Skip to main content Accessibility help

A neglected pre-Han philosophical text: Ho-kuan-tzu

  • A. C. Graham


The book which passes under the name of Ho-kuan-tzu, ‘Master of the pheasant cap’, an otherwise unknown teacher supposedly of the third century B.C., expounds a variety of syncretism which roots the fa decreed by the ruler in a Yin-Yang cosmological scheme. It is remarkable for presenting no less than three contradictory programmes for ideal government, two of them described at length and ascribed to sage emperors missing in orthodox historiography; one is roughly classifiable as Legalist, one as Yin-Yang, the third as Taoist anarchist. It attracted little attention until recently because of lack of evidence as to its date. But with the discovery in 1973 of the Ma-wang-tui manuscripts of Laotzu it was noticed that the ‘Yellow Emperor’ documents attached to manuscript B are closely related to it; and Wu Kuang has since pointed out that the first two chapters observe the taboo on the personal name of the Ch'in First Emperor.



Hide All

1 Wu Kuang 157. Cf. p.506 below.

2 Chang Chin-ch'hng and Neugebauer as in finding list.

3 cf. footnote 36, below. Ho-kuan-tzu ch. 11 is reproduced complete without commentary in a surviving volume of Yung-lo ta-tien (ch. 19,743), from the Yung-lo period (1403–1424), so earlier than the Ming Taoist Patrology. It preserves a passage of 222 characters which has dropped out before at (ch. 11) 81/3 but is restored in the modern edition of Chang Chin-ch'hng. The Taoist Patrology leaves two blanks for it; the SPTK preserves its first chapter but wrongly adds . This implies that the SPTK text had an exemplar less defaced than the Taoist Patrology, but also that the process of filling gaps conjecturally had already begun. Earlier in the chapter, at (ch. 11) 76/9, the Yung-lo ta-tien agrees with the Taoist Patrology in omitting the of the SPTK, but unlike it, does not leave blanks for the two characters. Here, as Chang notices, the SPTK characters fit badly and look as though they are supplied from the commentary.

4 The restorations are in the Ho-kuan-tzu p'png-chu , reproduced in the Shih-tzu ch'hian-shu. This has notes by Wang Yüs, one of which on p. 6 acknowledges the first restoration as his own. Sometimes one graph is restored where the SPTK edition has two blank spaces; then one space is still left blank. The Tzu hui edition of 1577 eliminates these blanks (although not those where no restoration was made), and seems therefore to be later, in spite of the fact that Yen Ling-feng (v. 2, 397) records no edition of Wang Yüs corrected text before 1625. I have failed to find evidence of the date of this Wang Yüi , style Yung-chün , of Fukien.

5 Fu, op. cit. The manuscript may be presumed lost with the fall of the Kuo-min-tang regime. In mainland China the Tun-huang yi-shu tsung-mu so-yin, 328 lists it among dispersed manuscripts; in Taiwan Yen Ling-feng (v. 2, 391f.) notes only that it was formerly in Fu Tseng-hsiang'g collection. A Tun-huang manuscript of the whole book has also been reported, but is a bibliographical ghost springing from a hasty reading of Tun-huang yi-shu tsung-mu so-yin, 486 (Neugebauer, 1, 7, 11).

6 In the table of contents reproduced by Kao Ssu-sun (chin-shih 1184), mu 10B/9.

7 The figure is ‘16’ in Han Yii'i note as it stands at the head of Lu Tien'n Ho-kuan-tzu (also as cited by Ch'hn Chen-sun 280), and Lu Tien himself notes (preface 2/3) that it implies that Han Yii had an incomplete text. The figure has been corrected to ‘19’ in Han Yii'i collected works (Han Ch'hng-li ch'han-chi, (ch. 11) 283).

8 Ch'hng-wen tsung-mu, 136. Bruce Williams, whose thesis became available to me only at proof stage, gives a detailed account of the text history (Williams, 37–43, 72–113). He has a very persuasive explanation of the different numbers of chapters given by T'Tng and later authorities. There is firm testimony that the 15-chapter version covered chs. 1–15 (Williams, 173f., nn. 294–96). The 16-chapter version mentioned by Han Yü seems to have belonged to a compilation of Mo-tzu chs. 1–13, the 19-chapter Ho-kuan-tzu and 19unknown essays mentioning Han and post-Han events, current during the T'ang and Sung (Williams, 87–9). Assuming that the divisions between sections were unmarked, readers could easily suppose that the Ho-kuan-tzu section ended with the last reference to Hokuan-tzu (ch. 15) or the last to P'ang Hsüan (ch. 16). The 19-chapter version now current is unlikely to be an extract from this compilation, the attestation of which is much later than the earliest quotations from chs. 16–19 (listed Williams, 256).

9 Needham, n, 547.

10 Wu Kuang, 154–6.

11 cf. p.499 above.

12 For this idiom cf. Li chi SPPY (Van kung B) 3/8A ‘There is acting directly on one'e genuine feelings, which is the way of the barbarian’.

13 Neugebauer, 23–37.

14 Neugebauer, 273–5.

15 Kuan-tzu, BSS (ch. 2) 1/5, annotated (ch. 64) 3/33. Huai-nan hung-lieh chi-chieh (ch. 6) 6/5A.

16 Kuan-tzu, (ch. 55) 3/14.

17 Neugebauer, 29.

18 cf. Daor, 44ff.

19 For the views of these three scholars, see Chang Hsin-ch'hng, 861–5.

20 cf. p.502 above.

21 ‘The date and composition of Lieh-tzu’, in G., Studies, 216–82.

22 Neugebauer, 121.

23 cf. ‘The origins of the legend of Lao Tan’, G., Studies, 111–24.

24 Hou Han shu (chih 30) 3670. Cf. Ch'hen Mu, 485.

25 Shih chi, (ch. 34) 1560, (ch. 43) 1831, (ch. 81) 2450.

26 cf. p.500 above.

27 Shih chi, (ch. 16) 768.

28 Wu Kuang, 157. Three chapters of Han Fei tzu (chs. 6, 16, 20) observe the same taboo (Lau, 236–240).

29 cf. Beck, op. cit. Lau holds that the taboo as observed under the First Emperor Lau, 222, 224). Williams does not discuss the taboo, but turns out to have been ahead of ne in noticing the overlooked Ma-wang-tui parallel to (ch. 2) 4/4 (Williams, p. 217, n. 43).

30 cf. G., Studies, Yin-Yang, 84–8.

31 Shih chi, (ch. 6) 237f.

32 Blank graph.

33 Chang Hsin-ch'hng, 865–7.

34 Chingfa, 167–89.

35 cf. p.507 above.

36 Here the graph is blank in the Taoist Patrology but present in the SPTK and the later editions. It is confirmed by the parallel in the Ma-wang-tui manuscript, where the graph is clearly legible in the facsimile (Ma-wang-tui Han mu pai shu, vol. 1, in column 91B, p. 7). This is important in showing that the text is less defaced in the SPTK than in the Taoist Patrology (cf. n. 3 above).

37 Ku shih pien, vo. 1, 105–50.

38 Forke, 528–37.

39 Shih tzu, SPPY A. 6A/8, 6B/5, 8A/7, 14B/6.

40 L'-shih ch'hn-ch'hu chi-shih, (ch. 5/2) 5/6B.

41 cf. (ch. 9) 48/10, (ch. 11) 77/1.

42 This obscure sentence should perhaps be read in the light of Chuang-tzu'u ‘Treat even the not so as so’ (Chuang-tzu HY 2/90) and the doctrine criticised in the Canons in the form ‘There is no victor in disputation’ (Mo-tzu HY 41/23). Ho-kuan-tzu holds (unlike Chuang-tzu) that the sage does distinguish between right and wrong alternatives, but also that Heaven does not.

43 cf. G., Disputers, Appendix 1.

45 cf. (ch. 3) 6/10f. ‘In reward and punishment it is the covenant’.

46 Hall and Ames, 11–21.

47 Needham, II, 547.

48 Cited p.5I2 above.

49 The second alternative would be allowable only on the assumption that the author was in a rebel area in allegiance to Ch'h after 209 B.C but returned to the shrinking Ch'hn area before 207 B.C.

50 Kuo yü (Ch'h yü) 231–8, also Kuan-tzu, ch. 20. The parallels are listed in Neugebauer, 261–4.

51 ‘Using punishment to get rid of punishment’ , Shang-chun shu chu-yi (ch. 13) 108, Han Fei tzu chi-shih, (ch. 53) 1123.

52 For the ch'hng of X as its essentials for being called ‘X’, cf. ‘The Mencian theory of human nature’, G., Studies, 59–65.

53 cf. pp.518 above.

54 cf. Hansen, 30–54.

55 The missing graph is blank in Taoist Patrology and SPTK but easily deducible from the context and restored in later editions.

56 cf. pp.507 512 above.

57 cf. p.507 above.

58 Chuang–tzu, HY 7/31, which I formerly translated (G., Chuang-tzu 98)‘Don'n be a medium possessed by your name...Don'n be the man-in-charge of wisdom’. The sub-commentary of Ch'hng Hsiian-ying (fl. 631–50) explains shih (the impersonator of the dead man at a funeral) as chu ‘master, host’ here and also in Chuang-tzu, HY 1/24, 33/1, where Yao as emperor and the head of the Mohist school are both called shih. The Ho-kuan-tzu examples force one to reconsider the Chuang-tzu passage; but although the re-translation ‘Don'n be the authority for names’ fits the context well, the examples are too sparse for a confident judgement. The ming shih of Huai-nan hung-lieh chi-chieh (ch. 14) 14/1B is in a borrowing of the Chuang-tzu passage.

59 Ho-kuan-lzu, (ch. 11) 76/2. Lu Tien, however, cites a variant wu ch'h for yü an ch'h. The other possible earliest example of üuan ch'h (Huai-nan hung-lieh chi-chieh (ch. 3) 1A/9 cf. Liu Wen-tien'n note) is also doubtful textually.

60 Shih-chi, (ch. 6) 236.

61 Shih-chi, (ch. 12) 473. For T'Ti-yi, T'Ti-huang and the Nine Majesties, cf. Ku shih pien, vol. 7, A 175ff., B 50ff.

62 Mencius, 5A/6. For other examples of the questioning of hereditary rulership, cf. G., Disputers, Part 3/4.

63 Ogata, 64, n. 18.

64 cf. p.501 above.

65 Chuang-tzu, HY 9/8, 10/32.

66 Chuang-tzu, HY 11/119, 29/31.

67 Teng Hsi tzu (Tzu hui edition), 7A/8–7B/6.

68 cf. ‘How much of Chuang-tzu did Chuang-tzu write?’, G., Studies, 305–307, 308.

69 cf. p.500 above.

70 Ho-kuan-tzu, (ch. 1) 2/4. Cf. p.499 above.

71 Chan-kuo ts's, (Yen 1) BSS 3/62/2.

72 cf. p.506 above

73 Paul Thompson pointed out to me after the submission of the manuscript that my Block B could well be the latest, from the beginning of Han. We would then have a neat dialectical development, first the semi-Legalist Utopia, then disillusionment with all government during the interregnum, then recovery of faith in government after the Han reunification and the proposal of a new and more humane Utopia.

I wish to thank Michael Loewe and Carine Defoort for their help in finding my way round the bibliography of Ho-kuan-tzu.

A neglected pre-Han philosophical text: Ho-kuan-tzu

  • A. C. Graham


Full text views

Total number of HTML views: 0
Total number of PDF views: 0 *
Loading metrics...

Abstract views

Total abstract views: 0 *
Loading metrics...

* Views captured on Cambridge Core between <date>. This data will be updated every 24 hours.

Usage data cannot currently be displayed