Working with the Shiji is less about reading it and more about surviving its textual problems
Most people approach Sima Qian Records Of The Grand Historian as if it is a single clean text you can read cover to cover. It is not. What exists is a transmitted manuscript tradition that has been copied, rearranged, edited, and partially lost over roughly two thousand years. If you are trying to use it for anything beyond casual reading, you will run into problems fast unless you understand what edition you are working from and which passages are genuinely problematic. The base text everyone uses comes from the Zhonghua Book Company 1959 edition, which was collated by Xia Nai and others using the Wang Niansun and Duan Yucai traditions. That edition is reliable for the vast majority of the 130 chapters. But it is not authoritative everywhere. Some chapters have known lacunae. The "Basic Annals" sections are generally stable. The "Hereditary Houses" and some of the later biographical sections have more variant readings across manuscripts.If you are doing serious work with the text, you need at minimum three reference editions side by side.
The standard sequence is the Zhonghua punctuation edition, the Fang Yizhi annotated version from the Qing, and the Siku Quanshu compilation notes where they differ. I spent about six months trying to pin down a specific passage in Chapter 63 about the Duke of Zhou's regency because every modern translation I found was silently emending the text without saying so. The workaround was pulling the Ming dynasty Jiyu edition facsimile and comparing it line by line against the Zhonghua apparatus. The passage in question had been altered at least twice in the transmission chain. Once you know which version you are reading, the argument changes completely.Sima Qian Records Of The Grand Historian primary source workflow
The practical workflow most people skip involves establishing your base chapter, checking the Zhonghua editorial notes for every passage you plan to quote, and then verifying against the earliest available printed edition for that chapter. The Shiji was first fully printed in the Song dynasty. Southern Song editions survive in fragments. The most complete early witness is the Jiu Min edition from the Jin dynasty, but most scholars work from the Song-Xia period reconstructions. Here is where beginners get tripped up. The 130-chapter structure is Sima Qian's own framework, but several chapters are widely believed to contain later insertions. Chapters 104 through 108, the biographies of certain figures from the HanWudi period, have textual features that do not match the surrounding material. Some scholars attribute parts of these to Sima Qian's daughter Guo Sheng or to later editors filling gaps after the original text was damaged. When you encounter a passage that feels tonally inconsistent with the rest of the chapter, that feeling is often correct. The other thing nobody warns you about is the date problem. Sima Qian died sometime after 91 BCE, probably around 86 BCE. The Shiji covers history from the Yellow Emperor down to about 104 BCE in the main narrative, with some updates extending to the reign of Han Wudi. But the dates given in the text for early legendary periods are notoriously unreliable. The chronology in the "Basic Annals" of the five ancient sovereigns is built on conflicting sources that Sima Qian himself acknowledged were uncertain. He wrote that he did not dare to set them down definitively. Modern readers treat those dates as fact. They are not.I once corrected a graduate student's citation that traced a ritual practice back to 2852 BCE based on the Shiji. The date itself was a later Han interpolation. The passage she quoted did not appear in the earliest receivable text of that chapter. It showed up in a Tang dynasty commentary that had been mistakenly integrated into the main body by a Song editor. The error propagated through at least twelve secondary sources before anyone checked the primary edition.
For anyone actually reading the Chinese text, the Guodian bamboo slips and the Shanghai Museum collection have been useful for crossreferencing certain classical references Sima Qian makes, particularly around the early Zhou period. They do not contain Shiji material directly, of course, since the bamboo texts predate Sima Qian by centuries. But they clarify some of the source material he was working from, which helps you judge when he may have been misinformed or when he made an editorial choice that changed the meaning. The English translations are serviceable but uneven. Burton Watson's twovolume version covers the major sections and is readable. Ralph Sykes covered more ground but is older and less careful with names. David Hawkes' selected translations are excellent for individual chapters but not comprehensive. If you are quoting from an English translation in any academic context, always verify the Chinese original. Watson sometimes chooses a rendering that favors narrative flow over textual precision, and Sykes has occasional misreadings of classical Chinese grammar that flip the meaning of a sentence. The most useful companion resource is the Shiji zhengyi, the Tang dynasty commentary compiled by Pei Yin. It preserves readings from editions that no longer exist and cites source materials Sima Qian used that have since been lost. Without Pei Yin, many passages in the Shiji would be unintelligible. With it, most become manageable. The Zhonghua edition includes the zhengyi notes in a separate section for each chapter, which is convenient but makes the text heavy. I usually keep the main Zhonghua text open alongside a standalone zhengyi reference for whichever chapter I am working on.
