Skip to content
News

Young people obsessed with “unpopular secrets”

Qu Qiang lay for a long time in front of a display cabinet in the Turpan Museum in Xinjiang. The documents inside were covered with creases, and the light shone down, forming shadows. It is filled with dense words written in Chagatai script, and the curved strokes are connected with each other. Some of the writing has been blurred, and several places are covered by creases and damage.

He read it carefully and found that a familiar name appeared at the beginning of the document—that of a Turpan ruler whom he had studied in his doctoral thesis. He realized that this might be an ancient decree that had not been studied by scholars.

After returning to school, Qu Qiang spent more than a month translating the document and writing it into a thesis. The next year, he took the photos and asked foreign tutors for advice. The professor was so fascinated that he didn’t even bother to eat. He looked at it over and over for a long time.

Qu Qiang was born in 1995. He loved watching historical documentaries since he was a child. Now he is studying the history of the Western Regions in the Ming and Qing Dynasties at Nankai University. In order to understand these ancient documents, he learned Manchu, Mongolian, Persian, Chagatai and other languages. In the eyes of others, this is some obscure knowledge, but he can often find fun in it.

According to the announcement issued by the National Office of Philosophy and Social Sciences in 2020, unpopular subjects mainly refer to some traditional humanities fields and research directions with low academic attention, difficulty in producing results, and small research groups; unique sciences are endangered disciplines in unpopular subjects with unique cultural value, high academic threshold, extremely difficult research, small research groups or even no successors.

Oracle bone studies, slip studies, Dunhuang studies, paleography, endangered languages ​​(dialects) research, ethnic minority languages ​​and history research (Tibetology, Mongolian studies, Xixia studies, etc.), research on characteristic regional cultures, compilation and research of traditional documents and unearthed documents, etc., all fall into the scope of unpopular secrets.

In 2016, the state proposed to pay attention to “unique” and unpopular subjects with important cultural value and inheritance significance. In 2018, the National Social Science Fund set up a special project to support related research.

On November 6 this year, the country’s first unpopular secret arts academy was established at the University of the Chinese Academy of Social Sciences. On the same day, the National Unpopular Secret Arts Alliance was launched, with more than 80 universities, scientific research institutions and cultural institutions joining.

These knowledge, which seems far away from real life, attract some young people to enter. Some people use AI to study ancient texts from more than 3,000 years ago, some learn languages ​​that are on the verge of being lost, and some try to restore a vague history from ancient documents scattered around the world.

The Beijing News reporter interviewed several young scholars who are engaged in the research of unpopular secrets and listened to their stories about these ancient texts.

▲Yarkand Khan’s tax exemption document from 1600 written in Chagatai script. Photo provided by interviewee

An ancient document made the professor and I forget to eat

Narrator: Qu Qiang, 31 years old, a lecturer at the School of History at Nankai University. He studies the history of the Western Regions from the late Yuan Dynasty to the mid-Qing Dynasty; Persian, Chagatai, Mongolian documents; Manchu documents and Qing history, etc.

When I was a child, I lived in Fularji District, Qiqihar, Heilongjiang, which is an old industrial town. “Fularki” means “Red Bank” in Daur. When I was a child, I had many Daur, Korean, Mongolian, Manchu, and Evenki classmates and friends. Looking back, there seems to be a strange connection between this and my later research on ethnic history.

After I entered elementary school, I lived with my grandma and grandpa for a long time. My grandpa loved reading, and the big bookshelves at home were filled with classical novels, poems, and history books. I read through them when I had nothing to do. At that time, I was obsessed with books such as “Dream of Red Mansions”, Qing history, emperors of the past dynasties, and ancient poetry.

I also like to watch CCTV’s “Lectures of Hundreds of Scholars” and “Exploration and Discovery”, especially the documentary “New Silk Road”. The desolate, deep, open and tragic opening theme, combined with the images of the desert, Populus euphratica, Xiaohe Cemetery, camel caravans, Lop Nur, the ancient city of Loulan… the Silk Road immediately attracted me. The opening song is still on my playlist now.

I really got to know the Western Regions deeply after I went to college. In the 2014 college entrance examination, I ranked in the top 30 in liberal arts in Heilongjiang Province. When filling out the application form, I wanted to follow my own interests, and finally chose the Bachelor’s and Master’s program at Renmin University of China.

We didn’t have many professional classes in the first semester of our freshman year, so everyone joked that we were in a “nursing home.” The roommates let themselves go and went to various courtyards to listen. My interests are also quite diverse, and I have taken courses such as general chemistry, sociology, and news criticism. There are many courses on the Western Regions in the training program of the Chinese Academy of Sciences. My curiosity drove me to sit in on Professor Wuyun Bilige’s Manchu class, and I became more and more interested in it.

The Western Region is an important stage for Sino-Western exchanges and a key node on the Silk Road. When everyone mentions the Silk Road, they will naturally think of historical events such as the ancient Loulan Kingdom, Zhang Qian’s mission, Xuanzang’s westward journey, the Mongolian Western Expedition, and Zuo Zongtang’s recovery of Xinjiang. But when it comes to what the local situation was like from the end of the Yuan Dynasty to the time when Emperor Qianlong of the Qing Dynasty unified the Western Regions, many people have no idea. Many books on the history of the Western Regions mention this section briefly, as if these hundreds of years have disappeared out of thin air.

During that time, the Western Regions were closely related to the Genghis Khan family. The Eastern Chagatai Khanate and the Yarkand Khanate both evolved from the Chagatai Khanate established by Genghis Khan’s descendants. Later, the Yarkand Khanate was destroyed by the Junggars, and the Junggars were destroyed by the Qing Dynasty.

I always want to know how local people lived during those hundreds of years, how officials managed land and population, and how local government operated. However, there is a paucity of Chinese historical materials of that period, and other historical materials are also few and scattered. Most of the surviving documents are written in Persian and Chagatai script, and the research threshold is very high.

In the Western Regions, various civilizations collided and communicated with each other, and languages ​​were borrowed from each other. Often I recognize the words but don’t know what they mean when put together. When I interpret Chagatai documents, I always have four or five dictionaries on hand. If a word cannot be found in Chagatai, you have to look it up in an Arabic, Mongolian, or Uighur dictionary, because it is impossible to determine which language it is a loanword from.

On the eve of my doctoral graduation in April 2024, I set foot on the land of the Western Regions for the first time and came to Turpan to participate in the graduate student forum. In the corner of an exhibition hall in the Turpan Museum, I saw a document in Chagatai.

I was so excited when I recognized the first line of names. The publisher is the king of the Yarkand Khanate in Turpan and a descendant of Chagatai, the second son of Genghis Khan. The decrees he issued were called “yarligh” in the documents, which meant “imperial edict” in the Mongolian and Yuan Dynasties.

Such political decrees issued by the descendants of Chagatai have been found very rarely in Xinjiang, less than ten copies so far, and most of them are collected abroad. Even without reading the content, you can tell how precious it is.

After I returned to school, I compared the photos and spent more than a month interpreting the documents and writing a thesis. This is probably the fastest paper I have ever written, because I have a sense of urgency in my heart. For people in our field, it is really rare to encounter a piece of material that no one has studied.

This document contains very rich official and geographical information. The Tuyugou, Yanghai and Subashi mentioned in it are all very famous places in Turpan today. The documents also record the local official setup and land management system, providing rare materials for us to understand local social conditions and administrative operations.

In 2025, I went to Kyoto, Japan to visit Professor Kenichi Isogai, my mentor when I was studying abroad, and mentioned this document at the dinner table. I took out the photos on my phone and showed them to him. Unexpectedly, he didn’t even bother to eat. He looked at them over and over again for a long time and even helped me correct a few interpretation errors. I was touched that an experienced teacher could be so excited.

▲In August 2025, Qu Qiang checked the Persian manuscript at the Haneda Memorial Hall of Kyoto University. Photo provided by interviewee

After graduating with my Ph.D. in 2024, I came to Nankai University to teach. My topic “Compilation and Research of Multilingual Documents on Border Governance in the Western Regions of the Ming and Qing Dynasties” was approved by the National Social Science Fund for a special project for research on unpopular secrets.

I opened the elective course “History of the Northwest Frontier” in 2025. There were nearly 30 students in the classroom, more than I expected. Later, some students wrote in the course evaluation that they were very interested in those languages ​​and wanted to try to learn one.

The college’s Manchu language course is only available at the master’s level, but there are sophomores in my class who also want to learn it, so I recently opened a “Manchu reading class” after class, and 14 students signed up. I am very pleased that everyone is aware of the importance of learning language in historical research.

Some people call the research we do “sitting on the bench,” but I don’t like to sound like an ascetic. I loved those distant historical stories when I was a kid, and I still love them when I grow up. If one day, this knowledge is no longer so unpopular, I will only be happier.

People two thousand years ago also wanted to find answers to the unknown

Narrator: Zhang Ting, 31 years old, is a doctoral candidate at the School of Liberal Arts of Nanjing University. She participated in the compilation and research of Mawangdui silk books.

Today, some people read their zodiac signs and draw tarot cards before exams or interviews; some people look through the almanac when they have a wedding at home. More than two thousand years ago, the ancients would also use a set of calculation methods about dates, directions, and good or bad luck to try to find answers to unknown things.

I came into contact with these issues because I participated in the compilation of Mawangdui silk books. Silk books are books written on silk. In the 1970s, silk scripts with more than 120,000 characters were unearthed from Han Tomb No. 3 in Mawangdui, Changsha, Hunan. Due to the relatively airtight, anoxic and stable temperature and humidity environment, they can be preserved from the Western Han Dynasty more than 2,000 years ago to the present day, which can be called a “miracle among miracles”.

Chapter A of “Yin Yang and Five Elements” that I participated in the research records how the ancients chose days and judged good and bad times, including military affairs, marriage, migration, sacrifices, etc. It is one of the most damaged and difficult chapters in the silk books unearthed in Mawangdui: it was broken into 32 larger silk pieces and nearly 400 small fragments. Some of them have only one word or even the remains of a certain word. As you can imagine, in front of you is a puzzle whose original appearance is unknown and with many pieces missing.

The things it talks about are also very complex. Today we turn to the almanac and see the words “appropriate” and “taboo”, and all we get is a conclusion. This silk book also attempts to explain how the conclusion is calculated and the rules of its operation, so it is difficult to understand. The font style of this article is also relatively unique. The whole article is written in official script, which retains more of the physical structure of seal script. It also contains a large number of Chu characters. The text is ancient and difficult to interpret.

▲A partial picture of Chapter A of the silk book “Yin Yang and Five Elements” unearthed in Mawangdui. Photo provided by interviewee

The compilation of Mawangdui silk books was completed by successive generations of scholars. Mr. Qiu Xigui, a paleographist, has been involved in related research since the 1970s, and later edited the “Changsha Mawangdui Han Tomb Slips and Silk Collection”. Since then, the team has continued to revise based on newly discovered fragments and other research results. My mentor, Mr. Cheng Shaoxuan, was the direct organizer of Chapter A of “Yin Yang and Five Elements”. Under his guidance, I also made adjustments and improvements to the order of chapters, text interpretation, and patchwork of fragments.

In 2024, the revised version of “Collection of Slips and Silks from Mawangdui Han Tomb in Changsha” will be published. The overall appearance of Chapter A of “Yin Yang and Five Elements” has also been more completely restored through our efforts. However, there are still more than 200 small fragments that have not been put together. We have only moved the puzzle one step forward and are still far from completing the goal.

I like this kind of work because it’s like a detective investigating a crime. Whether the edges of the fragments can be connected, whether the creases, handwriting, and context can match up, we need to find evidence. Senior scholars often say: “We need to be knowledgeable about what we can say.” This does not mean that debate is not allowed, but that every judgment must withstand scrutiny. I like this feeling of finding clues and getting closer to the answer bit by bit.

I had no idea initially that I would be doing something like this in the future. When I was an undergraduate student, I studied Chinese at Southwest Jiaotong University. The school did not have teachers who specialized in studying ancient characters. I was just curious about glyphs and learned by self-study. Later, I was admitted to Fudan University for postgraduate study at the Center for the Study of Unearthed Documents and Ancient Characters.

There was no grand reason for choosing this major at first, I just thought Chinese characters were interesting. The early glyph for “dance” in the oracle bone inscriptions resembles a dancing person with open arms and holding ornaments such as feathers or animal tails; the character “peng” was originally shaped like a string of shells, which was an ancient currency unit; the character “bao” is the image of a person carrying a baby on his back; the character “book” is like a string of bamboo slips connected with a rope; and the character “dian” is like holding a book in both hands, or placing it on several tables.

▲The oracle bone inscriptions include Wu, Peng, Bao, Xu, and Dian. Picture source: “Oracle Bone Inscriptions” Photo provided by interviewee

Today, when someone is called “stupid”, it means stupidity, but in ancient times, it also meant turmoil and chaos. Teacher Jiang Yubin of Fudan University once deciphered an oracle bone script that had been difficult to decipher for a long time. Based on careful comparison of glyphs, he believed that the oracle bone character was “tun”, but ancient people sometimes used it to mean “stupid”.

For example, in ancient times, some turbulent and restless tribes in Fang State were called “stupid states”. Once a word is recognized, the originally unintelligible war records are explained. For this research, he received a 100,000-yuan Chinese Character Interpretation Award from the Museum of Chinese Characters.

The achievements of these predecessors made me understand the charm of ancient writing research. But rather than a grand narrative, I want to know the real life of the ancients. In those fragments, people are concerned about when to get married, have children, go out and build houses, petty officials want to choose a good day to meet their superiors, and generals want to predict good or bad luck before sending troops.

I am still studying “numerology” literature, which is a type of book used by the ancients for divination, date selection and calculation. Some of these concepts were later absorbed into almanacs, and similar habits of choosing dates can still be seen in today’s old almanacs. Times have changed, but people’s anxiety when facing the unknown and their desire to make the future more certain are the same from ancient times to the present.

I also feel more and more that history is not just about the emperors and generals in the history books, but also the bits and pieces of ordinary people’s lives. From the few words they left behind, I can “travel through thousands of years” and reconnect with ordinary people’s worries and expectations for life. Empathizing with them is what I think is the most romantic part of this knowledge.

There are some words that you won’t be able to hear if you don’t record them.

Narrator: Han Yumo, 30 years old, is a research intern at the Inner Mongolia Academy of Social Sciences, studying the Manchu and Tungus languages ​​(Manchu, Hezhe, Sibo, Oroqen and Ewenki).

The first time I saw the white Spring Festival couplets, I was also very surprised. The Han people post red Spring Festival couplets during the New Year and are also accustomed to using red for weddings. But in the early customs of the Manchu people, it was just the opposite. White was an auspicious color, and white was used for weddings and couplets.

The ancestors of the Manchu people have lived among the white mountains and black waters for a long time, where the winter is long. They wear white clothes when hunting, making it easier to hide themselves in the snow. The natural environment and fishing and hunting life give the color white a special meaning in their culture.

My master’s thesis research is on Manchu color words. At first I just wanted to understand why the same color has such different meanings in the lives of different peoples. After research, I discovered that behind a word, there is often a connection with how a nation lives and understands the world.

▲The Hezhen folk art troupe went to the fishing beach of Tongjiangkou to perform the Hezhe ethnic group’s Imakkan and Marry Lingkuo performances. Source: IC photo

I come from Chifeng, Inner Mongolia, and I studied Manchu in my bachelor’s, master’s and doctoral studies. When I filled out my application form for the college entrance examination, I just thought this major was quite special. Later, I studied with the teacher and gradually understood the culture behind the words, and then I really fell in love with this subject.

For example, “Niu Lu”(Manchu: niru), the original meaning of Manchu is “arrow; covered with arrows”, and later became the name of the grassroots organization in the Eight Banners system. When the Manchu ancestors hunted together, they used to work in groups of 10 people, which was called “Niu Lu”. Later, Nurhaci transformed this hunting organization into a grassroots unit for managing military and civilians, and once stipulated that 300 people would be included in each unit.

The ancestors of the Manchu people valued riding and archery, and arrows were a very important thing in their lives. A word related to “arrow” later extended the meaning of organization and leader. Following the changes in word meanings, we can see changes in lifestyles and social systems. In this way, a word entered the institutional system from hunting life, and the extension of its semantics is itself a miniature national history.

When I was studying for my Ph.D., I began to do linguistic research on the “Manchu-Tungus language family” with Professor Zhao Aping from the Institute of Manchu Studies at Heilongjiang University. We mainly study the five languages ​​​​in China: Manchu, Hezhen, Xibe, Oroqen and Evenki.

The situation for these languages ​​is not optimistic. Manchu Tungus language expert Chao Ke once mentioned in the “Comparison of Manchu and Tungus Language Vocabulary” published in 2014 that at that time, among the more than 10 million Manchus, there were less than 10 elderly people who could still speak some mother tongue; some middle-aged and elderly people living in Manchu villages could only speak simple mother tongue. Four other languages ​​are also endangered or critically endangered.

During my fieldwork, I found that parents often ask their children to learn more Chinese or a more widely spoken language in the local area because of practical considerations of further education and employment. For example, in some Ewenki families in Inner Mongolia, children in pastoral areas are also exposed to learning Mongolian from an early age. It’s not that many elders don’t cherish their mother tongue, it’s just that children have fewer and fewer opportunities to communicate in their mother tongue from an early age in their families and communities, so the intergenerational inheritance of language faces practical difficulties.

UNESCO’s World Language Map includes 8,324 languages. According to statistics, one language disappears every two weeks. What may be lost is not only a method of communication, but also the knowledge, memory and culture related to it.

The “Imakan” of the Hezhe people is an example. It is a folk art form that tells stories in the Hezhe language. When we were collecting records in Tongjiang, Heilongjiang, we met You Wenfeng, the provincial inheritor of Imakan. Wearing a fish-skin hat with the characteristics of the Hezhe people, she sang Imakam in high spirits. Yimakan does not require musical accompaniment, but only relies on the cadence of his voice and the rhythm of his body to bring the listener into the story. The Hezhe language has no traditional written language, and “Imakan” is like their “oral encyclopedia”.

▲Fish skin clothing made by the Hezhe people. Photo provided by interviewee

I believe that cultural confidence is also reflected in the cherishment of minority languages. These languages ​​carry the life experiences and cultural memories of different ethnic groups and are an integral part of the diversity of Chinese culture.

Teacher Zhao Aping has been doing field research for about 40 years. She often said that to study a national language, one must go among the nation. I agree with this sentence more and more now. When we go to collect and record now, we will record the words, sentences, folk songs and stories together. Through transcription and translation, we will create a searchable corpus and preserve the sounds and usage of these languages.

After graduating with my Ph.D. in 2025, I came to work at the Inner Mongolia Academy of Social Sciences and continued to study Manchu Tungus language and culture. I became more and more concerned about: Who will these laboriously collected materials be used for in the end? Some corpora are handed over as project results after they are completed, or are only kept in the hands of a few scholars.

But what the endangered language corpus needs most is sharing: if a language only remains in papers and hard drives, and people who want to learn it cannot hear it, that is not enough. Only by making the corpus retrievable, accessible, and reusable can the collected corpora truly serve the protection and inheritance of endangered languages. For example, we can cooperate with universities, local museums, intangible cultural heritage centers, etc., so that students can access and learn, so that researchers can easily access corpus, and so that children can find sounds and images when they want to understand the stories of their own nation, instead of just facing a difficult dictionary.

I want to teach AI to read text written three thousand years ago

Narrator: Li Yimeng, 21 years old, a senior student in the Chinese Language and Literature (Paleography) Strong Foundation Program of the School of Humanities at Huazhong University of Science and Technology, using AI to assist in deciphering oracle bone inscriptions.

At the end of 2022, when ChatGPT was released, I was still in high school. Large models of artificial intelligence are beginning to attract attention, and I am also curious: What would happen if such new technologies were used to study ancient texts?

I have loved reading historical and cultural books and programs since I was a child. When I read “Lecture Forum of Hundreds of Scholars”, I learned for the first time that people today can still try to decipher the characters carved on the bones of tortoise shells more than 3,000 years ago, and there are still many characters that have not been understood.

When I was a child, the school held a cultural activity related to “horse”. I discovered that there are many horse-related glyphs in oracle bone inscriptions. The ancients would use different glyphs to record the color and use of horses. After I went to college, I went to the Yin Xu Museum, where I saw 22 kinds of oracle bone characters related to horses, and I recalled my childhood curiosity.

▲Different oracle bone inscriptions related to the character “horse” displayed in the Yinxu Museum. Photo provided by interviewee

There are only 5 students in our strong basic class this year. The school has assigned tutors to each of them, and also offers separate courses such as the introduction to “Shuowen Jiezi” and unearthed documents.

When I was a freshman, I joined the computational linguistics team of the college, mainly doing basic work such as corpus collection. The model needs a large amount of reliable data to answer ancient Chinese questions. If you want the AI ​​to answer accurately, someone must first sort and check the data.

In 2024, I joined the interdisciplinary student team of “Bu Suanzi” led by Professor Bai Xiang of the School of Software and Professor Liu Yuliang of the School of Artificial Intelligence and Automation. The team hopes to make a large model that uses AI to assist in deciphering oracle bones. I am the first ancient writing major in the team.

In order to teach AI to recognize oracle bone inscriptions, we need to sort out a large number of oracle bone rubbings and glyph materials. The program can be used for batch screening. I want to check whether there are any mistakes acknowledged or marked, and check how a word is written in different materials. I will also consult the research of predecessors such as Qiu Xigui and Li Xueqin, and tell AI the literature reviews of predecessors so that the answers given by the model can be based on evidence.

Does this method work? A model training at the end of 2024 gave us a specific example. We gave a large number of rubbings to the model for screening, and specifically tested the large model to pick out difficult words. It picked out a controversial character from the oracle bone rubbing No. 307 of “Tuennan”, and the candidate answers given included “lao” and “pawn”.

The same character in oracle bone inscriptions is often written in different ways, and different characters may also look similar. I spent a month searching the literature, combined with the glyphs and the context of the oracles, and found that the word “Lao” has more basis as a controversial word. “Lao” here does not mean labor, but may be related to the place name or national name at that time.

Later, we invited Professor Song Zhenhao, an oracle expert from the Chinese Academy of Social Sciences, to participate in the discussion. He agreed with our judgment, and later the authoritative domestic Oracle database “Yin Qi Wenyuan” re-labeled this character and included it in the range of variants of “Lao”.

▲Li Yimeng’s process of researching and proving that this oracle bone inscription is the word “Lao”. Photo provided by interviewee

Initially, the team mainly started with glyphs: letting the model learn the changes in oracle bone inscriptions, bronze inscriptions, and seal scripts to today’s Chinese characters, and then working backwards to figure out what an unrecognized oracle bone inscription might be.

In 2025, we went to 9 places, including Anyang, Zhengzhou, Shanghai, Nanjing, and Chongqing, and consulted 21 experts in the fields of paleography, phonology, exegesis, etc. After the experts reminded us, we found that paleography experts do not just look at what they look like, but make comprehensive judgments. Therefore, we have perfected a set of ideas of “sound, shape, meaning, and reason”: from pronunciation, glyphs, meanings to reasoning basis, we can determine how a word should be interpreted.

In the past two years, the team has continued to improve the system, and the model has been iterated 56 times. Initially, if you enter a word, it can only give you one answer; now it can provide several candidate words, related explanations, and locate literature discussing the word. The “Bu Suanzi” project we participated in also won the special prize in the “Challenge Cup” National College Student Extracurricular Academic Science and Technology Works Competition.

However, AI cannot replace paleographers in making the final judgment. The interpretation of a word requires mutual confirmation of the glyph, pronunciation, literature and historical background, and often requires long-term discussions in the academic community. We hope to use AI to help researchers find clues faster and leave more time for truly important judgments.

▲Oracle AI large model page produced by the “Bu Suanzi” team of Huazhong University of Science and Technology. Photo provided by interviewee

About Us · 關於我們