This collaborative volume on the classical Chinese text called Shangshu 尚書 (usually translated as Documents)Footnote ¹ is the fruit of two related conferences, held at Princeton and Oxford in 2013 and 2014, respectively, with chapters by contributors ranging from current graduate students to full professors. As Martin Kern and Dirk Meyer note in their introduction (p. 1f.), it stands as one of the few extensive treatments of the text in contemporary Western scholarship. The reasons for this neglect are straightforward and embarrassing: Shangshu is difficult because of its language, and sadly also tedious because of its content.Footnote ² Origins of Chinese Political Philosophy opened my eyes to several features of the text that I had never noticed, but I think most readers hoping to find philosophically engaging material in Shangshu will still be disappointed. Those who need to study it (whether out of genuine interest or the historicist recognition that the text is simply too important to ignore) will want to begin with this book, which represents the best of current scholarship.

Origins of Chinese Political Philosophy is unusually coherent for a collaborative volume (and lucidly written, it should be added). Constraints of space preclude a detailed consideration of each chapter; in addition, some contributors refer in detail to looted manuscripts, which I shall not cite out of principle.Footnote ³ But the book can readily be discussed as a whole because the methodological orientation varies little from chapter to chapter. While most twentieth-century scholarship (both Asian and Western) was devoted to determining plausible dates of composition for all the sections of Shangshu—and accordingly whether they should be designated as “genuine” (zhen 真) or “forged” (wei 僞)—Kern and Meyer (p. 6, n.20) revive an important insight by Jiang Shanguo蔣善國: the relevant question is not when each chapter of the text was composed, but when it was redacted.Footnote ⁴ Origins of Chinese Political Philosophy demonstrates impressively why Jiang's observation was correct.

One reason is that chapters in the received text often turn out to comprise two or more strands. In the first study in the book (“Language and the Ideology of Kingship in the ‘Canon of Yao,’” pp. 23–61),Footnote ⁵ Kern shows that “Yao dian” 堯典, the opening chapter of Shangshu, contains divergent narratives of the Sage Kings Yao and Shun 舜, which not only are marked by distinct rhetoric, but also present the two sovereigns in contrasting modes: Yao as a dynamic and charismatic figure, Shun as an invisible force operating behind the impersonal machinery of state.Footnote ⁶ Kern identifies the second half as later than the first, and associates it with the ideology of the Qin 秦 and Western Han 西漢 empires. If he is correct, then asking when “Yao dian” was written would qualify as a category mistake, because different parts were written at different times (and probably for different purposes). Once again, the relevant question is when the text was redacted: in this case, sometime during the early empire, in order to provide doctrinal cover for a revolutionary ideology that emphasized efficient bureaucratic administration rather than the unique powers of the king. Immediately after Kern, Kai Vogelsang (“Competing Voices in the Shangshu,” pp. 62–105) shows that two other chapters, “Gao Yao mo” 臯陶謨 and “Lü xing” 呂刑, can similarly be divided into “A” and “B” sections, the first portraying the emperor as the virtuous head of his lineage, the second as a figurehead surrounded by meritorious and genetically unrelated ministers (e.g., p. 71).Footnote ⁷

In a later chapter (“The ‘Harangues’ [shi 誓] in the Shangshu,” pp. 281–319), Kern again relies on structural and rhetorical features (pp. 289–303) to show that three “rousing battle speeches” (p. 282), namely “Gan shi” 甘誓, “Tang shi” 湯誓, and “Mu shi” 牧誓, belong to a discernible genre that he calls “harangue”.Footnote ⁸ Building on Jan Assmann's notion of cultural memory (kulturelles Gedächtnis), Kern argues that these speeches, though placed in the mouths of kings of high antiquity, “were created as idealized artifacts to literally overwrite the actual historical events and to make history conform to the moral norms of a later age”.Footnote ⁹ (p. 284) Toward the end of this study, Kern makes another important point: “harangues” cannot be dated by linguistic features because they were not composed at a single historical moment. Rather, drawing on an old suggestion by Nomura Shigeo 野村茂夫, Kern contends that they are best understood as repertoires of constantly revised and renegotiated material, hypertexts “of a common underlying hypotext”.Footnote ¹⁰(p. 307) Here too, the simplistic “genuine-or-forged” discourse of earlier textual criticism proves inadequate. The “harangues” are not plausible as contemporaneous accounts (p. 288), but this does not mean that they were simply forged.

A final strength of the book is that it sheds considerable light on the processes by which Shangshu was closed and canonised, but the evidence is spread over several chapters. Yegor Grebnev (“The Yi Zhoushu and the Shangshu: The Case of Texts with Speeches,” pp. 249–280) compares chapters from Shangshu with those from a similar collection, Yi Zhoushu 逸周書 (which he translates as Remnant Zhou Documents). Although one would have liked to see an account of the latter text's history,Footnote ¹¹ it is clear that Grebnev has hit on an important criterion when he observes that “dramatic” speeches are more typical of Shangshu, “non-dramatic” speeches more typical of Yi Zhoushu (pp. 266–270). As Grebnev explains, “the former are emotionally laden and personalized and have a richer repertoire of emphatic devices, while the latter appear as treatises superficially furnished [with] emphatic devices reminiscent of dramatic speeches” (p. 255); “dramatic” speeches, similarly, contain an even distribution of “first- and second-person pronouns, vocatives, and exclamations” (p. 256). Thus if the material in the two collections derives from the same fund of lore, someone (or some committee) used certain selection criteria when deciding which to include in Shangshu (p. 276f.). Grebnev also reminds us that each chapter of the received Shangshu includes a preface purporting to explain the circumstances of its composition (p. 271), another obvious sign of canonisation.Footnote ¹²

Joachim Gentz (“One Heaven, One History, One People: Repositioning the Zhou in Royal Addresses to Subdued Enemies in the ‘Duo shi’ 多士 and ‘Duo fang’ 多方 Chapters of the Shangshu and in the ‘Shang shi’ 商誓 Chapter of the Yi Zhoushu,” pp. 146–192) observes that Shangshu chapters are more similar to one another than to any other genre (pp. 176–177), and David Schaberg (“Speaking of Documents: Shu Citations in Warring States Texts,” pp. 320–359) draws attention to the number 28, which appears in too many places to be coincidental (p. 350f.). (I never would have noticed this.) For example, the sky is divided into twenty-eight “lodges” (xiu 宿).Footnote ¹³ Thus it is striking that the jinwen 今文 edition of Shangshu contains precisely twenty-eight chapters. Someone governing the final form of the text chose a numerologically significant length.

I did detect one latent contradiction throughout the book—at any rate, an assumption that might have been examined explicitly. If scholars such as Kern, Vogelsang, Meyer, and Gentz are right that Shangshu chapters exemplify repertoires based on a shared hypo-text (Kern) or are constructed out of “modules” that can be combined and recombined in a variety of permutations (Gentz, pp. 160–163), then—as they themselves emphasize—one has to be careful about appealing to such material as proof of particular beliefs or practices in the Bronze Age. But occasionally they make what seems like the same mistake that they criticise in others. For example, at one point Meyer writes:

Which “ancient sources”? He does not say, but I can only suspect that he is referring to Edward L. Shaughnessy's analysis of “Shao gao” 召誥 (a very famous chapter of Shangshu that is scarcely mentioned in this book).Footnote ¹⁴ One has to ask why “Shao gao” should be accepted unproblematically as an “ancient source” if the other chapters, as we have been learning, are to be construed as late reconceptualisations of earlier traditions. Or is there a deeper problem: is it the case that some chapters are truly “ancient sources” but others are more complicated?

Several contributors rely on one text in particular as representative of Bronze-Age values: “Shi fu” 世俘, a notoriously bloody narration of the Zhou conquest currently found in Yi Zhoushu.Footnote ¹⁵ I counted no fewer than five instances in this book (pp. 152, 181, 285, 297, and 308) when a contributor contrasted material in Shangshu with the supposedly unvarnished account in “Shi fu”. The suitability of “Shi fu” for this purpose is never challenged—yet one can imagine how “Shi fu” would also have to be reinterpreted if it were subjected to the same sort of textual destabilisation that is performed on other texts in this book. After all, nobody has given us more reason to be wary of reading them at face value than the contributors themselves.

Lastly, some readers may wonder about the title: Origins of Chinese Political Philosophy. On the one hand, the contributors are not to blame for the fact that Shangshu is less than thrilling as philosophy; on the other hand, while they are thoughtful and persuasive when they show how to read the text (and how not to read the text), they do not make it sound any less preachy. One telling detail is that “Kang gao” 康誥 is mentioned more than once as part of “the earliest layer” (e.g., pp. 173 and 370), but no one was inspired to devote a chapter to it. I cannot help thinking that the reason is all too disheartening: “Kang gao” is an infernal bore. It contains tireless exhortations to treat widows and widowers kindly, to revere the people who ought to be revered, to be ever mindful of Heaven's awesome mandate—but never any doubt as to what these injunctions mean (or why we should act on them). That, in my view, is necessary for philosophy: the awareness that there can be other perspectives, that a moral life requires thinking for oneself and not simply living up to the expectations of some unquestioned authority. If there were people in the Bronze Age who ruminated along such lines, we do not have any record of their ideas.

There are occasional philosophically interesting moments. Yuri Pines (“A Toiling Monarch? The ‘Wu yi’ 無逸 Chapter Revisited,” pp. 360–392) and Michael Hunter (“Against (Uninformed) Idleness: Situating the Didacticism of ‘Wu yi’ 無逸,” pp. 393–415) disagree over the meaning of the word zhi 知 (to know, to be aware of, to understand) in the declaration that the ruler must zhi “the hardship of sowing and reaping” 先知稼穡之艱難 (Pines's translation, p. 363). Does this mean that the ruler must experience manual labour personally (Pines, pp. 367f. and 372) or merely that he must understand its importance (Hunter, p. 406)? The text is too underdetermined to be sure of the answer, but philosophical readers might take it as an early suggestion of the unity of knowledge and action (zhixing heyi 知行合一), a cardinal idea of the much later philosopher Wang Shouren 王守仁 (1472-1529).Footnote ¹⁶ Can a king be said “to know” manual labour if he does not do it? A student of Ming philosophy might even be motivated to check what Wang himself had to say about the “Wu yi” chapter (which he probably knew by heart).

In sum, Origins of Chinese Political Thought might not induce readers to drop their Zhuangzi 莊子 or Xunzi 荀子 and dust off their mouldering copy of Shangshu instead, but it does show how the text slowly yields its patterns under sound methodological investigation. Moreover, the interpretive techniques so effectively deployed in this volume could undoubtedly be applied to other texts as well. Several of the contributors might already be contemplating such research.