¹ For an engaging, and at the same time peropulently detailed, biography of Stein see Mirsky, Jeannette, Sir Aurel Stein, archeological explorer, Chicago and London, University of Chicago Press, 1977Google Scholar. The account of Stein's discovery of the Tun Huang cache is given in ch. xiv, pp. 254–80.

² Ching tien shih wen l/33a, Ssu-k'u shan pen ts'ung shu ed., Taipei, I Wen reprint, n.d. In fact Lu enters the commentary under the name Hsiang Yii, writing yÜ erroneously for erh (i.e. ). This would suggest that he had never actually seen the text, and raises the question of its existence at this time. This is the first known mention of the Hsiang erh chu in any bibliographic record. Neither the Han shu, I wen chih nor the Sui shu, Ching chi chih lists it.

³ Encylopedia Britannica (15th ed.), 1975: ‘Taosim, history of’ (M. Strickmann); and Michel, Strickmann, ‘On the alchemy of T'ao Hung-ching’, in Welch, H. and Seidel, A. (ed.), Facets of Taoism, New Haven, Yale University Press, 1979, 164–7.Google Scholar

⁴ There are no second-century texts extant as far as has been determined, except for the possibility that some fragments have been preserved in the later writings of Shang ch'ing Taoists such as T'ao Hung-ching (A.D. 456–536), who cites, for example, the Ch'ien erh po kuan i , identified by Strickmann as a second or early third-century Celestial Master document from the Szechuan/Han-chung area. See Strickmann, Michel, Étude sur le Taoisme du Mao chan, Ph.D. thesis (Paris), 1979, 56Google Scholar.

⁵ San kuo chih 8 (Wei shu 2), Peking: Chung hua shu chü ed., 1959, 263–5. See also de Groot, J. J. M.,‘On the origin of the Taoist Church’, Transactions of the third International Congress for the History of Religions, I, Oxford, 1908, 138–49Google Scholar, esp. 139. Essentially the same thing is repeated in the Hou Han shu where the works Chang Lu allegedly compiled are called fu shu ‘talismans and writings’ (or ‘talismanic writings’); see Wang, Hsien-ch'ien, Hou Han shu chi chieh 75, Taipei, Commerical Press, 1968, 2668Google Scholar. The Shen hsien chuan , attributed to Ko Hung (A.D. 283–343) but actually a much later text, records that Chang Lu wrote twenty-four pien worth of Taoist tracts (Han Wei ts'ung shu ed. 4/lla).

⁶ de Groot 1908, passim; and Levy, Howard S., ‘Yellow Turban religion and rebellion at the end of the Han’, JAOS, LXVI, 4, 1956, 214–27Google Scholar.

⁷ Strickmann ‘On the alchemy of T'ao Hung-ching’, 167.

⁸ Tsung-i, Jao, A study on Chang Tao-ling's Hsiang-er commentary of Tao Te Ching, Hong Kong, Tong Nam Printers and Publishers, 1956Google Scholar ( ); and ‘Lao tzu Hsiang erh chu hsu lun’ in Fukui hakase shoju kinen Tōyō bunka ronshū, Tokyo. Waseda University Press, 1969, 1155–71.Google Scholar

⁹ Shih-hsiang, Ch'en , ‘Hsiang erh' Lao tzu Tao ching Tun huang ts'an chüan lun cheng’ Tsing Hua Journal of Chinese Studies I, 2, 1975, 41–62.Google Scholar

¹⁰ Ying-shih, Yü, ‘Life and immortality in the mind of Han China’, HJAS, XXV, 1964–1965, 80–122.Google Scholar

¹¹ Robert G. Henricks points out that the claim that the characters of an Emperor's name became ‘taboo’ when he first assumed the highest position is only one of two possibilities. It has also been claimed that the ‘taboo’ did not take effect until the Emperor's death, in which case the date of the A manuscript could be as late as 194 B.C. See R. G. Henricks, ‘Examining the Ma-wang-tui silk texts of the Lao tzu’, T'oung Pao, LXV, 4–5,1980, 166–99; esp. n. 3, p. 167.

¹² Further introductory remarks and details on the silk manuscripts of the Lao tzu can be found in these articles: Kao Heng and Ch'ih Hsi-chao ‘Shih t'an ma wang tui Han mu chung ti po shu Lao tzu’ (Discussion of the silk manuscripts of the Lao tzu from the Han Ma wang tui tombs), Wen Wu 11, 1974, 1–7; Jan, Yü-hwa, ‘The silk manuscripts on Taoism’, T'oung Pao, LXIII, 1, 1978, 65–84Google Scholar; and Henricks 1980.

¹³ The point has been argued convincingly by Ma Hsü-lun in his preface (hsü ) to Lao tzu chiao ku ( ), Hong Kong, 1973 (2nd ed.), and by Pelliot, P., ‘Autour ďune traduction sanscrite du Tao tö king’, T'oung Pao, XIII, 1912, 366–70Google Scholar. Ma bases his con clusion both on the absence of any bibliographic record of the Ho Shang-kung commentary before the Sui, and on the content of the commentary itself, which deals with cosmic-somatic correlations of a kind most likely unknown in the Former Han, even less in pre-Han times. Ma also suggests the possibility that the Ho Shang-kung commentary is later than the Wang Pi commentary because it includes notes to chapter 31, for which a Wang Pi commentary is missing. Prior to the discovery of the Ma-wang-tui silk manuscripts it was common to explain the lack of a Wang Pi commentary to this chapter (and to chapter 66 as well) by saying that the lost commentary had become irretrievably integrated into the basic text. Ma's observation was that since Ho Shang-kung annotated chapters 31 and 66 which ostensibly embodied Wang Pi's commentary somewhere within, his must have been a later commentary. This argument falls when we compare these two chapters with the MWT manuscripts and find them essentially unchanged from the early Han versions. This means Wang Pi's missing commentary does not constitute any part of them, and we can no longer claim that Ho Shang-kung was annotating passages that included anything by Wang Pi. The basic conclusion (on other grounds) that Ho Shang-kung does not predate the end of the Han dynasty still stands in spite of this one correction to Ma's analysis. See also Erkes, E., Ho-Shang-kung's Commentary on Lao tzu, Ascona (Switzerland), Artibus Asiae, 1950, 124–8Google Scholar.

¹⁴ The most comprehensive, and by far the most valuable variorum edition, done by a philologist of the highest repute, is the Rōshi kōsei by Kunio, ShimaTokyo: Kyukoshoin, 1973Google Scholar. For some reason this work seems not to be widely known in North America. I am grateful to Derek Herforth, of the University of British Columbia, for providing me with a copy of it.

¹⁵ This does not mean, of course, that the Wang Pi and Ho Shang-kung commentaries should be taken as in any way representing a common interpretation of the text. In fact they are quite different, and have long been regarded as complementary. See, for example, Hung, William, ‘A bibliographic controversy at the T'ang Court A.D. 719’, HJAS, XX, 1–2 1957, 74–134, esp. 81.Google Scholar

¹⁶ The graph cheng is attested as a chia-chieh loan for cheng , OC in both the Hsiao ching and Hsiin tzu. See Chu Chün-sheng , Shuo wen t'ung hsün ting sheng s.v. cheng The Old Chinese (OC) reconstructions given in this paper follow grosso modo the system proposed by Pulleyblank, E. G. in his most recently available articles: ‘Some new hypotheses concerning word families in Chinese’, Journal of Chinese Linguistics, I, 1973, 111–25;Google Scholar ‘The final consonants of old Chinese’, to appear in Mon. Serica (in honour of Father Paul L.-M. Serruys); and ‘The Chinese cyclical signs as phonograms’, JAOS, XCIX, 1, 1979, 24–38. Since there are still a great many unanswered questions about the reconstruction of OC in general some of the specific forms posited herein may not be reconstructed exactly as Pulleyblank would have done, and it should not be assumed that he would endorse the specific forms in every case.

¹⁷ The term sheng , conventionally translated as ‘sage’, sometimes more exaltedly as ‘saint’, is regularly written in the MWT A manuscript as and in the B manuscript as (i.e., without the t'ng as the bottom element). At first glance the A manuscript's may look like a mere chia-chieh loan, and the B manuscript's an ‘abbreviated’ character. But taken together these two graphic forms of the word sheng ( ) give us a valuable clue to the precise meaning of the word from the Chinese point of view, (‘sage’ and ‘saint’ each being a rough approximation, and heavily coloured, especially in the case of ‘saint’, by Western connotation and usage.)

The two words sheng ‘sound’ and sheng ‘sage’ differ only in tone in modern Chinese, and differed only in the presence or absence of a final in Old Chinese: was and was . Clearly the of is a ‘phonetic’. What is curious is that the form of the B manuscript, without the ‘phonetic’, shares both a graphic constituent element (the ) and a close pronunciation ( ) with . The implication of this is that we can correlate the shared graphic element with the common pronunciation, thus appears to function as a ‘phonetic’ with the reading in both and . Moreover, the closely allied word t'ing has also the same element, and a nearly identical pronunciation. The case for is made even stronger when we examine the set semantically:

‘sound’, ‘hear’, and (‘ear’), clearly all having to do with audibility, or an audient capacity. As a noun, meaning ‘ear’ was, of course, read erh, but it must have had a second reading with a meaning associated with ‘hearing’, ‘that which is heard,’ or ‘hearability’, and as such functions not only as ‘phonetic’ in ‘sound’ and , ‘hear’, but also as a semantic component. This leads us to speculate that it is also functioning semantically in (where it is ‘phonetic’ in the reading as we noted above), and that the primary meaning of is related to the set , and should be understood as something like ‘one who is adept at “hearing”’.

Notice, in support of this hypothesis, the important status given to ‘hearing’, having one's ‘ear in accord’, in Lun yü III: ‘at sixty one's ear is in accord’, i.e. one ‘hear's correctly’. This is what we think the sheng jen or sheng wang was: someone who heard (and understood).

In the Indo-European linguistic tradition knowledge and understanding is largely a matter gained through the sense of sight; cf. wisdom, vision, wit, Veda, etc. all traceable to the root for SEEING. There is no a priori reason why the Chinese could not have looked upon these same capacities as being gained via HEARING, and this is what I think the sheng represents.

¹⁸ Ch'ien-chih, Chu , Lao tzu chiao shih , Tainan, P'ng p'ng publishers, 1975, 10Google Scholar.

¹⁹ Jao Tsung-i, A study on Chang Tao-ling's Hsiang-er commentary, 86.

²⁰ I am adopting Ma Hsü-lun's suggestion that shu read as OC is a loan for su ‘quickly’, OC (both are Shih ching rhyme group wu. See Ma, 1973, 41.

²¹ Lau, D. C., Lao tzu, Too te ching, Penguin, 1963, 61, 187Google Scholar; Wing-Tsit, Chan, The way of Lao tzu, Indianapolis and N.Y., Bobbs-Merrill, 1963, 107–8Google Scholar; and Waley, A., The way and its power, N.Y., Grove Press, 1958, 147Google Scholar.

²² Ch'eng-hai, Cheng , Lao tzu Ho Shang kung chu kou li , Taipei, Chung hua shu chü, 1971, 39Google Scholar.

²³ This was also pointed out at the June 1979, Berkeley Workshop on the Ma wang tui texts by E. G. Pulley blank in his paper ‘Linguistic notes on the Laozi Manuscripts’, 3.

²⁴ This discrepancy between the Wang Pi text and the Wang Pi commentary has been noticed and remarked upon by Chan, The way of Lao tzu, 162–3.

²⁵ Chu Ch'ien-chih, Lao tzu chiao shih, 77.

²⁶ Jao Tsung-i, A study on Chang Tao-ling's Hsiang-er commentary, 82–92, compares the HE commentary with the Ho Shang-kung commentary, and concludes that the former in some cases shows further refinement of the latter, and therefore must be a later compilation. When the texts are compared with the MWT silk manuscripts it turns out that this conclusion is not defensible, and the HE commentary is more likely to pre-date the Ho Shang-kung. In chapter 19, for example, Jao interprets the variation we noticed above as item B.3 between tzu san che of the R texts (including the Ho Shang-kung) and tzu san yen of the HE, together with their respective commentaries to this line, as reflecting a further refinement and more developed sense on the part of the HE in changing the text from an original che to yen (pp. 87–8). But in fact, as was noted above, the HE's yen is validated by the MWT manuscripts, both of which have yen also. Thus it is Ho Shang-kung text that is the altered one, and Jao's argument of textual priority for it cannot stand. Other cases are not always so clear cut, but it can be said that not one instance of textual variation incontrovertibly gives textual priority to the Ho Shang-kung version.

²⁷ Loewe, Michael, Ways to paradise: the Chinese quest for immortality, London, George Allen and Unwin, 1979Google Scholar.

²⁸ Chan, 1963, 105.

²⁹ Waley understands this line similarly. See Waley, 1958, 146.

³⁰ Juan Yüan (1764–1849), in his compendious Ching chi tsuan ku , gives only one other significant textual occurrence where huo is glossed as ch'ang , this is Lun yü XIII.22 and comes in the citation of a line from the I Ching, hexagram 32, heng ‘constancy’. There is no compelling reason in this line to read huo as ch'ang; I suspect that that gloss has arisen through the contextual association with heng ‘constancy’.

³¹ Clearly a textual emendation from A to B where B is a near homonym of A is much less disruptive and more acceptable than it otherwise would be. Beyond this the significance of the homonymy here ( ) and in example C.5 below ( )is not obvious.

³² The term shih chieh is mentioned in Hou Han shu 82B as an accomplishment of a certain Wang Ho-p'ing , said to have been a partisan of Taoist techniques (hoo too shu ) where it was apparently unfamiliar enough to have required an explanatory note. The scholars appointed by the heir-apparent, Li Hsien , in the seventh century to enter a commentary to the text of the Hou Han shu explain that shih chieh is a process for release and transformation of one who is about to rise to transcendence (Wang Hsien-ch'ien, 1968, 3027). See also Ngo Van, Xuyet, Divination, magie et politique dans la Chine ancienne (Bibliothèque de l'Éicole des Hautes Études, Sec. des Sciences Religieuses, vol. 78), Paris, Presses Universitaires de France, 1976, 146–7; and 204–9, ‘Art de longue vie et d'immortalité’.Google Scholar

³³ It should be noted that the R texts all have tzu hi chi chin , matching the HE text, and thus inconsistent with their use of chung fu . The clear sense of ‘tracing, following back' of the MWT manuscripts' shun ) is also diluted somewhat by the R texts’ yüeh ‘scrutinize’. This may have been the reason why the inconsistency of saying ‘starting from antiquity reaching down to the present’ in connexion with a search for origins has not been more noticeable.

³⁴ Waley, 1958, 249.

³⁵ Lau, 1963, 71.

³⁶ Chu Ch'ien-chih, 1975, 39–10.

³⁷ Yü Ying-Shih, 1964–65, 85–S6.

³⁸ See R. A. Stein, ‘Religious Taoism and popular religion from the second to seventh centuries’, in H. Welch and A. Seidel (ed.), Facets of Taoism, 1979, 53–81.