Dating the Shizi

Carine Defoort, Leuven
WSWG 17, Leiden University, 18 September 2003

Carine Defoort


The Shizi has since the Han dynasty been attributed to one of Shang Yang's advisers, who fled to Shu when the latter was executed in -338. Nowadays, some scholars (Shui Weisong, Wei Qipeng) indeed date the book to the -4th century. Others (Sun Xingyan, Forke, Graham) consider it a late Zhou or Qin syncretic text (late -3rd c.). Most contemporary research on pre-Han thought does not mention the Shizi at all, because of its questionable authenticity (e.g. Loewe, Early Chinese Texts).

1. The Text

Three Chinese scholars in the '20s laid the basis for a critical evaluation of the Shizi and published their views in the Gushibian (Zhang Xitang in v4, 646-653; Sun Cizhou in v6, 101-112; Jin Dejian in v6, 306-313). The critical verdict is that the current Shizi is a mixture of textual fragments of at least three different times:

  • The first author was a legalist adviser of the -4rd century.
  • A second layer was written in the Eastern Han by a Confucian Guliang zhuan scholar.
  • The third layer was rewritten during the Wei dynasty between +220 and +226, after 9 chapters had been lost.

The text had disappeared by the end of the Song, and was only preserved through quotes in other books. Five Qing scholars made reconstructions of the text on the basis of these quotes. The last reconstruction, by Wang Jipei (fl. c1800, preface 1811), is based on his predecessors' work and is generally considered the best. Wang's reconstruction consists of three parts:

  • Part I (13 chapters): taken from the Qun shu zhi yao, chapter 36.
  • Part II: based on quotes from other sources, such as Er Ya (-3rd century?) and Bei tang shu chao (before +618)
  • Part III: a short collection of what Wang considers doubtful attributions to the Shizi.

I want to show that a core of the Shizi (the part preserved in the Qun shu zhi yao) could very well date from around the late Zhou or Qin dynasty. I will therefore first refute the Gushibian criticism of the Shizi, and then suggest some positive arguments in favor of a pre-Han dating.

2. Response to Criticism of the Shizi

The division of the Shizi in three layers by the Gushibian scholars is based on Parts I and II of the current Shizi without discrimination. There are some problems with their arguments:

  • First, the distinction between the two first layers is partly based on their attribution of the text to a legalist adviser of the -4rd century. Later concepts (e.g. the combination of ren and yi, speculations on the heart/mind) and the combination of various strands of thought (e.g. fa + ru) makes the Shizi dubious in their eyes. I believe that these concepts are not problematic if one accepts that the core of the text is syncretist and, at the earliest, from the -3rd century.
  • Second, most of their dating arguments are based on fragments from Part II, some of which may indeed postdate the Qin. My focus is on a core preserved in the Qun shu zhi yao. This compilation of texts was presented to the court by Wei Zheng in 631, got lost after the Tang, but was preserved in Japan and returned almost completely during the reign of Qianlong.
  • Third, and most interestingly, even when dating arguments are based on fragments in Part I, several of these fragments do not occur in the Qun shu zhi yao, 36 (SPCK, 472-482), but are inserted on the basis of other sources (eg in ch. 9 and ch. 10) by Wang Jipei.

The only remaining arguments in favor of a post-Qin date (Han or later) of Part I are:

  • Shizi 1 quotes Zengzi, and Shizi 6 quotes Zengzi while attributing the saying to Confucius. Sun Cizhou (p108) believes that Shizi quotes from the Da dai liji. But what is the direction of citation? What is the source?
  • Shizi 11 quotes an Ode, "Nan feng ge" (title + content) of which the content was perhaps not yet known in the Han. Jin Dejian (p312) argues that Zheng Xuan (127-200) says in his commentary to the "Yueji" that "its lines are not yet (or "not quite," wèi?) known". This point is at best tentative.

3. Positive arguments in favor of the Qun shu zhi yao core of the Shizi

Grammar (the lists of criteria may not be totally reliable):

  • Checked against Graham's (Studies, 249-264) ten linguistic criteria for determining pre-Han texts, the thirteen Shizi chapters preserved in the Qunshu zhiyao fit the test.
  • Karlgren's criteria ("The Authencticity...," 61-65) to distinguish 3rd century B.C. texts confirm this date.

Vocabulary (only some expressions):

  • Zhengming occurs in a political context in ch. 5 and 6 as in most pre-Han texts. It is most closely related to Lüshi chunqiu, 17:1. More abstract and philosophical speculations on zhengming that predominate in Han texts, such as the Chunqiu fanlu and the Yinwenzi, are totally absent from Shizi, Part I.
  • Yinyang in the sense of cosmic forces and primary binary opposites is absent from Shizi, Part I.


All lectures and abstracts posted on this site are Copyright © by their authors.

 Back to Conference Program

20 May 2003 / Contact The Project / Conferences Page