The Book Behind the Book
Sima Guang is famous for the Zizhi Tongjian. Few people know that he also wrote a second book that was essential to the first. The Kaoyi (考异, “Examining Differences”) was Sima Guang’s companion volume to the Tongjian, documenting every case where his sources disagreed and explaining which account he chose and why. It was one of the earliest examples of systematic textual criticism in Chinese historiography, and it reveals the research process behind the polished narrative.
In addition to the Kaoyi, Sima Guang and his team maintained working documents called the congmu (丛目), or “clustered entries,” which were raw chronological notes organized by year and event. The congmu served as the bridge between the original sources and the final narrative. Sima Guang would take the clustered notes, resolve contradictions using the methods documented in the Kaoyi, and draft a continuous narrative. This three-stage process, from sources to congmu to final text, was remarkably systematic for its time.
What the Kaoyi Contains
The Kaoyi survives as a separate volume of roughly 300 entries, each addressing a specific textual problem. The typical entry follows a pattern: Sima Guang states the event in question, lists the different accounts from different sources, notes where they conflict, and explains his decision. The conflicts range from minor date discrepancies to fundamentally different versions of the same event.
For example, when dealing with the assassination of a Tang dynasty minister, Sima Guang found that the official biography recorded the event in one year, a court diary recorded it in a different year, and a private letter mentioned it in yet another year. He cross-referenced the dates with other events mentioned in each account, identified which dating system each source used, and concluded that the court diary was correct because its relative dates matched other independently confirmed events.
The Kaoyi is where you see Sima Guang not as a storyteller but as a detective, weighing evidence and building arguments. The Tongjian presents the conclusion. The Kaoyi shows the work.
Types of Textual Problems Sima Guang Addressed
The problems documented in the Kaoyi fall into several categories. Date conflicts are the most common, because different sources used different calendar systems, reign names, and methods of counting years. Sima Guang standardized everything to the Western Han calendar and resolved conflicts by cross-referencing.
Attribution errors are another frequent category. Biographies sometimes attributed actions to the wrong person because of name confusion, shared titles, or deliberate historical revisionism. Sima Guang traced the transmission of stories through multiple texts to identify where errors were introduced.
Political bias was a subtler problem. Official histories were often written by historians of the succeeding dynasty, who had incentives to portray the previous dynasty’s final years as corrupt and illegitimate. Sima Guang compared official accounts with unofficial sources, private letters, and inscriptions to detect distortions. Learn more about this process in the material collection process.
Types of Textual Problems in the Kaoyi
| Problem Type | Frequency | Resolution Method |
|---|---|---|
| Date conflicts | ~120 entries | Cross-reference with independently dated events |
| Name attribution | ~60 entries | Trace through multiple sources, compare biographies |
| Political bias | ~40 entries | Compare official vs unofficial sources |
| Geographic confusion | ~30 entries | Cross-reference with administrative geography |
| Chronological table errors | ~50 entries | Reconcile tables with narrative text |
The Congmu: Working Notes as Historical Artifact
The congmu were Sima Guang’s working notes, and they no longer survive as a complete document. We know about them from references in the Kaoyi and from Sima Guang’s own descriptions of his methodology. The congmu were essentially a massive chronological database of raw historical material, organized by year, with each event documented through multiple source excerpts.
Each assistant was responsible for compiling congmu for their assigned period. Liu Shu compiled notes on the Five Dynasties. Liu Ban worked on the Han period. Fan Zuyu handled the Tang dynasty. Sima Guang would then take these raw notes and draft the narrative, cutting redundancies, resolving contradictions using the methods documented in the Kaoyi, and adding his commentary. This division of labor allowed the team to cover 1,300 years of history in roughly twenty years of work. See how the team operated in the assistant scholars overview.
A quick note: The loss of the congmu is a significant gap in our understanding of the Tongjian‘s creation. Without the working notes, we cannot fully reconstruct Sima Guang’s decision-making process. The Kaoyi is our only window into the research behind the final text, and even it only documents the cases where sources disagreed, not the vast majority of passages where sources were consistent.
The Legacy of Sima Guang’s Textual Criticism
The Kaoyi established a standard for textual criticism in Chinese historiography that influenced all subsequent official histories. Later historians adopted the practice of documenting source disagreements, though few did it as systematically as Sima Guang. The Qing dynasty scholar Qian Daxin (钱大昕, 1728-1804) praised the Kaoyi as the most rigorous piece of textual criticism in pre-modern Chinese historiography, and modern historians continue to use it as a primary source for understanding the transmission of early historical texts.
The Kaoyi also had an indirect influence on the development of evidential scholarship (考证学, kaozheng xue) in the Qing dynasty. The methodological principle that Sima Guang demonstrated, that claims should be evaluated against multiple sources and that contradictions should be documented rather than hidden, became a cornerstone of the Qing scholarly movement. For connections to other foundational texts, visit the Tongjian pillar and the Shiji comparison.
| Companion Work | Purpose | Survival Status |
|---|---|---|
| Kaoyi (考异) | Document source disagreements and resolutions | Survives, ~300 entries |
| Congmu (丛目) | Raw chronological notes from sources | Lost |
| Fanli (凡例) | Editorial conventions and rules | Survives, partially |
| Mulu (目录) | Index and chapter guide | Survives, modified over time |
Why the Kaoyi Matters to Readers Today
Most readers of the Tongjian never look at the Kaoyi. The companion volume is technical, dense, and assumes familiarity with the sources Sima Guang consulted. But for anyone interested in how history is actually written, not just what it says, the Kaoyi is invaluable. It shows that the smooth, authoritative narrative of the Tongjian is built on a foundation of conflicting, messy, sometimes contradictory sources. The narrative hides the seams. The Kaoyi exposes them.
This transparency about methodology is modern in spirit. Modern historians are expected to document their sources, explain their decisions, and acknowledge uncertainties. Sima Guang did all of these things in the eleventh century. He may have been the first historian anywhere to practice systematic textual criticism on this scale. The fact that he did it within a project designed for imperial education makes it more remarkable, not less. He understood that a history without documented sources is just a story, and he wanted the emperor to know the difference.
For readers who want to explore the broader culture of historical accuracy in Chinese thought, the connection to the Five Arts tradition shows how textual criticism and metaphysical study intersected in imperial China.
Frequently Asked Questions
Is the Kaoyi available in English?
Not as a complete translation. Selected entries have been translated in academic articles, particularly those dealing with the Tang dynasty, where the Kaoyi is most detailed. For most English readers, the best access is through secondary scholarship that cites and discusses specific Kaoyi entries.
How does the Kaoyi compare to Western textual criticism?
The Kaoyi predates the systematic Western tradition of textual criticism by several centuries. The methods are similar in spirit: comparing sources, cross-referencing dates, identifying bias. The main difference is that Sima Guang worked within a tradition that treated historical writing as a moral enterprise, not just a scholarly one, which meant his source criticism was always connected to questions of moral judgment.
Did other Chinese historians adopt Sima Guang’s approach?
Yes, though with varying degrees of rigor. The official histories of the Song, Liao, Jin, and Yuan dynasties all include companion volumes documenting source disagreements, directly modeled on the Kaoyi. The practice became a standard feature of Chinese historiography, though few historians matched Sima Guang’s systematic thoroughness. For the broader tradition, see the Tongjian vs Shiji comparison.
The Scale of the Research Operation
Understanding the Kaoyi requires understanding the scale of the research operation behind the Tongjian. Sima Guang and his team consulted over three hundred sources. These included the seventeen official dynastic histories that existed at the time, covering everything from the Shiji to the Old Tang History. They also drew on court diaries, private letters, family genealogies, local gazetteers, inscription collections, Buddhist and Daoist texts, and previously compiled chronologies. The total volume of source material was staggering, perhaps ten times the final text.
The working library that Sima Guang maintained in his Luoyang residence was described by visitors as a room overflowing with bundled notes and excerpts. Each source was excerpted into the congmu system, with the relevant passage, its source name, and its date noted on a separate strip or sheet. Over nineteen years, the accumulated raw notes filled a substantial room. Sima Guang reportedly told his assistants that the notes alone, if stacked, would reach the ceiling. This is the invisible mountain of material that the Kaoyi hints at and that the final Tongjian distills into narrative.
The physical demands of this project were enormous. Sima Guang worked every day, including during his years as chancellor. His eyesight deteriorated. He developed calluses on his fingers from writing. He revised the text multiple times, and there are records of him rewriting a single paragraph dozens of times until the wording satisfied him. This level of dedication is what separated the Tongjian from the many other chronological histories that had been attempted before it and failed.
The Fanli: Editorial Conventions
Another companion document that partially survives is the Fanli (凡例), or “Editorial Conventions.” This was Sima Guang’s style guide for the Tongjian, specifying how different types of events should be recorded, what terminology to use, how to format dates, and when to include commentary. It ensured consistency across 294 chapters written with the help of three different assistants whose prose styles naturally diverged.
The Fanli covered details that seem trivial but mattered deeply to Sima Guang. How should a ruler’s death be recorded? What verb should describe a military victory versus a military defeat? Should an official’s full name be used every time, or can a shortened form appear after the first mention? These were not just stylistic preferences. They were political signals. The choice of verb for a military action could imply legitimacy or illegitimacy. The context of how Sima Guang framed rulers is explored in the lessons for rulers article.
| Fanli Rule Category | Example | Why It Mattered |
|---|---|---|
| Verbs for military actions | “Conquered” vs “invaded” vs “punished” | Signaled legitimacy of the action |
| Emperor titling | Full temple name vs posthumous name | Indicated editorial judgment of the reign |
| Rebellion terminology | “Rebelled” vs “rose up” vs “resisted” | Determined moral framing of the event |
| Commentary placement | End of event vs end of chapter | Controlled reader interpretation |
What Was Lost When the Congmu Disappeared
The loss of the complete congmu is one of the great gaps in Chinese historiographic record. With the working notes, we could see exactly which passages Sima Guang took from which sources, how he modified them, and what he chose to omit. Without them, we rely on the Kaoyi for the cases where sources disagreed, and on textual comparison for the cases where the Tongjian follows a single source closely.
Modern scholars have attempted to reconstruct parts of the congmu by comparing the Tongjian text with surviving source texts. When a passage in the Tongjian closely matches a passage in the Old Tang History, for example, we can infer that Sima Guang used that source for that section. When a passage in the Tongjian does not match any surviving source, it likely came from a source that is now lost. This reverse-engineering approach has been productive but can never fully recover the congmu. For more on Sima Guang’s methods, see the reading guide and the complete guide.
Leave a Reply
You must be logged in to post a comment.