Rampant plagiarism in the Chinese literary world
« previous post |
"It cannot read the human heart" by Yan Ge (b/1984), London Review of Books Blog (2/20/26)
Since November 2024, a book influencer on RedNote has been publishing posts featuring side-by-side excerpts from works by different authors that contained similar, and in many cases identical, sentences and paragraphs. Among those whose sentences, similes, descriptions, scenes and plotlines appeared to have been copied and pasted were Eileen Chang, Hsien-yung Pai, William Faulkner, Orhan Pamuk, Annie Proulx and Gabriel García Márquez. The perpetrators of the apparent plagiarism were a number of contemporary Chinese authors.
‘Why are so many writers “borrowing” from others’ work?’ my friend asked. ‘Is this some kind of open secret in the literary world?’
I had no answer. In more than twenty years as a writer, I have previously encountered only a couple of incidents of outright literary theft (as opposed to quotation or allusion). Both times, I was baffled by it. Plagiarism, it seems to me, is a humiliating admission of artistic failure.
Digging deeper into the causes for the widespread plagiarism that she was encountering, Yan discovered one potential reason for the rapid rise in these corrupt practice cases:
The discovery was made possible by AI-powered plagiarism-checking applications, but some people have suggested that the plagiarism itself may have been fostered by the use of large language models. Given the data that AI models are trained on, wasn’t it possible – inevitable, even – that any writer who used AI for prompting or editing would end up copying, inadvertently, the work of others? The trouble is that much of the apparent plagiarism was published in the early 2000s or the 1990s. So unless someone invents a time machine, the theory doesn’t hold.
Moreover, says Yan,
If plagiarism is defined as having sentences flagged as identical by a checker, then so be it. But the software can only scan texts mechanically; it cannot read the human heart … This so-called reader who exposed the identical texts, you are not a reader in any real sense. You just used the software, being too lazy to read anything yourself … You are merely a reader who is not illiterate.
There is yet one more outré hypothesis about what may have served to promote plagiarism:
Other online analysts noted that a number of the authors involved had attended creative writing MFA programmes, which have been a feature of Chinese universities for the last fifteen years or so. ‘So this is how they teach writing in the universities,’ people speculated. ‘They simply get the students to memorise the classics and graft the masters’ sentences into their imitations.’ The opinion echoed a long-running scepticism towards the institutionalisation – or, as some would have it, the industrialisation – of writing.
In the final analysis, after consulting with another friend, Yan came to the conclusion that the plagiarizers were doing it for money. Creative writing, especially for state-funded journals, is so highly lucrative that, if you steadily churn out one or two stories a month for them, before long you will be in the top five per cent income bracket.
Yan has been writing in English in addition to Mandarin and Sichuanese. Her first English book is a 2023 short story collection Elsewhere: stories. Reviewer Chelsea Leu wrote
Yan Ge’s English debut is preoccupied with language, its failures, and its relationship to human emotions and the raw reality – the 'food' – of life. … These stories map out the distance between the head and the gut – the way language can fail to convey the deepest, most visceral facts of life."
Reviewer Sindya Bhanoo wrote that the stories "explore the power of language across the Chinese diaspora to either bring people together or push them apart."
If there's not a dramatic turnaround soon, these practices will take all of the fun out of writing — and reading.
Selected readings
- "Tik Tok and Red Book" (1/19/25)
- "AI plagiarism" (1/4/24)
- "John McIntyre on varieties of plagiarism" (3/30/13)
[h.t. John Rohsenow and thanks to Jing Hu]]
Martin Schwartz said,
February 25, 2026 @ 11:19 pm
Why should AI have to plagiarise?? It now writes stories online,
e.g. dramas about children foiled at trying to rip off their aged parents,
with robotic narrators, but one would not necessarily guess.
I've seen some AI poetry which may pass muster (for at least a plastic hotdog). Last night BBC reported a robot conductor for the Swedish Philharmonic Orch.; a reviewer complained about the robot's
expressionless face. That may be resolved by now, easy enough.
And no, I'm not a
Martin Schwartz
robot
(yet)
Oy.
Lucas Christopoulos said,
February 25, 2026 @ 11:47 pm
AI does not have the ability to achieve the same specificity of analysis from different perspectives. This is part of what makes the human mind so remarkable. The way an individual’s mind works is inherent to that person, and AI cannot follow the same stream of thought. Diversity of knowledge certainly helps, but the perspective of expression is intrinsic to the biological mind.
John Swindle said,
February 26, 2026 @ 2:41 am
Not only sentences but whole books can be copied, in which case the English term is "pirating."
I don't remember whether you've discussed the remarkable 2012 novel 花冠病毒 (huāguān bìngdú, styled “The Corolla Virus" in English), by Bi Shumin 毕淑敏. Bi, a Beijing psychiatrist and writer, fought the SARS epidemic and in her novel predicted the next one, which did occur and became known as COVID-19. When I was able to find a copy to read (nope, my Chinese wasn't good enough) and donate to my local library, it claimed to be written by someone else entirely. I forget who—the library accepted the donation and listed the author correctly as Bi Shumin. What little I could find online confirmed that it was the same work. As far as I know it hasn't been translated into English yet.
Scott P. said,
February 26, 2026 @ 8:44 am
Why should AI have to plagiarise?? It now writes stories online,
e.g. dramas about children foiled at trying to rip off their aged parents,
with robotic narrators, but one would not necessarily guess.
That's how AI works, it digests the content of the training material and uses that to predict what is most likely to follow a prompt. So everything a LLM produces is plagiarized.
ajay said,
February 26, 2026 @ 9:21 am
If plagiarism is defined as having sentences flagged as identical by a checker, then so be it. But the software can only scan texts mechanically; it cannot read the human heart … This so-called reader who exposed the identical texts, you are not a reader in any real sense. You just used the software, being too lazy to read anything yourself … You are merely a reader who is not illiterate.
It might be worth clarifying that this quote is not from Yan Ge herself, but from one of the plagiarists!
Yan's own complaint also seems a little unfair:
When I lived in China, I fretted over the fact that I couldn’t write more than a couple of short stories a year, failing to realise that I was missing out on making money unless I broadened my resources.
I don't think there has ever been a time when any professional writer could realistically expect to survive on an output of two short stories a year.
ajay said,
February 26, 2026 @ 9:22 am
John Swindell: out of curiosity, how does a novel that has not been translated into English have an English title?
Victor Mair said,
February 26, 2026 @ 10:22 am
That's a reasonable question, and it even passed through my mind fleetingly as I read John Swindell's comment, but I quickly recalled that it is not unusual for Chinese books, films, etc. to have English titles when their primary content is in Chinese.
ajay said,
February 26, 2026 @ 10:54 am
Ah, I did not know that. On searching I see it has a Chinese title as well, and both appear on the cover of the book. Interesting! Why do they do that, do you know?
Ted McClure said,
February 26, 2026 @ 11:05 am
@ajay: It's Library of Congress cataloging practice to create a translated title of a foreign language work for its English language record if the work doesn't already provide one.
cervantes said,
February 26, 2026 @ 11:17 am
This is not really anything new. It's a fairly modern standard that you don't copy from extant works. Check out the Gospels if you don't believe me. Tristam Shandy is a famous novel and it's heavily plagiarized. I could go into this at greater length but you get the idea.