Computational Criticism of Classical-Chinese Residues in Lu Xun (1903–1936): A Six-Function Taxonomy, Sublanguage Detection, and the Register × Stratum Two-Dimensional Model

The Ghost in the Machine: Why Lu Xun's Words Were Never Just "New"

Imagine you are trying to understand a person by listening to them talk. If they grew up speaking one language at home but switched to a different, modern language for school and work, you might expect them to sound completely different in every situation. But what if they didn't? What if, deep down, their brain was still arranging thoughts in the old way, even while using new words? This is the kind of mystery that "computational criticism" tries to solve. It's a field where computer scientists and literary detectives team up. Instead of just reading a book and saying, "This feels old-fashioned," they use software to count every single word, sentence structure, and punctuation mark across thousands of pages. They treat literature like a giant dataset, looking for hidden patterns that human eyes might miss.

The big question here is about the "ghosts" of the past. For centuries, Chinese writers used a formal, ancient style of writing called wenyan (classical Chinese). In the early 1900s, a massive movement called the New Culture Movement tried to kill that old style and replace it with baihua (modern vernacular), which sounded like how people actually spoke. The leader of this movement was a writer named Lu Xun, who is famous for writing in the new, modern style. But everyone suspected that Lu Xun was still haunted by the old style. The big debate was: Did he truly abandon the old language, or was it just hiding? This paper uses a massive computer analysis to finally answer that question, not by guessing, but by measuring exactly how the "ghosts" moved around.

---

The Detective Work: Counting Lu Xun's "Old Ghosts"

Lu Xun (1881–1936) is the superstar of modern Chinese literature. He is the guy who helped invent the modern Chinese novel. He famously said that the old classical language was like "ancient ghosts" on his back that he couldn't shake off. For a long time, scholars argued about what those ghosts were doing. Were they just accidental habits? Or was Lu Xun secretly using them on purpose?

This paper, written by Jun Zeng, decided to stop guessing and start counting. They built a massive digital library of Lu Xun's entire life's work—about 4.37 million characters. This wasn't just his published novels and essays; they also included his private letters and diaries, which had never been studied together on this scale before. They used a computer to scan this mountain of text and look for six different ways Lu Xun used the old language.

Think of it like a chef who grew up cooking with a specific set of ancient spices. The computer asked: Is the chef still using those spices in the main dish (public writing)? Is he saving them for a secret family recipe (private writing)? Or is he just sprinkling a pinch in when he runs out of modern ingredients?

The Big Surprise: The Ghosts Didn't Leave; They Moved House

The most shocking discovery is that the old language didn't just fade away. It didn't retreat uniformly. Instead, it did a strange dance.

When Lu Xun wrote for the public—his famous essays, stories, and translations—the amount of old-fashioned language dropped sharply. It went from about 0.0994 (early in his career) down to 0.0655 (later on). It looked like he was successfully switching to the new language.

But here is the twist: When the researchers looked at his private letters and diaries, the story was completely different. In his private notes, the old language didn't drop; it actually increased. It went from 0.0383 early on to 0.0695 later on. In fact, in his later years, his private writing was more full of old language than his public writing.

This proves that Lu Xun didn't lose his classical training. He just moved it. He kept the "ghosts" in his private room where no one could see them, while presenting a cleaner, modern face to the world. The paper calls this a "register-shift mechanism." It wasn't a retreat; it was a migration.

The Deep-Down "Sublanguage": The Brain That Never Changed

Even more interesting is what the computer found "under the hood." The researchers built a special detector to find the sublanguage. This isn't about using old words; it's about using old thinking patterns.

Imagine you are speaking English, but you are arranging your sentences like you are speaking Latin. You might use modern words, but your brain is still building the logic in the old way. The computer measured this "sublanguage index" and found it was incredibly stable. It stayed almost exactly the same from the beginning of Lu Xun's career to the end (around 0.67).

Whether he was writing a public essay, a private letter, or a scholarly book, his brain kept using these deep, ancient patterns. The paper suggests that the "old ghosts" weren't just on his back; they were in his brain. Even when he switched to modern words, the way he connected his thoughts remained deeply rooted in the classical tradition.

The "Stopgap" Mystery: Why He Used Old Words When He Was Stuck

The paper also looked at why Lu Xun used the old language. They found two main reasons:

1. Expressive: Using it to tell a story or argue a point (this dropped by 71.1% over time).
2. Stopgap: Using it when he couldn't find a modern word for something.

Here is a funny finding: You might think that if you are writing a private diary, you are free to use whatever words you want, so you'd use the old "stopgap" words the most. But the computer found the opposite. Lu Xun used the "stopgap" old words most in his public essays (360.68 times per 10,000 characters) and least in his diaries (129.23 times).

Why? Because his public essays were the hardest. They were complex, argumentative, and full of big ideas. When he got stuck trying to explain a difficult concept in modern language, he would reach for an old classical word as a "backup tool." His diary was just him talking to himself, so he didn't need the backup. The paper argues that he didn't use old words because he was "free" to; he used them because he was complex and needed help.

The "European" Myth: Old and New Were Not Best Friends

There was a popular theory that Lu Xun used the new, European-style sentences (long, complex, with many "and" and "because" words) to help translate his old Chinese thoughts into modern language. The idea was that European grammar was the bridge between the old and the new.

The computer tested this, and it was wrong. The data showed no connection between the "European" style and the "old Chinese" style. In fact, they seemed to be independent. The paper suggests a new idea called the "tension model." Instead of one helping the other, Lu Xun used the long, winding European sentences and the short, punchy classical phrases at the same time, creating a unique, tense, and powerful style. They weren't working together; they were fighting, and that fight made his writing so strong.

The Final Verdict: A Two-Dimensional Map

So, what is the final picture? The paper proposes a new way to look at Lu Xun's writing using a two-dimensional map: Register (Public vs. Private) and Stratum (Surface words vs. Deep thinking).

• Surface Level: In public, the old words faded away. In private, they stayed strong.
• Deep Level: The old thinking patterns stayed strong everywhere.

The paper concludes that the "victory of modern language" wasn't a simple story of the new replacing the old. It was a selective migration. Lu Xun didn't abandon his past; he compartmentalized it. He kept the "ghosts" in his private life and his deep brain, while wearing a modern mask for the public.

This study doesn't just tell us about Lu Xun; it shows us how computers can help us see the hidden layers of literature. It proves that sometimes, the most important things aren't the words you say, but the invisible patterns of how you think. And in Lu Xun's case, those patterns were as old as the ghosts he claimed to shake off, but as alive as the modern world he helped create.