 |
| Chinese Gazzeters |
What 地方志 can tell us, what they cannot, and why contemporary evidence matters
Chinese martial arts history presents an unusual problem.
Many of the people who supposedly shaped famous martial arts lived during the eighteenth and nineteenth centuries. Yet much of what is commonly known about them was written down decades—or sometimes more than a century—after they supposedly lived.
A lineage may preserve the name of a teacher. A student may remember what his teacher told him about an earlier generation. A family may preserve a story about an ancestor. A martial arts association may publish a history of its style.
All of these sources can be valuable.
But they present a basic historical problem:
How can we determine which parts of these stories describe historical events and which parts developed during later transmission?
One answer is to search for evidence produced closer to the period being investigated.
For Chinese local history, one of the richest bodies of such evidence is the 地方志 (difangzhi)—usually translated into English as local gazetteers.
They have become increasingly important in my research into Wing Chun, White Crane and the history of Chinese martial arts. This article explains what these remarkable books are, why historians use them, and why their evidence must still be approached critically.
What is a Chinese local gazetteer?
The English word gazetteer can be misleading.
To a modern English reader, a gazetteer might sound like little more than a geographical directory—a list of towns, mountains, rivers and administrative divisions.
Chinese 地方志 were considerably more than that.
A local gazetteer was an organized record of a particular geographical or administrative area. Depending upon the period and the work, that might mean a province, prefecture, department, county, town or other locality.
Joseph Dennis, whose major study Writing, Publishing, and Reading Local Gazetteers in Imperial China, 1100–1700 examines their production and use, describes them as cumulative records of territorial units, normally organized according to subjects such as geography, institutions, population, taxation, biographies and literature.
The Max Planck Institute for the History of Science similarly describes 地方志 as major primary sources for Chinese local history. More than 8,000 titles dating from the tenth through twentieth centuries survive, covering most populated areas of historical China. Their contents extend far beyond geography to temples, schools, officials, prominent individuals, local customs, plants, animals, products and many other aspects of local life.
A gazetteer can therefore sometimes resemble an extraordinary combination of:
a geographical survey, administrative handbook, census, local history, biographical dictionary, institutional directory, collection of inscriptions, literary anthology and record of local customs.
For historians interested in people who never became important enough to enter the great histories of China, that local perspective can be enormously valuable.
Why did Chinese governments produce them?
This is essential to understanding their evidentiary value.
Gazetteers were not originally created for twenty-first-century historians.
They had practical functions.
Li Zonghan (李宗翰), in his study 〈清代地方志的知識性質——以光緒《金門志》為例〉 (The Nature of Knowledge in Qing Local Gazetteers: The Guangxu Kinmen Gazetteer as an Example), emphasizes that local gazetteers traditionally possessed the character of official books, 官書.
They helped governments understand and administer the places under their authority. This gave them a degree of official authority, although, as we will see, it certainly did not make them neutral or infallible.
Their historical development was long.
Li notes that from the 圖經 (tujing) tradition of the Sui and Tang periods, the state and local officials systematically collected information about local areas in order to govern them more effectively. From the Song through the Qing, the compilation of most local gazetteers was initiated by government officials. Consequently, they contained precisely the sorts of information governments needed: population, taxation, military organization and other administrative matters.
The genre became particularly important from the Song onward and flourished during the Ming and Qing periods. Modern scholarship treats the gazetteer as one of the fundamental bodies of evidence for studying premodern Chinese local society.
By the Ming and Qing periods, a hierarchy of gazetteers roughly reflected the hierarchy of administration itself:
通志 — provincial gazetteer
府志 — prefectural gazetteer
州志 — departmental/state gazetteer
廳志 — subprefectural gazetteer
縣志 — county gazetteer
Below these levels there could also be locally produced town and village histories.
Li points out that these smaller works were not necessarily official. Many surviving Ming and Qing town gazetteers were privately compiled by local elites rather than reviewed by government authorities.
That distinction matters enormously.
Not everything called a 方志 was produced in exactly the same way or carried the same institutional authority.
What information did they contain?
There was no single table of contents followed by every Chinese gazetteer.
Nevertheless, recurring categories appear again and again.
Depending upon the period and locality, we can encounter sections dealing with geography, mountains and waterways, administrative boundaries, population, taxation, government offices, military defenses, temples, monasteries, schools, academies, examination graduates, officials, prominent families, local industries, commerce, medicine, notable inhabitants, customs, disasters, inscriptions, literature and biographies.
Modern gazetteers can be even more detailed.
They may contain sections on police organizations, schools, hospitals, associations, sports, newspapers, political organizations, local industries and—in some cases—武術, martial arts.
This is why gazetteers are potentially so important to martial arts historians.
A nineteenth-century martial arts teacher might never have attracted the attention of an imperial historian in Beijing.
But if he operated a medical clinic, trained local militia, held a military position, belonged to a guild, became locally prominent, participated in community affairs, maintained a school, treated injuries or was remembered as an unusual local personality, he might leave a trace in a local record.
That trace may be only a few lines.
Sometimes those few lines are more historically valuable than twenty pages written about him a century later.
Gazetteers as historical databases
There is a useful modern way to think about these enormous books.
Researchers working with digitized gazetteers have described the genre as something approaching a historical database.
A recent digital-humanities study notes that individual gazetteer compilations can be understood as data-collection projects. Editors gathered information about a locality and organized it into recurring categories. Successive editions often updated that information. Because thousands of gazetteers cover enormous geographical and chronological ranges while retaining broadly comparable structures, researchers can investigate them collectively rather than simply reading one book at a time.
This is particularly useful for martial arts research.
Suppose we are investigating a person who supposedly lived in Foshan around 1870.
We do not merely type his name into Google.
We can ask:
Was he mentioned in the county gazetteer?
What about the prefectural gazetteer?
Is he included among physicians?
Merchants?
Military personnel?
Artists or specialists?
Local personalities?
What about the street where he supposedly lived?
Does his alleged business appear?
Do his relatives appear?
Are people of comparable occupation and social status recorded?
Was the information copied from an earlier edition?
Does a later edition suddenly introduce him?
These are much more productive historical questions than simply asking:
“Was the story true?”
Why are gazetteers primary sources?
Here we need an important distinction.
A gazetteer is not a primary source merely because it is old.
Its evidentiary character depends upon the particular information we are using.
Imagine a gazetteer published in 1926.
If it reports the establishment of a local association in 1923 from information gathered during its contemporary compilation, that passage can function as near-contemporary or contemporary primary evidence for that association.
But suppose the same 1926 gazetteer recounts something supposedly occurring in 1650.
For that seventeenth-century event, the 1926 book is obviously not contemporary evidence merely because the physical book itself is now old.
We must determine where the compilers obtained the information.
Perhaps they copied an eighteenth-century gazetteer.
Perhaps they reproduced an inscription.
Perhaps they used a genealogy.
Perhaps they copied a government document.
Perhaps they recorded a local tradition.
Each creates a different evidentiary situation.
This is why historians must investigate the source inside the source.
The correct question is not:
“How old is this book?”
It is:
“How close is this particular information to the event it claims to describe, and where did that information come from?”
Compilation date is not necessarily publication date
Li Zonghan's study of the Guangxu 《金門志》 (Kinmen Gazetteer) provides an excellent example.
The book was ultimately printed in 1882.
But its history began much earlier.
林焜熿 began the project around 1833–1835. His son 林豪 subsequently continued and expanded the work. Officials and other editors became involved at different stages.
The manuscript was reviewed and altered.
At one stage, an editor reportedly removed approximately 30 percent of the earlier manuscript while preparing another attempted publication.
Therefore, saying simply:
“The Kinmen Gazetteer dates from 1882.”
tells us remarkably little about when a particular piece of information entered the text.
For serious research we want to know:
Who collected it?
When?
From whom?
Was it present in the earlier manuscript?
Was it added later?
Was it copied from another document?
Was it removed or rewritten during editing?
This principle becomes extremely important when martial arts traditions claim that a document “proves” something.
The existence of an old book is only the beginning of the investigation.
Official does not mean infallible
There is a dangerous temptation when researchers first discover gazetteers.
After spending years dealing with legends and oral traditions, an official nineteenth-century document can feel wonderfully solid.
But we must resist replacing one form of unquestioning belief with another.
Gazetteers can be wrong.
They can contain errors.
They can repeat earlier errors.
They can omit people.
They can reproduce legends.
They can reflect political priorities.
They can privilege elite families.
They can classify people according to contemporary moral and social expectations.
Their editors could have personal interests.
Local families could influence what was recorded.
Officials could influence what was considered appropriate.
Li's study is particularly valuable precisely because he examines this tension.
Gazetteers possessed official authority, but their compilation frequently depended upon local elites. Consequently, they could simultaneously reflect state interests and local interests.
In the case of 《金門志》, Li shows how local elite interests affected the presentation and classification of prominent people.
He even identifies an instance in which a newly successful examination candidate's name entered the gazetteer shortly before publication—an excellent reminder that these books were produced by real people living inside local networks of prestige and power.
So our principle is never:
The gazetteer says it, therefore it happened.
It is:
The gazetteer says it. Now we know that this claim existed in this textual context, and we can investigate where it came from and whether other evidence supports it.
Contemporary evidence and later oral history are not equivalent
This brings us directly to martial arts history.
Consider two hypothetical pieces of evidence.
The first is a local document written in 1875 describing a martial instructor active in the town in 1873.
The second is an interview conducted in 1975 in which a practitioner recounts what his teacher told him about a martial instructor supposedly active in 1873.
Either account could be correct.
Either account could be wrong.
But they do not present the historian with the same evidentiary problem.
The 1875 account was produced two years after the alleged event.
The 1975 account was produced 102 years later and may contain several stages of transmission:
1873 event → witness → student → later student → interviewee → interviewer → published text.
At every stage, information can be preserved.
But it can also be forgotten, simplified, misunderstood, combined with another story, embellished, reinterpreted or deliberately changed.
This does not make oral history worthless.
Quite the opposite.
Oral histories can preserve names, relationships, terminology and events absent from surviving written records. They can lead us to photographs, documents, graves, family records, buildings and other evidence we would otherwise never discover.
But oral tradition and contemporary documentation answer different historical questions.
A late oral account certainly demonstrates that a tradition was being transmitted when the account was recorded.
Whether it accurately describes events a century earlier is a separate question requiring corroboration.
The problem of lineage transmission
Martial arts history makes this particularly difficult because information is often transmitted through lineages.
A student naturally trusts his teacher.
The teacher trusted his teacher.
Eventually a story can acquire enormous authority simply because generations of respected practitioners repeated it.
But repetition is not independent corroboration.
If ten books published between 1980 and 2020 repeat the same story derived from one 1960 interview, we do not possess ten independent sources.
We possess one source reproduced ten times.
This is why dating the earliest identifiable appearance of a claim is so important.
Instead of asking:
“How many people tell this story?”
we ask:
“What is the earliest source in which we can actually demonstrate that this story existed?”
Then we work backward.
Contemporary does not automatically mean truthful
The opposite mistake must also be avoided.
A contemporary document can lie.
A government official can exaggerate his accomplishments.
A family can inflate its prestige.
A newspaper can publish misinformation.
A tombstone can exaggerate.
An association can rewrite its own history.
An inscription can be political propaganda.
A gazetteer compiler can favor one family over another.
Being contemporary reduces one important problem—chronological distance—but it does not eliminate bias, deception or error.
Therefore our preferred method is not simply to locate one contemporary source.
It is to find independent contemporary or near-contemporary sources that intersect.
A gazetteer might identify a person.
A newspaper might place him in the same town.
A genealogy might establish his family.
A government record might establish his occupation.
An inscription might confirm participation in a local institution.
A grave might provide dates.
A photograph might establish association with particular people.
None necessarily proves the entire later story.
But together they can gradually reconstruct a historical person.
Corroboration: when different sources meet
Li's 《金門志》 study demonstrates this principle beautifully.
He does not simply extract information from the gazetteer and assume it is correct.
He compares the gazetteer's picture of nineteenth-century Kinmen elites with surviving stone inscriptions. The inscriptions independently support important aspects of the social pattern reconstructed from the gazetteer.
That increases our confidence.
This is precisely how martial arts history should be approached.
Imagine finding a martial artist in a gazetteer.
Good.
Then find him in a newspaper.
Better.
Find his medical shop in a commercial directory.
Better still.
Find his family in a genealogy.
Find an association record bearing his name.
Find an inscription mentioning him.
Now we are no longer dealing with a single lineage narrative.
We are reconstructing a person through independent documentary intersections.
What does absence from a gazetteer mean?
This is one of the most difficult questions.
Suppose a famous martial arts tradition describes a nineteenth-century master as an extremely prominent local figure.
We search the contemporary local gazetteer.
He isn't there.
Have we proved that he never existed?
No.
No gazetteer recorded everybody.
Its categories were selective. Compilers made decisions. Records disappeared. People were overlooked. Social status mattered. Some professions were considered more worthy of documentation than others.
Therefore:
Absence is not proof of nonexistence.
But absence is not always meaningless either.
Imagine that the same gazetteer contains biographies of:
physicians, merchants, martial practitioners, military men, charitable figures, unusual local personalities and people operating precisely the sort of establishments our missing individual supposedly operated.
Now his absence becomes more interesting.
Suppose he is absent from several successive gazetteers.
More interesting.
Suppose his supposed business is absent from commercial records.
His family cannot be located.
No contemporary newspaper mentions him.
No inscription mentions him.
No genealogy can be located.
And the first detailed biography appears seventy years after his supposed death.
We still have not proved that he did not exist.
But the historical problem has changed considerably.
This is why negative evidence must be treated cumulatively and contextually rather than absolutely.
Successive gazetteers can show when a tradition appears
One of the most powerful techniques is comparing different editions of a gazetteer from the same locality.
Imagine that we possess editions from:
1692 → 1757 → 1787 → 1930
A famous local founder is supposedly active around 1700.
The 1692 edition says nothing.
The 1757 edition says nothing.
The 1787 edition says nothing.
The 1930 edition suddenly contains a detailed biography.
That does not prove the biography was invented in 1930.
Perhaps the compilers discovered an older document that no longer survives.
Perhaps a genealogy became available.
Perhaps an earlier manuscript was overlooked.
Perhaps local oral tradition preserved genuine information.
But now we have identified a very important historical problem:
Why does this person enter the surviving gazetteer tradition at this particular point?
We can then investigate the source used by the 1930 compilers.
That is far more useful than simply arguing whether the founder was “real.”
Gazetteers preserve local information that central histories may not
This is perhaps their greatest value for martial arts research.
Most traditional martial artists were not emperors, generals or nationally significant officials.
They were local people.
Some were soldiers.
Some ran businesses.
Some practised medicine.
Some worked in transport.
Some belonged to opera companies.
Some taught privately.
Some participated in local associations.
Some were members of powerful families.
Some were nearly invisible outside their neighborhood.
Central imperial archives had little reason to record many of them.
Local gazetteers did.
Modern research on gazetteers specifically emphasizes that their regional information frequently preserves material absent from central-government archives.
That makes them particularly suitable for reconstructing the social environments in which Chinese martial arts developed.
We may not find:
“Master X founded Style Y.”
Instead we may discover something historically more valuable:
Master X actually existed.
He lived at this location.
He practised medicine.
He belonged to this family.
He served in this organization.
He knew these people.
He was described using this contemporary terminology.
From those small pieces, a real historical world begins to emerge.
Chinese Local Gazetteers: Primary Sources for Martial Arts History
This is another important application.
Modern martial arts names are often projected backward.
A style currently known as 詠春拳, 白鶴拳, or another standardized name may not have been described by exactly that terminology in an earlier period.
Therefore we should not search only for modern style names.
We also search period vocabulary:
拳 — boxing/fist
拳術 — boxing arts
武術 — martial arts
武藝 — martial skills
技擊 — fighting/combat skill
武師 — martial instructor/expert
教習 — instructor
棍 — staff
刀 — saber/knife
槍 — spear
武勇 — martial valor
武略 / 武績 — military accomplishment
Terminology itself has a history.
Finding when a particular name first appears can therefore be as important as finding a particular person.
Why one gazetteer is never enough
Li explicitly warns against generalizing from a single case.
Thousands of Chinese gazetteers were produced under different political, social and institutional conditions. What is true about one compilation cannot automatically be assumed about another.
Before using a gazetteer seriously, we therefore ask:
Who commissioned it?
Who compiled it?
When was the information collected?
When was it published?
Was it official or privately compiled?
What earlier editions existed?
What sources did the compilers use?
What categories did they include?
Whose biographies were considered worthy of inclusion?
What political or local interests might have shaped it?
Was a later edition copied from an earlier one?
Can important statements be checked against other evidence?
In other words, the gazetteer itself must become an object of historical investigation.
Only then can we properly use the information inside it.
A primary source is evidence, not a verdict
This may be the most important point of the entire article.
Historians do not use primary sources because primary sources always tell the truth.
They use them because primary sources bring us closer to the historical process we are attempting to reconstruct.
A police report may contain bias.
A newspaper may contain an error.
A genealogy may exaggerate.
A gazetteer may omit someone.
An inscription may praise its donor excessively.
An eyewitness may remember incorrectly.
Nevertheless, these sources allow us to establish what particular people or institutions recorded at particular times.
History emerges through comparison.
When independent sources converge, confidence increases.
When they contradict one another, the contradiction itself becomes evidence requiring explanation.
When only a much later tradition survives, we say exactly that.
There is nothing wrong with saying:
We do not know.
That is preferable to turning a possibility into a fact.
Tradition is also historical evidence
None of this means that martial arts traditions should be discarded.
A legend is evidence.
A lineage genealogy is evidence.
An oral history is evidence.
A master's recollection is evidence.
The question is:
Evidence of what?
A story recorded in 1980 is unquestionably evidence that the story existed in or before 1980.
It is not automatically evidence that the event described actually occurred in 1780.
Those are two different propositions.
And sometimes investigating that difference reveals something fascinating.
We may discover how a martial community understood itself, how its identity developed, how different lineages competed, how political circumstances changed a story, or how one person's biography gradually became a founder legend.
The history of a tradition is itself history.
Why Martial Custodian uses gazetteers
The purpose of Martial Custodian is not to prove martial arts legends false.
Nor is it to prove them true.
The objective is simpler:
to find out what the surviving evidence allows us to say.
For that reason, Chinese local gazetteers have become an important part of our research.
They allow us to move beyond lineage histories and search the wider society in which martial artists actually lived.
Instead of looking only for a famous master's name, we can reconstruct his environment:
his town, family, occupation, institutions, military organization, medical community, commercial world, temples, associations, contemporaries and terminology.
Sometimes the traditional story will survive that examination remarkably well.
Sometimes only part of it will.
Sometimes the documentary evidence will contradict it.
And sometimes, after searching everything available, the answer will remain:
We don't know.
All four outcomes are legitimate historical conclusions.
Contemporary evidence before retrospective certainty
When reconstructing an event, we generally give greater evidentiary weight to evidence produced close to that event than to accounts recorded generations later.
Not because old documents are sacred.
Not because oral tradition is worthless.
And certainly not because officials never lied.
The reason is methodological.
Every additional stage separating an event from its surviving account creates another opportunity for information to be altered through memory, interpretation, selection, institutional interests or storytelling.
So we begin as close to the event as surviving evidence permits.
Then we work outward.
Contemporary documents.
Near-contemporary documents.
Inscriptions.
Newspapers.
Official records.
Gazetteers.
Genealogies.
Association records.
Personal writings.
Later recollections.
Oral histories.
Lineage traditions.
The exact evidentiary value of each must be evaluated individually. This is not a mechanical ranking system.
A well-documented oral testimony can correct an official record.
A genealogy can preserve information absent from a gazetteer.
A newspaper can be completely wrong.
The principle is simply that claims should be traced toward their earliest demonstrable sources whenever possible.
From legend back to history
Chinese martial arts have extraordinarily rich traditions.
They deserve something better than blind belief.
They also deserve something better than blind debunking.
Local gazetteers give us one way to move beyond both.
They allow us to place martial artists back into the towns, families, institutions and historical circumstances in which they supposedly lived.
And when combined with newspapers, inscriptions, genealogies, government records, photographs, association documents, manuscripts and carefully collected oral histories, they can sometimes transform a legendary name into a historically recoverable human being.
That is ultimately why we use them.
The purpose of historical research is not to defend the story we inherited.
It is to discover how much of that story the surviving evidence allows us to understand.
A note on sources and further reading
The central methodological inspiration for this article is Li Zonghan's 〈清代地方志的知識性質——以光緒《金門志》為例〉, published in 漢學研究 33.3 (2015), pp. 241–274. Li approaches gazetteers not merely as containers of historical facts but as products of interaction between government authority and local elites.
For the broader history and production of the genre, Joseph R. Dennis's Writing, Publishing, and Reading Local Gazetteers in Imperial China, 1100–1700 is particularly important. Modern work by the Max Planck Institute for the History of Science also demonstrates the extraordinary potential of treating thousands of surviving gazetteers collectively as a historical research infrastructure.
Li's own bibliography provides another useful path into Chinese and international scholarship, including work by James Hargett, Timothy Brook, Peter Bol, 林開世, 倉修良, 黃葦, 巴兆祥 and 劉緯毅.
For Taiwan specifically, there is already a substantial scholarly literature on local gazetteers, including work on Qing Taiwan gazetteers, Japanese-period official local histories, and the historiographical value of different Taiwan gazetteer traditions.