1. The characteristic disdain for Han Confucian thought is well reflected by the following comment of the late Angus Graham: “Competition was now for the ear of a single ruler, and adopted to the lowest common denominator of rulers, with the aim of distracting them from the realities of political power by frightening them with omens or promising them the elixir of life”; Disputers of the Tao (La Salle: Open Court, 1991), 372 Google Scholar. A similarly low level of expectation is displayed by K.C. Hsiao, who notes that two systems of official organization advocated in different parts of the Chunqiu fanlu are inconsistent, but never thinks to question the authenticity of either; A History of Chinese Political Thought volume 1, translated by Mote, F.W. (Princeton: Princeton University Press, 1979), 491 n. 69Google Scholar. It seems he hardly expects consistency of Han Confucians.

2. Even some early editors of the Chunqiu fanlu were not entirely convinced of its reliability. For instance, the preface by Lou Yu 樓郁 which survives from the 10-juan edition printed in 1047 states: “Its authenticity I leave to be argued by worthies”; Chunqiu fanlu edition of A.D. 1211 ( Beijing tushuguan guji zhenben congkan 北京圖書館古籍珍本叢刊 2, Peking: Shumu wenxian chubanshe, n.d. ?1990), 507 Google Scholar. I will refer to this edition hereafter as Beijing Chunqiu fanlu.) The most prominent pre-modern skeptic was Zhu Xi 朱熹 (1130–1200), who replied to one direct inquiry by saying, “I have taken a look at it, and it is not Master Dong's work”; Guji kaobian congkan 古翁考辨叢刊 vol. 1, ed. Jiegang, Gu 顧頡岡！！ (Peking: Zhonghua shuju, 1955), 48.158 Google Scholar. A minority of Japanese scholars have always maintained a healthy suspicion of the Chunqiu fanlu: one early example is Masami, Harada 原田正己 in his essay “Kan Ju no bunshitsu setsu” 漢儒の文質說, Tōyō shisō kenkyū 東洋思想硏究 2 (1938), 509 n. 19Google Scholar. More recently, in addition to Tanaka Masami (discussed below), the authenticity of the text has been attacked by Fukui Shigemasu 福井重雅 (” Jukkyo seiritsushijō no ni, san no mondai” 儒敎成立史上の二三の問題, Shigaku zasshi 史學雜誌 76.3 [1967])Google Scholar and Kenji, Iwamoto 岩本憲司 (“Kandai Shunjūgaku ni kansuru ni, san no mondai” 漢代春秋學にん十る二三の問題, Atomi Gakuen Joshi Daigaku kōyō 跡見學園女子大學紀要 16 [1983], 129–139.)Google Scholar

3. This standpoint is epitomized by the following remarks from Lixue, Lin 林麗雪, Dong Zhongshu (Taipei: Taiwan Shangwu Yinshuguan, 1978), 12 Google Scholar: “Summing up, the present edition of the Chunqiu fanlu was certainly not put into shape by Dong Zhongshu personally, but is a work edited from scattered texts by later men. The confusion, lacuna, and errors in it cannot be overlooked, and it is also possible that a portion was added on by scholars after Dong. Nonetheless, its content is more than adequate to shed light on the writings of Dong passed down in the Honshu, and also matches the consciousness of the age of intellectual development at the beginning of the Han. It is proper to rely on it for research into Dong Zhongshu's scholarship.”

4. Xu Fuguan 徐復觀, for instance, brushed aside the usual evasions to boldly claim that “the Chunqiu fanlu text we have today has only damage and lacuna; it is entirely free of admixtures of unreliable material”; Liang Han sixiangshi 兩漢思想史, volume 2 (Taipei: Xuesheng shuju, 1979), 315–16Google Scholar. He gave two arguments to support his radical conclusions. The first was a deduction from what he alleged was the “not yet entirely merged” state of development exhibited by the Yin-Yang and Five Forces portions of the text. “And the doubts that many hold towards [the Chunqiu fanlu] are chiefly due to their not being able to master this characteristic from the standpoint of the whole history of Chinese thought, and so considering that Dong Zhongshu should not have been characterized by such a jumble.” His second argument was that passages cited elsewhere and attributed to Dong are found in the Chunqiu fanlu. However, none of the texts he mentioned —or could have mentioned — makes any mention of the Five Forces. Some of Xu's disciples, such as Lai Qinghong 賴慶鴻, have taken his already shaky arguments to the point of absurdity (see the remarks on textual authenticity in Lai's, Dong Zhongshu zhengzhi sixiang zhi yanjiu 董仲舒政治思想之硏究 ( Taipei: Wenshizhe chubanshe, 1981), 14–23 Google Scholar.

5. This approach began in earnest with the essay by Mitsuo, Keimatsu 慶松光雄, “ Shunjū hāmo gogyō shohen gisaku kō” 春秋繁露五行諸篇偽作考, Kanazawa daigaku hō-bun gakubu ronshū (tetsugaku bungaku) 金澤大學法文學部論集 (׳哲學文學） 6 (1959), 25–46 Google Scholar. In his bibliographic review of Dong Zhongshu studies, Timoteus Pokora mentions Keimatsu's study with approval; “Notes on new studies on Tung Chung-shu,” Archiv Orientální 33 (1965), 266 and note 51Google Scholar. Junren, Dai 戴君仁 came to almost the same conclusions independently a few years later;, “Dong Zhongshu bu shuo wuxing kao” 董仲舒不說行考, Guoli zhongyang tushuguan guankan 2.2 (1968), 9–19 Google Scholar; repr. in Junren, Dai, Meiyuan lunxueji 梅園論學集 (Taipei: Kaiming shuju, 1970), 319–334 Google Scholar. The work of Keimatsu and Dai provided the first conclusive evidence that an entire theme in the Chunqiu fanlu was unrepresentative of Dong. In 1969, Tanaka Masami 田中麻紗己went on to point out that the Five Forces material in the Chunqiu fanlu was not even internally consistent (repr. in his Ryō Kan shisō no kenkyū 兩漢思想 の 硏究 [Tokyo: Kenbun shuppan, 1986], 34–51)Google Scholar. Earlier, however, the late Qing scholar Yu, Su 蘇輿 had denied the authenticity of several chapters on the basis of content (Chunqiu fanlu yizheng 春秋繁露義證 [Changsha: 1914 Google Scholar; repr. Kyoto: Chūbun shuppansha, 1973; Taipei: Heluo tushu chubanshe, 1975], 1.23b, 6.16a, 7.25b, 7.30a, 7.32a), and Toshio, Shigezawa 重澤俊郞had accounted for an inconsistency in cosmological theory by suggesting both early and late work by Dong was present in the text (“Tō Chūjo”” 董仲舒, in Shū-Kan shisō kenkyū 周漢思想硏究, ed. Toshio, Shigezawa, [Tokyo: Kōbundō shobō, 1943], 236–244)Google Scholar. The analytic approach is followed to some extent in the discussion of textual authenticity in Queen, Sarah, “From Chronicle to Canon: The Herrnaneutics of the ‘Spring and Autumn Annals’ according to Tung Chung-shu” (Ph.D. dissertation: Harvard University, 1991)Google Scholar, and the whole text is analyzed at length in my own dissertation: “Restoring Dong Zhongshu (BCE 195–115): An Experiment in Historical and Philosophical Reconstruction” (Ph.D. dissertation, University of British Columbia, 1991)Google Scholar; see in particular pp. 263–314 (on the development of the Gongyang school after Dong Zhongshu) and pp. 315–542 (on the Chunqiu fanlu itself).

6. The Shunjū hanro 春秋繁露 by Toshikuni, Hihara 日原利國 (Tokyo: Meitoku, 1977)Google Scholar translates only the first five chapters.

7. Yanyuan, Lai, Chunqiu fanlu jinzhu jinyi 春秋繁露今註今譯 (Taipei: Shangwu yinshuguan, 1983)Google Scholar.

8. All are ostensibly studies of “Dong Zhongshu,” but all accept the authenticity of the Chunqiu fanlu and make extensive and uncritical use of it. The first, and by a considerable margin the best, is “Tung Chung-shu's System of Thought, Its Sources and Its Influence on Han Scholars” by Tain, Tzey-yueh (Ph.D. dissertation: University of California, Los Angeles, 1974)Google Scholar. The others, both completed in 1982, are “The Political Philosophy of Tung Chung-shu (179–104 B.C.): A Critical Exposition” by Vuylsteke, Richard R. (University of Hawaii)Google Scholar and “Tung Chung-shu and the Origins of Imperial Confucianism” by Davidson, Steven Craig (University of Wisconsin-Madison)Google Scholar. In contrast to Tain's sincere but insufficiently detailed investigation of the authenticity question, these two latter dissertations ignore it entirely, while contributing nothing of worth in other areas. The only people who might profit from reading either are connoisseurs of turgid style and typographical carelessness (see in particular the typo on Davidson p. 287, where “the shift from classical Confucianism” is rendered sans the “f” in “shift.”)

9. For Dong's writings during the Han dynasty, see the discussion of the authenticity of the “Section A” chapters given below. More detail will be found in Arbuckle, , “Restoring Dong,” 55–79 Google Scholar, and Queen, , “From Chronicle to Canon,” 11–35 Google Scholar.

10. Preserved in a quotation from the Shiji zhengyi 史記正義, in Kametarō, Takigawa 瀧川龜太郞, Shiki kaichū kōshō 史記會注考證 (repr. Taipei: Yiwen yinshuguan, n.d.), 121.28 Google Scholar.

11. Beijing, Chunqiu fanlu, 507 Google Scholar.

12. An eight-juan version does exist (in the Liang jing yipian 兩京遺篇, a collection edited by Hu Weixin 胡維新 in the Ming) but it is nothing more than a reprint of the first eight juan of the received text. This makes it impossible that it descends from the Song or pre-Song ten-juan editions, which contained at least some of the chapters now found in the last seven juan.

13. Beijing, Chunqiu fanlu, 606–607 Google Scholar.

14. Beijing, Chunqiu fanlu, 507–519, 577 Google Scholar.

15. Yu, Su, Chunqiu fanlu yizheng, 11.1a Google Scholar.

16. With the exception of “Chu Zhuang wang,” only four chapters use their unmodified opening words as titles: 77, “Xun tian zhi dao” 循天之道; 78, “Tian di zhi xing” 天地之行 (which is not a chapter at all, but a commentary on Chapter 18, misplaced and mistitled here; cf. Arbuckle, , “Restoring Dong,” 334–341)Google Scholar; 81, “Tian di yin yang” 天地陰陽; and 82, “Tian dao shi” 天道施. These are clustered at the end of the book, in what I refer to as Section G, the miscellaneous group. The only apparent exception is with Chapter 66, “Jiao Yi” 郊義 ( Yu, Su, Chunqiu fanlu yizheng, 15.1a Google Scholar), but here the jiao yi which begins the text makes no sense in context, and appears to be either a mistaken duplication of the title, or a relic of an earlier stage of the text where titles stood directly at the heads of their essays.

17. See Arbuckle, , “Restoring Dong,” 324–342 Google Scholar, where I demonstrate that much of the textual disarray falls into a pattern indicating it can be traced back to a single copy that contained around 420 characters per double page.

18. Apart from the three early chapters whose titles Ban Gu mentions, there are four more that might go back to Dong himself. These are Chapter 32, “Dui Jiaoxi wang Yue dafu bu de wei ren” 對膠西王越大夫不得爲仁 (see Arbuckle, , “Restoring Dong,” 67–76), Chapter 44Google Scholar, “Wang dao tong san” 王道通三 (“Restoring Dong,” 386–387), Chapter 73, “Shan Chuan song” 山川頌 (see Arbuckle, Gary, “A Note on the Authenticity of the Chunqiu fanlu: The Date of Chunqiu fanlu Chapter 73, ‘Shan Chuan song” 山川頌 (‘Praise-ode to Mountains and Rivers’),” T'oung Pao 75 [1989], 226–234)CrossRefGoogle Scholar, and possibly Chapter 71, “Jiao shi dui” 郊事對 (“Restoring Dong,” 76–77, and “The jiao 郊 ‘suburban’ sacrifice in the Chunqiu fanlu”, paper for the annual meeting of the Association for Asian Studies, New Orleans, April 1991).

The use of cosmological political code in six other chapters indicates that they were written as political propaganda by Han loyalists in the final years of Former Han or during Wang Mang”s Xin dynasty. These are Chapters 38, “Wu xing dui” 五行對; 42, “Wu xing zhi yi 〃五行之義; 43, “Yang zun yin bei” 陽尊陰卑; 58,”‘Wu xing xiang sheng” 五行裙生; 59, “Wu xing xiang sheng” 五行相勝; and 60, “Wu xing shun ni” 五行順逆. The rationale for this dating will be found in Arbuckle, , “Restoring Dong,” 488–92Google Scholar, developed further in “‘New” Evidence on the Gongyang School,” paper for the 202nd meeting of the American Oriental Society, Cambridge, Massachusetts, March 1992.

19. The present text of Chapter 18 may be out of place here. See my analysis of the structure and content of this group in Arbuckle, , “Restoring Dong,” 459–470 Google Scholar.

20. Sarah Queen tends to accept the authenticity of these chapters, without much in the way of detailed discussion; see “From Chronicle to Canon,” 98–99, 137–138, 141.

21. Hanshu (shuju, Zhonghua ed.), 56.2515–2516 Google Scholar.

22. Hanshu, 30.1714, 30.1727Google Scholar.

23. It is often assumed that the eighty-one chapters of the Chunqiu fanlu represent the portion of the Dong Zhongshu which survived the battering of time between the end of the Han and the early sixth century. The fact that these chapters, several of which are found in the present Chunqiu fanlu (at least in name), do not seem to have been part of the Dong Zhongshu is evidence against this.

24. Yu, Su, Chunqiu fanlu yizheng, 1.17a Google Scholar.

25. Yu, Su, Chunqiu fanlu yizheng, 1.6a Google Scholar.

26. Hanshu, 27A.1332 Google Scholar.

27. For a discussion of Dong's cyclical theory of history, see Arbuckle, , “Restoring Dong,” 158–169 Google Scholar.

28. Cf. Arbuckle, “The Jiao 郊 ‘Suburban’ Sacrifice in the Chunqiu Fanlu.”

29. Yu, Su, Chunqiu fanlu yizheng, 3.14b Google Scholar.

30. This account is limited to points in the first six chapters, the ones which Gassmann has translated. Chapters 7–17 have a number of other problems, including an attitude to Duke Huan of Qi 齊桓公 sharply at variance with what we know of Dong Zhongshu's, and what may even be a veiled attack on Dong himself. For a more detailed review of the discrepancies between these first chapters and what we know of Dong's thought from reliable sources, see Arbuckle, , “Restoring Dong,” 447–459 Google Scholar.

31. This is Huwu Sheng 胡毋生 of Qi. See the remarks on him below and in Arbuckle, , “Restoring Dong,” 276–280, 456–459 Google Scholar.

32. See Arbuckle, , “Restoring Dong,” 2–3, 17 Google Scholar.

33. The most one can expect is a brief sketch of the chapter contents in the first footnote, and a certain amount of cross-referencing in some of the other notes.

34. For one thing, its notes are printed in a size so close to the text that it is some-times difficult to keep the two apart. Most researchers prefer Su Yu for his abundant annotations.

35. In fact, Hihara does not follow Ling Shu all the time, but when he introduces an emendation, it is based either on another text or on his own supposition, not on an independent textual tradition. For instance, at one place in Chapter 5 (p. 177), Hihara has sha 殺, “kill,” where all texts, including Ling Shu, have shi 試, “assassinate.” This seems clearly an error, and footnoting it, as Gassmann does (276 n. 54), is rather a waste of space. In at least one other place, Gassmann attributes an emendation to Hihara which was actually first proposed by the Qing dynasty editor Lu Wenchao 盧文招 (Gassmann 182 n. 7; cf. Yu, Su, Chunqiu fanlu yizheng, 3.21a Google Scholar).

36. Cf. the chart of Chunqiu fanlu textual filiation in Queen, “From Chronicle to Canon,” 329 ff.

37. William Seufert's old translation of Dong's, Hanshu 56 memorialsGoogle Scholar, “Urkunden zur staatlichen Neuordunung unter der Han-Dynastie,” Mitteilungen des Seminars fūr Orientalische Sprachen, 23.24 (Berlin, 1922), 1–50 Google Scholar, is cited in Gassmann's bibliography.

38. Yu, Su, Chunqiu fanlu yizheng, 1.25b–26a Google Scholar.

39. See Lunyu zhushu 論語注疏 9, Shisanjing zhushu 十三經注疏 (repr. Taipei: Wenhua tushu gongsi, 1970), 2491c Google Scholar.

40. Franke's translation of this chapter renders pian as “plötzlich” (“abruptly”) which is also incorrect but closer; see Franke, Herbert, Studien zur Geschichte des konfuzianischen Dogmas und der chinesischen Staatsreligion: Das Problem des Tsch'un-ts'iu und Tung Tschung-shus Tsch'un-ts'iu fan lu (Hamburg: L. Friedrischsen & Co., 1920), 276 Google Scholar.

41. It is in fact the title of Chapter 27 of the Chunqiu fanlu.

42. One of the few extended treatments of the cosmological cycle of “the refined” and “the bask” (wen 文 and zhi 質)is Harada Masami, “Kan Ju no bunshitsu setsu” (see note 2). It is not listed in Gassmann's bibliography.

43. The corresponding Guliang passage is largely identical, but renders the two final clauses 疾其始 and 樂其終 as “hated its initiation” and “took joy in its ending.”

44. The Guliang has ji qi yi huo gong ye 疾其以火攻也, “hated his attacking with fire.”

45. Gassmann also fails to note that Su Yu cites the authority of the Tianbi 天辟 edition of the text for his reading oifang here.

46. The line reads 止基適理, which Karlgren renders “the settlements were well distributed,” with a note that “chï-ki [止基] is a binome = 'settlement’”; Glosses on the Book of Odes (Göteborg: Museum of Far Eastern Antiquities, 1964), 81 Google Scholar.

47. Karlgren, Bernhard, Glosses on the Book of Documents (Göteborg: Museum of Far Eastern Antiquities, 1970), 270–271 Google Scholar.

48. The clearest of these is found in Han Feizi 韓非子 44 (“Shuo yi” 說疑): “Therefore when the worthy and good advanced, the wicked and deviant all retired, and so, by a single bringing (of the worthy) to prominence, they were able to subdue the feudal lords. It is found in the records, which say (qi zai ji yue 其在記曰), ‘Yao had Dan Zhu, and Shun had Shang Jun’”; see Han Feizi jishi 韓非子集釋, ed. Qiyou, Chen 陳奇猷 (rpt. Taipei: He-Luo chubanshe, 1974), 924 Google Scholar.

49. This is written 勿勿諸其欲其饗之 in the “Ji yi” 祭義 chapter of Li ji (Shisan jing zhushu ed., 47.1593a)Google Scholar, but 勿勿乎其欲其養之 in an otherwise parallel passage in the “Li qi” 禮器 chapter of the same work (24.1441c).

50. He also provides what he condemns as a “partially inadequate” rendering of the original passage by Franke, which still more ironically is superior to his own: for one thing, Franke has managed to identify the correct subject for the verbs ming and xian!