Hostname: page-component-586b7cd67f-rcrh6 Total loading time: 0 Render date: 2024-11-27T13:59:41.439Z Has data issue: false hasContentIssue false
  • English
  • Français

Some Observations on the Composition of the “Core Chapters” of the Mozi)

Published online by Cambridge University Press:  26 March 2015

Erik W. Maeder
Erik W. Maeder*
Affiliation:
rue Collet 5, 1800 Vevey, Switzerland
Get access Rights & Permissions [Opens in a new window]

Abstract

A.C. Graham understood the “core chapters” of the Mozi as the collected literary remains of three independently evolving traditions of early Mohism. The present study makes use of the remarkable coexistence of parallel texts in a single book to get a clearer picture of how such traditions may have been formed. Examining textual recurrences and “marked” lexical and functional terms in the core chapters, the author argues that the “Ten Theses” may not all have been parts of some “original” teaching by the “master” referred to, but documents attesting to the very process of transmission within still living, competing traditions.

These parts were, it is argued, summarily incorporated within three distinctive “documents” relatively late in the history of transmission, that is, just before or during early Han.

A.C. Graham 認爲墨子 • 十論是早期墨家三個各自獨立發展的學派存留下來的文字的結集. 本文試圈利用在同一著作中引人注目的相互獨立的文本共存的事實更淸晰的描述這些學派的形成過程. 在對十論中某些文句的重複出現, 以及帶有標記的詞彙和語法成分進行考査時, 本文作者指出十論很有可能幷非墨子本人的原著. 反之, 卻恰好表明墨子的成書是在不同學派競相演進中進行的. 作者還進一步指出這三個明顯不同的文本是在墨子成書較晚期, 也就是大約在漢前或漢初,未加鑑別的收入進來的.

Type
Articles
Information
Early China , Volume 17 , 1992 , pp. 27 - 82
Copyright
Copyright © Society for the Study of Early China 1992

Access options

Get access to the full version of this content by using one of the access options below. (Log in options will check for institutional or personal access. Content may require purchase if you do not have access.)

References

1. See Jiulong, Wu 吳九龍, Yincjueshan Hanjian shiwen 銀雀山漢簡釋文 (Beijing: Wenwu chubanshe, 1985)Google Scholar.

2. Yujiang, Wu 吳敏江, Mozi jiaozhu 墨子校注 (Chongqing: Duli chubanshe, 1944)Google Scholar, as reprinted in Mozi jicheng 墨子集成, ed. Lingfeng, Yan 嚴靈峰 (Taibei: Chengwen chubanshe, 1975), vols. 43–44Google Scholar, points to five such instances in the extant text of the Mozi. The list is given in Graham, Angus C., Later Mohist Logic, Ethics and Science (London: SOAS and Hong Kong: Chinese University Press, 1978), 89 Google Scholar.

3. See Tsuen-hsuin, Tsien, Written on Bamboo and Silk. The Beginnings of Chinese Books and Inscriptions, (Chicago: University of Chicago Press, 1962), 109 Google Scholar, on the usages of ce and pian. I will have something more to say on this topic below.

4. See Graham, A.C., Divisions in Early Mohism Reflected in the Core Chapters of Mo-tzu (Singapore: IEAP, 1985), 28 Google Scholar.

5. See in particular Forke, Alfred, Me Ti des Sozialethikers und seiner Schuier philosophische Werke (Berlin: 1922), 5–15 Google Scholar; Durrant, Stephen W., “An Examination of Textual and Grammatical Problems in Mo-tzu,” (Ph.D. diss.: University of Washington, 1975), 45–90 Google Scholar; and Graham, , Later Mohist Logic, 64–76 Google Scholar. The textual authorities are described in detail in the Appendix to Wu Yujiang, Mozi jiaozhu. The Daozang text is reprinted in Lingfeng, Yan, ed., Mozi jicheng, vol. 1 Google Scholar.

6. Graham, , Later Mohist Logic, 65 Google Scholar.

7. Graham, , Later Mohist Logic, 68, and 68 n. 94Google Scholar. The Yu manuscript is collated by Wu Yujiang in Mozi jiaozhu.

8. See Harvard-Yenching Institute Sinologicai Index Series. A Concordance to Mo Tzu (1948; reprint: San Francisco: Chinese Materials Center, 1974)Google Scholar.

9. See Graham, , Divisions in Early Mohism, 17 Google Scholar.

10. The Han Feizi quote is given in Graham, , Divisions in Early Mohism, 18 Google Scholar.

11. The “Luwen” 49/61–64 citation lists the authoritative statements or “theses” in an order slightly different from the one we find in the block of “Essays” (say, AB, EF, IJ, GH, and CD). The names of the “theses” are identical except in the cases of “Tianzhi” and “Minggui,” where each one of these two titles is replaced in turn by one of the two compounded terms of the expression zuntian/shigui 尊天/事鬼 (“Honor Heaven and serve the spirits”) found used three times independently in the “Dialogue” chapters (at “Gongmeng” 48/27 and 82, as well as in “Luwen” 49/3), once in chapter 4 from the “Epitomes” (4/18, with a negative form given at 4/20), once in “Essay” chapter 9 (“Shangxian” zhong 9/55, with a negative form given at 9/58), once in “Essay” chapter 35 (“Feiming” shang, with two parallel negative forms given in chapter 31” Minggui” at 31/84 and 89 respectively), and once in “Essay” chapter 26 (“Tianzhi” shang 26/25) albeit in a split form. (The parallel negative formula is given at 26/28). Forke, , Me Ti, 3–5 Google Scholar, gives several citations from ancient texts which show that the title of the “Minggui” chapter in particular was not well established.

12. See Graham, , Later Mohist Logic, 87–110 Google Scholar. Graham discussed the matter in somewhat more detail with regard to the transmitted text of the Gongsun Longzi 公孫龍子 in his “Two Dialogues in the ‘Kung-sun Lung tzu’,” Asia Major ns 11/2 (1965), 128–152 Google Scholar, reprinted with some changes in Graham, A.C., Studies in Chinese Philosophy and Philosophical Literature (Singapore: IEAP, 1986), 167–192 Google Scholar, under the title:” A First Reading of the ‘White Horse’.”

13. See Yates, Robin D.S., “The Mohists on Warfare: Technology, Technique, and Justification,” Journal of the American Academy of Religion 47/3 S (1979), 564–579 Google Scholar (“Thematic Issue: Studies in Classical Chinese Thought,” ed. by Henry Rosemont Jr. and Benjamin I. Schwartz).

14. Graham, , Later Mohist Logic, xi Google Scholar, commenting on Forke, Me Ti.

15. Graham, , Later Mohist Logic, xii Google Scholar.

16. Durrant, , “An Examination of Textual and Grammatical Problems,” 136 Google Scholar.

17. Forke, , Me Ti, 21–24 Google Scholar.

18. Forke, , Me Ti, 123–158 Google Scholar.

19. Forke, , Me Ti, 123–158 Google Scholar.

20. Karlgren, Bernhard, On the Authenticity and Nature of the Tso Chuan, Göteborgs Högskolas Aarsskrift 23:3 (Göteborg: Elanders Boktryckeri Aktiebolag, 1926)Google Scholar.

21. Durrant, , “An Examination of Textual and Grammatical Problems,” 2 Google Scholar.

22. Durrant, , “An Examination of Textual and Grammatical Problems,” 317–318 Google Scholar.

23. Durrant, , “An Examination of Textual and Grammatical Problems,” 315–316 Google Scholar.

24. Durrant, , “An Examination of Textual and Grammatical Problems,” 317–318 Google Scholar.

25. Graham, , Divisions in Eariy Mohism, 2 Google Scholar. This supposes that one accepts Graham's laying aside of the three parallel instances of hu clustered in a short corrupted passage found towards the beginning of chapter 8 (“Shangxian” shang 尙賢上 at 8/7).

26. Graham, , Divisions in Early Mohism, 4 Google Scholar.

27. Graham, , Divisions in Early Mohism, 3–4 Google Scholar. I will come back to this point below.

28. A remark made in parentheses in Graham, , Divisions in Early Mohism, 24 Google Scholar, on the use of weapons reveals that Graham at one time considered chapter 20, finally treated as a “Digest,” to be a possible candidate for inclusion in the J series.

29. See Graham, , Divisions in Early Mohism, 18–28 Google Scholar, and Graham, A.C., Disputers of the Tao. Pholosophical Argument in Ancient China (La Salle, Ill.: Open Court, 1989), 36 Google Scholar.

30. I include for comparison Forke's tabulation, as found in Me Ti, 23. It should be noted that Forke made a slight error in his tabulation: chapter 20, twenty lines long in the Harvard-Yenching concordanced text, should actually be placed below chapter 21, which is only nineteen lines long in the concordance. But the tabulation as it is allows for a better comparison with Graham's.

It will be noticed that Forke chose to include the present chapter 6 from the “Epitomes” in the slot left vacant by “missing” chapter 22. I will have more to say on this below.

31. Graham, , Divisions in Early Mohism, 10–12 Google Scholar. As Graham noted, however, the longer the “word-combination,” the greater the danger of textual corruption, and one has to be all the more strict about what he terms “verbal identity.” Even he could not quite respect his own rule, it seems, and instances of corruption he discounted as unmeaningful in his discrimination among parallel sentences or formulae found at both ends of a chapter within any one of his predefined series are not, in my view, in all cases very different from the ones found in cross-series parallels.

32. Durrant himself remarked on this point in an article Graham happened to stumble upon as he was finishing an article of his own on the uses of the particle hu in pre-Han texts: Durrant, S.W., “A Consideration of Differences in the Grammar of the Mo Tzu ‘Essays’ and ‘Dialogues’,” Monumenta Serica 33 (1977–1978), 249–250 CrossRefGoogle Scholar. The reference to this article by Durrant and to Graham, A.C., “A Post-Verbal Aspectual Particle in Classical Chinese: the Supposed Preposition HU (乎),” Bulletin of the School of Oriental and African Studies 41/2 (1978)CrossRefGoogle Scholar is given in Graham, , Divisions in Early Mohism, 2 n. 5Google Scholar.

33. The word “document” is used by Graham at the end of his study to mean the conflation of the various chapters in a series; see Graham, , Divisions in Early Mohism, 28 Google Scholar.

34. Graham, , Divisions in Early Mohism, 1, and 28 Google Scholar.

35. Compare Graham, , Divisions in Early Mohism, 7 Google Scholar, where the reference given at the top of the page is, I think, more correctly, 35/18–35; and ibid. 14–17, where the reference is consistently given as 35/18–33. The confusion comes from Graham's treatment of the matter in the last paragraph on page 14: 35/28–31 and 35/31–34 are two strictly parallel blocks, both carrying the H mark guanfu, the only difference coming from a shift from positive to negative and negative to positive signs. 35/34–35 does not carry any series criteria, but it could itself be treated as coming from H since it has correspondences with chapter 16 H (at 16/3 and 16/84).

36. Durrant, , “An Examination of Textual and Grammatical Problems,” 167–177 Google Scholar.

37. Watson, Burton, transl., Mo Tzu. Basic Writings (New York: Columbia University Press, 1963), 37 Google Scholar. I basically follow Watson's rendering of the phrase as it is found in chapter 11 Y since I take the chapter 12 H parallel to be partially corrupted: as we will see later, the whole passage in chapter 12 H is actually badly damaged and shows traces of interference with parallel passages found in another H chapter, chapter 27, “Tianzhi” zhong, at 27/28–30 and 27/9 respectively. Other examples of hu/yu inversions are to be found at: 9/49 H paralleling 10/24 J; 9/73 H paralleling 12/68 H; 37/8 H paralleling 35/11 Y; and 32/S H paralleling 6/30 from the “Epitomes.”

38. Compare Watson, , Mo Tzu, 66 Google Scholar. See also Yi-pao, Mei, The Ethical and Politicai Works of Motse (London: Probsthain, 1929; reprint: Westport, Conn.: Hyperion Press, 1973), 124 Google Scholar. 16/41 has jun 君 here; the Daozang text has a corresponding blank in 16/28, which has been filled out by Sun Yirang with shi 士 to accord with the context.

39. Compare Mei, , The Works of Motse, 67–68 Google Scholar.

40. Compare Watson, , Mo Tzu, 91 Google Scholar, and Mei, , The Works of Motse, 148 Google Scholar.

41. Graham, , Later Mohist Logic, 15–25 Google Scholar.

42. Durrant, , “An Examination of Textual and Grammatical Problems,” 315–316 Google Scholar.

43. Graham, , Divisions in Early Mohism, 8–10 Google Scholar.

44. One of the two occurrences of tiangui in J is parallel: see 10/48 J. The compounding of the two forms is also found repeated twice in the “Dialogues” within two closely parallel sentences clustered at 47/16.

45. As one can see from the text of the quotations given below, the compound guanfu, “government bureaus,” once replaces in chapter 37 H (37/33) the full phrase canglin fuku 倉魔庫府 (“granaries and treasuries”) which is given in parallel in chapter 32 H at 32/36, This shift is, however, exceptional.

46. Graham, , Divisions in Early Mohism, 1 Google Scholar.

47. See Watson, , Mo Tzu, 114–115 Google Scholar. Compare Mei, , The Works of Motse, 179 Google Scholar. For the versions in chapters 9 H and 25 H, see further Watson, 23–24 and 68; Mei, 37 and 126. The chapter 37 H version may be checked in Mei, 198–199.

48. Chapter 25 H here has a passage original to it which deals with artisans.

49. The textual affinities between” Feiming” chapter 37 H and “Feiyue” chapter 32 H by contrast are more ancilliary. The “Feiming” texts themselves do share with “Feigong” chapter 19 H what appears to be a common but otherwise exclusive concern for the not inconsequential notion of the “mandate of Heaven” or tianming 天命, but they do not otherwise directly connect with the “Jieyong” problématique. The bound form of the compound tianming appears at 19/33, 41, 42, 44 H as well as at 37/23, 24 H; 36/25, 29 J; and 35/40 Y. The unbound form is given at 19/37, 38, 39, 41 H as well as at 36/28, 30 J; and 35/45 Y. The compound is also found at 39/10 “Fei Ru.”

50. Graham, , Divisions in Early Mohism, 28 Google Scholar.

51. See Forke, , Me Ti, 24 and 136–138 Google Scholar; and Huanbiao, Wang 王煥鑣, Mozi jiaoshi 墨子校釋 (Hangzhou: Zhejiang Wenyi chubanshe, 1984), 176–186 Google Scholar.

52. The introductory formula is partially parallel to both 11/1 Y and 12/1 H.

53. See Mei, , The Works of Motse, 122 Google Scholar.

54. These two broken instances run simply: guzhe shengwang wei (…) 古者聖王爲, and may at one time have connected with the partial wei X X zhifa (yue) 爲 X X 之法 (曰) found at 6/2 and 6/11 in chapter 6. The whole section 21/9–13 Y in which the two instances of guzhe shengwang wei appear is itself structured around a recurring pattern in the form ci X zhi li ye 此X 之利也, “this is the benefit of X,” which appears five times in this section but nowhere else in the chapter. The section itself has a “digest” air to it. Slightly varying forms of each one of the formulae typical, respectively, of chapter 6 and 21 Y appear in turn in “Digest” chapter 20, at 20/6 and 20/11 respectively.

55. The text being introduced by the formula itself contains definitive traces consistent with its place in Y chapter 21, such as the use of zuyi 足以instead of keyi 可以, as found, for example, in chapter 6. See the note below.

56. See Forke, , Me Ti, 4–5 Google Scholar, referring to Wei Zheng's 魏徵 (A.D. 581–643) Qunshu zhiyao 群書治要. The parallel is at 5/19 = 6/8, 14, where chapter 5's bu keyi 不可以 is given as bu zuyi 不足以. We have seen in the extended quotation above the same type of variation occurring in passages where chapter 21 Y (keyi) runs parallel to chapter 6 (once again: zuyi).

57. The other chapters from the “Epitomes” are more difficult to place in terms of chronology. Their language most often resembles that of chapter 5, for example, or that of the stretches of unpatterned text which are found tucked in between the neatly patterned sentences which characterize the full, “Essays” as well as the shorter “digests.” As with chapter 4, one does find in chapter 3 (“Suo ran” 染), for example, scattered instances of parallelisms with the “Essays”: 3/10–11 is thus very close textually to 9/58 H (“Shangxian” zhong), with a sentence which is now found in the “Dialogues” at 49/7 (“Luwen”) helping us to fill in a break in the parallel. This reminds us that scattered instances of parallelisms with the “Essays” are also present within the “Dialogue” chapters. As their generic title implies, the chapters in this block are predominantly made up of short ce recording verbal exchanges between the “Master” Mozi and a number of his disciples or competing intellectuals, as well as a number of other personages including lords of states or principalities, in a manner which is reminiscent of the Confucian Analects or the Mencius. These exchanges are usually very short, only a few ever attaining the length of “Epitome” chapter 7 (“San bian”三辯), which is ten lines long in the concordanced text and also happens to be set in dialogue form. (Chapter 7 is alone among the “Epitomes” to use the dialogue form, and has indeed often been considered as a stray fragment from the “Dialogues” block). But the “Dialogue” part of the Mozi is not itself wholly made up of dialogues; it also in ־ eludes a fair number of short ce which are expository or argumentative in style and are thus hardly distinguishable from what is now found within the block of “Essays.” Close textual parallels with the “Essays” are nonetheless exceptional. Because they share this characteristic with Graham's” Digests,” I would tend to treat the “Dialogues”, as of later general composition than the “Essays,” thereby agreeing with Durrant, albeit on different grounds. (See my treatment of Durrant above. The reference is to Durrant “An Examination of Textual and Grammatical Problems,” 317.) This is said over a possible objection. Among the clear textual parallels between the “Dialogues” and the “Essays,” there is indeed one which appears to be in much better condition where it is found in the “Dialogues” than at its place within the block of “Essays.” This sentence, which describes the mores of some barbarian population on the fringes of civilization, runs smoothly at 49/27 (“Luwen”), but is more manifestly sundered at 25/78, 75 H (“Jiezang,” following the order of the sentence as given at 49/27). It is in my opinion characteristic of the “Essays” to work around what disruptions may have occurred in a way which often looks forced and rather awkward. I take it as a further sign of the probable anteriority of the “Essays” in relation to the “Dialogues”.

58. Graham, , Divisions in Early Mohism, 1 Google Scholar. The other instance has already been noted: this is where chapter 37 H from “Feiming” runs parallel to the whole of chapter 36 J for its first twenty-nine lines before running parallel to 32/34–42 H at 37/30–41 H.

59. These are, respectively: 13/8–9 J = 11/1–2 Y = 12/1–2 H; and 13/11–12 J = 11/5–7 Y = 12/6–9 H.

60. The parallel this time is “Shangxian”: 11/3–4 Y = 12/3 H = 10/42 J. These are but a few of the many occurrences in the block of “Essays”, where the smooth running of the text is more or less abruptly interrupted by a passage which has its equivalent in another chapter, even a chapter from quite another triad.

61. The parallel in chapter 35 Y (“Teiming”) cannot escape notice.

62. See Watson, , Mo Tzu, 34–35 Google Scholar. I have changed only the last line.

63. The phrase is whole only in the treatments of the district and state levels. Because of manifest disruption, it is almost non-existent in the treatments of the village and world levels at the beginning and end of the section.

64. Compare Mei, , The Works of Motse, 61–62 Google Scholar.

65. See Watson, , Mo Tzu, 36–37 Google Scholar.

66. See Mei, , The Works of Motse, 74 Google Scholar.

67. See Watson, , Mo Tzu, 35 Google Scholar.

68. See Watson, , Mo Tzu, 36 Google Scholar.

69. These are found at 12/49 H = 13/16 J; 12/54–55, 59 H = 13/21 J; 12/63–64 H = 13/51–53 J; and 13/66–68 H = 13/48–49 J. There are two other traces in this part of a possible interaction with chapter 9 H.

70. Forke, , Me Ti, 24 Google Scholar.

71. Forke, , Me Ti, 133–136 Google Scholar.

72. Graham, , Divisions in Early Mohism, 3–4 Google Scholar.

73. See Watson, , Mo Tzu, 51 Google Scholar.

74. See Mei, , The Works of Motse, 158 Google Scholar.

75. See Mei, , The Works of Motse, 159 Google Scholar.

76. The shorter instances of parallelism with both chapters 17 and 19 H from “Feigong” are at 28/58 J = 17/3 and 5; and at 28/50, 51, 54 J = 19/30, 48, and 50 H, where the phrase is a formula repeated verbatim in both chapters. A clear connection with a member of its own triad appears at 28/69–70 J with a quote from a “Book of the Former Kings” (xianwang zhi shu 先王之書), which is also given at 27/52 H.

77. See Graham, , Divisions in Early Mohism, 14–15 Google Scholar. That this passage, 28/45 or 47 to 68, is somewhat heterogeneous within chapter 28 J may be confirmed on yet another count. Graham addressed the interesting issue of the assumed “Audiences of the Three Series.” Noting that “the difference in audience shows up most clearly in the concluding formulae,” he remarked that “in H the appeals are to ‘kings, dukes, great men and officer gentlemen’ (wang gong daren shi junzi 王公大人士窘子), or, in the case of the more private issues, mourning, music, fatalism, simply to the ‘officer gentlemen.’ The officer gentlemen are described in ch. 32/34–40 H as those who hold offices and collect taxes, in contrast with the kings, dukes, and great men who attend court.” Graham also noted that “in Y on the other hand the concluding appeals are solely to the officer gentlemen, and these are identified, not as holders of office, but as … rival thinkers ….” As for J, it “addresses both audiences. The concluding formulae of his ‘T'ien chih’ and ‘Fei ming’ chapters appeal only to the officer gentlemen, whom like Y he identifies with his opponents on the issues of the will of Heaven and fatalism …. In the two political chapters on the other hand the appeal is to ‘kings, dukes and great men,’ whom like H he repeatedly specifies as the ‘kings, dukes and great men of today’”; Graham, , Divisions, 18–20 Google Scholar. If one considers all instances of both wang gong daren and shi junzi, one notes that, whereas H is indeed mostly characterized by its repeated appeals to the” kings, dukes and great men of today,” the matter is less clear with regard to the other series. The most numerous instances of the wang gong daren formula are to be found for both H and J in the “more political” “Shangxian” triad: chapter 9 H has a total of 13 instances of the formula, but chapter 10 J has even more instances of it: 19 instances in all, if we include the “concluding formula” which conflates wang gong daren and shi junzi. With only two exceptions, the shi junzi formula is regularly part of a longer formula which reads: jin tianxia zhi shi junzi 今天下之士君子. There are six instances of it in chapter 10 J (including the “conciuding formula”), and just as many in chapter 28 J. This is the highest number one gets of instances of the shi junzi formula within a single chapter. The next highest number is in chapter 26 Y, also from the “Tianzhi” triad: the total there is five instances. (Chapter 15 Y has three). Now, chapter 27 H from “Tianzhi” has no instance of the shi junzi formula, but two of the rival wang gong daren formula. Chapter 28 J also happens to have three instances of the wang gong daren formula, and these are all to be found concentrated in the section I have just discussed, at 28/59, 61, and 67. By contrast, the six instances of shi junzi are in the sections of chapter 28 J which run parallel to either chapter 26 Y or 27 H from the same triad: see 28/1, 18, 34, 45, and 46 (which would indicate that one should actually see the heterogeneous section starting at 28/47 instead of 28/45), as well as 28/71. Chapter 26 Y has no instance of the wang gong daren formula; chapter 19 H has three, including the “concluding formula” in which it is conflated with the single instance of shi junzi. Chapter 18 Y has two, but no instance of shi junzi.

78. See Mei, , The Works of Motes, 157 Google Scholar.

79. See Watson, , Mo Tzu, 50–51 Google Scholar.

80. Compare Watson, , Mo Tzu, 52 Google Scholar, and Mei, , The Works of Motse, 107 Google Scholar.

81. See Watson, , Mo Tzu, 92 Google Scholar.

82. This passage has a number of parallels in common with chapter 9 H from the” Shangxian” triad (35/22, 25–26 = 9/12–13 H; 35/31–32 = 9/28–29 H), and two with chapter 11 Y from the “Shangtong” triad (35/30 = 11/10 Y; 35/32 = 11/12 Y).

83. Graham, , Divisions in Eariy Mohism, 12–13 Google Scholar.

84. Graham, , Divisions in Early Mohism, 13–14 Google Scholar.

85. The difference between Graham's” simple” and more “complicated” solutions comes down to whether one chooses to leave “the sections on the sages' practice” (test number one, shengwang zhi shi 聖王之事, chapters 35/10–12, 36/9–13) “in the same chapters as before.” See Graham, , Divisions in Early Mohism, 14 Google Scholar.

86. Graham, , Divisions in Early Mohism, 13 and 6Google Scholar.

87. Graham, , Divisions in Early Mohism, 28 Google Scholar.

88. See Watson, , Mo tzu, 119 Google Scholar.

89. Compare Mei, , The Works of Motse, 197 Google Scholar.

90. The close parallels are: 37/21 H = 36/23–24 J; 37/23–24 H = 36/24–25 J = 35/39 Y; 37/25 H = 36/25 J = 35/40 Y; and 37/28 H = 36/27 J = 35/42 Y.

91. A close parallel with a sentence in a chapter from quite another triad is also found at this point, always a sure sign of disruption: 37/29 H = 27/67–68 H.

92. See 36/3 J. The awkwardness of test number two in chapter 36 J comes from the fact that test one, the “practice of the Sage Kings,” would seem to call by itself for some sort of consultation of the “Books of the Former Kings.” As we have seen, the “Books” are in fact abundantly referred to in both J and H, as well as in Y as far as the “Feiming” triad is concerned; see 35/39–42 Y. The presentation of the first two tests is announced by the formula: yu cji ben/yuan zhi ye 於其本/原之也, and test number two is presented thus: zheng yi xianwang zhi shu 徵以先王之書. This is one of only two occurrences of zheng in the whole Mozi corpus. The announcement of the third test, which follows immediately, is out of pattern: yong zhi naihe 用之奈1可, instead of the expected yu cji yong zhi ye.

93. Another textual basis for judging J's relative “conservatism” is chapter 13 J's mention of the ‘family” or clan (jia 家), in the patterned sets original to it at 13/22–29, instead of the village (li ) which is mentioned in parallel in both Y and H (11/13 Y and 12/18 H); see Graham, , Divisions in Early Mohism, 23 Google Scholar. The text is intact here, and the elements of proof consequently more firm. Another difference with both Y and H in the “Shangtong” triad is that J seems to have, within those same patterned passages, the “lords” (jun 君) of family, state, and empire call upon their subjects to conform, not to the power above them as is clearly stated in Y and H, but to themselves. (See Graham, , Divisions, 23 Google Scholar). But the matter is already less coherent than it seems at first sight: the lords of family, state, and empire are also told within the same sets to “organize the purposes of their family / state / empire and identify them with those of the state / emperor / and Heaven” (13/29, 36, 42 J). And when much of the same matter is brought up again in the context of the “Tianzhi” (“Will of Heaven”) triad, the passage in chapter 28 J which says that “the gentlemen of the world all understand that the emperor sets the standard for the world but do not understand that Heaven sets the standard for the emperor” (jin tianxia zhi shi junzi jie ming yu tianzi zhi zheng tianxia ye, er bu ming yu tian [zhi] zheng [tianzi] 今天下之士君子皆明於天子之正关下也。而不明於天 [之]正 [天丰]也⃘) the wording of the passage looks suspiciously close to that of a comparable passage found at the end of “Shangtong” chapter 11 Y. (See 28/13 J = 11/22 Y. The the Daozang text lacks both zhi and tianzi, but this does not seem to affect the integrity of the passage in any important way).

94. See Graham, , Divisions in Early Mohism, 14 Google Scholar.

95. The parallel stops at this point.

96. See Graham, , Divisions in Early Mohism, 16 Google Scholar. Chapter 37 H presently lacks the illustration for the “ears and eyes” test; Graham saw it as a lacuna in the text.

97. Graham, , Divisions in Early Mohism, 28 Google Scholar.