, meaning he spread Aristotle and Greek culture everywhere he went. That Hellenized world became the world we live in today. It shaped not just Europe but the Middle East and North Africa. St. Augustine wrote City of God from a town called Alexandria, named, you can guess, after Alexander. These were dangerous ideas precisely because they were powerful ones, not the kind of thing to disregard.
There's a strain of ignorance-celebration in the modern world that we have to escape if we want to preserve Western civilization, because Millennials and Gen Z are the ones who will decide whether it survives. That's an open debate right now that Gen Xers and Boomers don't even realize is happening, because their generation quietly took the steps that forced this conversation, including taking Aristotle out of the classroom. Most people didn't read Aristotle in high school. Some didn't read him in college. Nobody's teaching him anymore.
Aristotle is a placeholder here for all of the greats. We get the word "tactics" from a writer named Tacticus, who literally wrote the book on tactics, and it's genuinely good, though almost nobody's read it. Everyone knows The Art of War, but Tacticus's book deserves the same attention.
There's a huge amount of ancient literature getting lost, not because the books are being burned, but because they're being burned out of our minds. We're biased against them, treating new as automatically better and old as automatically foolish. That's a worldview, and it's a choice. There's ancient wisdom in these books, and if you have it, you can outmaneuver people who refuse to open their eyes, people who refuse to leave Plato's cave. And if you want to know what Plato's cave actually is, you have to read Plato, which the Daily Aristotle podcast doesn't cover.
What's next for the Daily Aristotle project?
Thomas: Once the translation harness is working, nothing's stopping me from building a daily Plato podcast, a daily Tacticus podcast, and others, especially once I bring the cost down.
Between the different AI models checking each other's work to keep the translation from skewing modern, the first version of an episode cost about $5. That adds up to more than $1,500 for a full year of episodes, and I'm honestly not sure I wouldn't be the only listener. I'm still working out the economics.
I'm learning a lot along the way, about AI, about ancient languages, about how language itself works. The very first chapter of Aristotle's Categories is literally about how language works and how words relate to each other. Rhetoric and Poetics aren't the only books by Aristotle authors should read. They need to start with the Categories.
