UBC Community, Partners, and Alumni Publications

Shangshu 尚書 ("Exalted Writings" ; also, "Documents") Poli, Maddalena


The Shangshu 尚書 (Exalted Writings; also known as Book of Documents) is one of the foundational texts in Chinese history, and with a most complex textual history, as recorded in Han 漢 dynasty (202BCE–220CE) documents. This text is a collection of speeches attributed to rulers and important ministers of the Zhou 周 dynasty (1046–256 BCE). It became one of the Five Classics 五經 during the Han dynasty (see Nylan 2001), i.e. one of the texts on which the imperial curriculum was based. Knowledge of these classics was prerequisite to assume any official position. While some chapters were put together after the fall of the Han dynasty, almost half of it are genuine compositions of early date, and present very challenging language. Wang Guowei 王國維, a most experienced scholars in the study of Chinese classics, admitted to understanding no more than half of them, and that it is true of previous scholars too,《詩》《書》(...) 於六藝 中最難讀。以弟之愚闇,於《書》所不能解者殆十之五;於《詩》亦十之一二。漢魏以來諸大師未嘗不強為之說,然其 說終不可通,以是知失儒亦不能解也。(In Guantang jilin 觀堂集林). ____ One of the loci classici to begin to understand the history of this text is the Treatise on Arts and Letters 藝文志, the oldest extant bibliography included in the Book of Han 漢書. In there, the Shangshu is presented as originally of 100 chapters (pian 篇) and a preface, of which only 28 survived. Han scholar Fu Sheng 伏生 (active second century BCE) is credited with having saved these 28 chapters from the burning of the books that occurred during the Qin 秦 dynasty (221–202 BCE), to which another one, the Taishi 太誓, was added. These became known as jin wen 今文 chapters, because they were written in clerical script (in use during the Han dynasty). However, Fu Sheng was an official serving under the Qin, thus belonging to a category that could possess books. Hence this narrative poses a contradiction on why he needed to hide this text at all. 16 more chapters were recovered from in the wall of the once house of Confucius, when Prince Liu Yu 劉餘 (d. 127 BCE) started construction to enlarge it. These were written in a script that preceded the standardization of the script ordered by Qin minister Li Si 李斯 (circa 280–208 BCE), and became thus known as guwen Shangshu 古文 尚書 (ancient script Shangshu). The Treatise seems to imply that a witch trial interrupted the establishment of an office for the study of these newly recovered chapters. In his memorial to Emperor Ai 哀 (r. 7–1 BCE), Liu Xin 劉歆 calls for the establishment of a boshi 博士 (“erudite”) position for the study of all texts in ancient, suggesting that these 16 guwen chapters were located in the imperial library. After this reference, we lose track of these 16 chapters. While the Book of Han most certainly refers to it as a somewhat defined collection, the Records of the Historian (which mentions a total of 68 chapters) seems to be using the term also to indicate a category of writing. ____ While Liu Xin’s collated edition of jinwen and guwen chapters seems to have remained within the library, the 29 jinwen chapters enjoyed wider readership, and were frequently referred to as Shangshu. As for many other texts during the Han dynasty, the Shangshu too presents different lineages and scholars who taught this text, most famously Ouyang Gao 歐陽高, Xiahou Sheng 夏侯勝 and his nephew Xiahou Jian 夏侯建. Thus, in the body of Han literature we have references to the Ouyang Shangshu 歐陽尚書, the Da Xiahou Shangshu 大夏侯尚書 and the Xiao Xiahou Shangshu 小夏侯 尚書, however without extensive details on the textual differences that existed among these lineages. Of these, the Ouyang Shangshu was the most renown, and one for which one textual change is mentioned: Ouyang divided the Pangeng 盤庚 chapters into three parts (as we still read it today), making it into a 31- chapter Shangshu. ____ The next central figure for the history of the Shangshu is that of Du Lin 杜林 (first CE), who made a lacquer edition of the Shangshu in 29 chapters. Because this edition was written in ancient script, it became known as Guwen Sangshu 古文尚書. This led to the mistaken assumption that his edition included the 16 guwen chapters found in the first century BCE. These 29 chapter Shangshu eventually reached Mei Ze 梅賾 (fl. Beginning fourth century CE), the Easter Jin 東晉 dynasty scholar who is responsible for the shape of the Shangshu in 58 chapters that we see today. After reorganizing the 29 chapters into 33, he fabricated (wei 偽) the remaining chapters on the basis of older material and quotations in other texts. In fact, there is no evidence a single person was responsible for all the fabricated chapters. However, it is plausible that he was the one who eventually presented this 58 chapter Shangshu to the Jin court. While Song 宋 scholars raised some eyebrows about language used in the chapters that Mei Ze composed, the Shangshu remained by and large unquestioned, until Yan Ruoqu 閻若璩 (1636–1704) and Hui Dong 惠棟 (1697–1758) identified the chapters added by Mei Ze as fabrications, showing their inconsistencies and linguistic mistakes. ____ A new chapter in the history of the Shangshu is being now written, following the recovery of bamboo manuscripts, in particular those now part of the Tsinghua University 清華大學 collection. Among these manuscripts, several bear titles similar to titles of Shangshu chapters that were listed as lost during the Han Dynasty, such as the Fu Yue zhi ming 傅說之命; others are written in what is now called shu 書 style (Allan 2012), such as the Feng xu zhi ming 封許之命, the Zhi zheng zhi dao 治政之道, and the Si gao 四吿. As with many of the recently recovered manuscripts, scholars lively debate on the nature of these manuscripts. The initial publication (Qinghua daxue cang Zhanguo zhu jian 清華大學藏戰國竹簡) identified many of these manuscripts as the de facto lost chapters of the Shangshu, including those not mentioned by Han texts. More recent scholarship has started to propose different interpretations on the nature of these manuscripts: while some resemble polished texts that might as well be originally part of the Shangshu, some manuscripts present mistakes and incongruities that suggest they might have been used as master copies or templates (most recently, Cheng Hao 2020; Jia Lianxiang 2020). [This entry was revised on September 27th, 2021.

Item Citations and Data


Attribution 4.0 International