On the Chapter Divisions in the "Lao-tzu"

Author(s): Robert G. Henricks

Source: Bulletin of the School of Oriental and African Studies, University of London ,
1982, Vol. 45, No. 3 (1982), pp. 501-524
Published by: Cambridge University Press on behalf of School of Oriental and African
Studies
Stable URL: http://www.jstor.com/stable/614922

JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide
range of content in a trusted digital archive. We use information technology and tools to increase productivity and
facilitate new forms of scholarship. For more information about JSTOR, please contact support@jstor.org.

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at
https://about.jstor.org/terms

Cambridge University Press and School of Oriental and African Studies are collaborating with
JSTOR to digitize, preserve and extend access to Bulletin of the School of Oriental and African
Studies, University of London

This content downloaded from

103.62.153.18 on Mon, 27 Feb 2023 07:38:21 UTC
All use subject to https://about.jstor.org/terms
 ON THE CHAPTER DIVISIONS IN THE LAO-TZU

By ROBERT G. HENRICKS
The Ma-wang-tui texts of Lao-tzu 1 have seemingly confirmed two
held to be true by many scholars: (1) that the text from an early d
right from the outset, was divided into two parts;2 and (2) that
not claimed that the Ma-wang-tui texts represent the original text
originally divided into chapters. This latter conclusion is open to d
although there are neither chapter numbers nor names in the Ma

texts, there are punctuation dots (-) in part II of text A, often pl

beginning of one of our present chapters.3 However, absence of p
elsewhere, and the failure of what there is to completely coincide wit
chapter divisions, would seem to indicate at least that the present
chapter divisions-thirty-seven in part I and forty-four in part
yet (c. 180 B.c.) uniformly observed.
This has led me to speculate about the punctuation of the Lao-t
were the current eighty-one chapters settled upon, and why ? Cou
other schemes for dividing the text ? Are the current chapter div
at the right places ? If not, is it possible to identify, by reading
sequence, the boundaries of distinct units of meaning ?

1 For a general introduction to the two Ma-wang-tui texts of the Lao-tzu found at
(in Changsha, Hunan) in 1973, see my article, 'Examining the Ma-wang-tui silk t
Lao-tzu: with special note of their differences from the Wang Pi text', T'oung P
1979, 166-99.

2 Much has been written on the division of the text into two parts, a' Tao ch
a 'Te ching' . . While most scholars accept the two-fold division as basic, s
titles ' Tao' and 'Te ' came later, and there is debate about when this first bec
a' canon' or ' classic' (though in most passages translated in this paper I use ' book
Ssu-ma Ch'ien in Shih-chi 63 (Peking: Chung-hua, 1959), vii, 2141, simply ref
with 'sections I and II' (I - ). There is nothing in the Ma-wang-tui text
that the Lao-tzu was yet (early second century B.c.) referred to as a ching. Howev
are divided into two; and the division embraces the same material as that embra
present twofold division of the text, and 'Tao' and 'Te' are used as the titles
sections, at least in text B. At the end of part I of that text we find ' Tao, 2426

(': - 7- I. - W A ), and at the end of part II, 'Te, 3041 [characters

t-I
~ -,-ft --). ForWen
~ (Peking: the wu,
Ma-wang-tui texts
1980), I, 124, 113.see Ma-wang-tui Han mu po-shu, . IE Jf i -
For the many opinions and arguments on the two-fold division and the references to ching,
'classic', see the following: Ma Hsii-lun .I ' ff Lao-tzu chiao-ku 5 }- j (Peking:
Chung-hua, 1974), 13-19; Chu Ch'ing-yiian ; [ j (or Chu Ch'ien-chih ;J ), Lao-tzu
chiao-shih ( ~ ; (Taipei: Shih-chieh, 1968), 1-2 (of the text itself); and Chiang
Hsi-ch'ang a, Lao-tzu chiao-k.u J - g j (Taipei: Ming-lun, 1971), 1-2 (of the
text itself). Also, in English, see Wing-tsit Chan, The Way of Lao Tzu (Indianapolis: Bobbs-
Merrill, 1963), 74-5.
* See my article, 'A Note on the question of chapter divisions in the Ma-wang-tui Manu-
scripts of the Lao-tzu ', Early China, Iv, 1978-9, 49-51. The dots occur at the beginning
of the present chapters 46, 51, 53, 57, 63, 64, 69, 73, 75, 76, 80, and 1, and in the middle of
chapters 46, 51, 52, 72, 75, and 81.

This content downloaded from

103.62.153.18 on Mon, 27 Feb 2023 07:38:21 UTC
All use subject to https://about.jstor.org/terms
502 ROBERT G. HENRICKS

My search for answers to these ques

(1) an examination of sources on the L
the years about the chapter divisions in t
of the text to see where I would break
I never read another edition.
In this paper I shall present the results of both of these enquiries. But first
let me say that I now consider the number eighty-one to be arbitrary. This
view is not new: Duyvendak reached the same conclusion many years ago.
In the introduction to his translation of the Lao-tzu he said: 4

The text numbers 81 chapters. This number is not accidental. The number
81, that is 3 x 3 x 3 x 3, is a sacred number in Taoism and it is certain
that the division was made artificially during or shortly after the Han
dynasty (206 B.C.-A.D. 221).
I concur with the view that the eighty-one-fold division was made in the
Han. However, I would relate it more to the conception of Yin and Yang-
i.e. 81 = 9 x 9, and 9 is the perfection of Yang.

I. Evidence from early sources

The evidence one can find on this problem in early records is not as complete
as one would wish. None the less, what there is is both interesting and
suggestive. Let me go over these items one by one, commenting as I proceed.
Some of this information has been noted earlier by Wing-tsit Chan,5 but the
evidence bears reviewing in the light of Ma-wang-tui. My list does not pretend
to be exhaustive; it represents my conclusions so far.
1. First, Liu Hsin glJ p (c. 46 B.C.-A.D. 23) in his Ch'i-liieh -L W says that
his father, Liu Hsiang lIj (79-8 B.c.), 'set the work at two sections with 81
chapters. The first book had 34 chapters and the second 47 '. ( g ~i

- two
for A t- - --First,
reasons. , ?[what
@ -is Et V;understood
to be - t g by .) 6 This 2is& ?interesting
ting-chu Its primary
meaning is to edit, collate, a text. This could mean that it was Liu Hsiang
who, working with various versions of the text, first divided the text into two
sections with eighty-one chapters. The second point of interest is the thirty-four
chapters in the first part and forty-seven in the second; modern versions have
thirty-seven and forty-four. It might be supposed that this is just a copyist's
error, and other evidence cited below seems to support a division at thirty-seven

4 See J. J. L. Duyvendak, Tao Te Ching : the book of the Way and its virtue (London: John
Murray, 1954), 4.
5 Chan, The Way of Lao Tzu, 75-6.

* This is quoted in the commentary to Tung Ssu-ch'ing's g , g (Sung) preface to his

Tao-te chen-ching chi-chieh @ , HY 705 (text no. 705 in the list of the works
in the Tao-tsang according to the Harvard-Yenching Index to the Taoist Canon, Tao-teang
tzu-mu yin-te ( $i ,- j i ~, Peking, 1936; reprint, Taipei, 1966). For the text of the
preface itself see Cheng-t'ung tao-tsang ]E ~ (Taipei: I-wen, 1977), xxI, 16875. That
Liu Hsin calls the sections of the Lao-tzu 'ching' would seem to confirm, if this is authentic,
that the Lao-tzu had 'classic' status by this time.

This content downloaded from

103.62.153.18 on Mon, 27 Feb 2023 07:38:21 UTC
All use subject to https://about.jstor.org/terms
 ON THE CHAPTER DIVISIONS IN THE LAO-TZU 503

at this time. Nevertheless, the number thirty-four occurs more

our sources. K'ung Ying-ta #L g S (593-648) is said to have
commentary on the Lao-tzu, and his text had sixty-four chapte
part I, and thirty-four in part II.7 And, as we shall see, there is r
that part II of Yen Tsun's gi g (53-24 B.C.) text was also d
thirty-four.
This may be no more than coincidence, and the thirty-four pertains to
part II of the text for Yen Tsun and K'ung Ying-ta, part I in the text of Liu
Hsiang. But, Yen and K'ung's part II could have been Liu's part I, since the
Ma-wang-tui texts reveal that, in at least some versions of the text, parts I
and II were at that time reversed. (The Ma-wang-tui texts begin with part II,
chapters 38-81, and end with part I, chapters 1-37.)

2. Secondly, the essay 'Shuo-mu' - (Explaining the Divisions) which

prefaces Yen Tsun's Tao-te chih-kuei lun g $ - states that his text
had seventy-two chapters; forty in part I and thirty-two in part II.* The
authenticity of this treatise has been questioned.9 Nevertheless it is worth-
while quoting the relevant passage. The text runs: 'The first book matches
Heaven and the second book matches Earth. The way of Yin is 8; the way
of Yang, 9. Since Yin activates (?, hsing i) Yang, there are seventy-two
divisions (shou -i); since Yang activates Yin, the text is divided into two.
Since 5 activates 8, the first book has 40 [chapters] and then begins again;
since 4 activates 8, the second book has 32 [chapters] and then comes to an
end.' 10
The connexion of Yin-Yang and numerology with the chapter divisions of
the Lao-tzu is an important one and is noted in more than one source as we
shall see. First, however, there is another important point to be made about
Yen Tsun's text. Hu Chen-hsiang iJ g (Ming, fi. 1600) in a piece entitled
'T'i Tao-te chih-kuei' g ~ $ (Notes on the Tao-te chih-kuei) says
that part II of the text he had-part II is all that survives, save fragments, of
Yen Tsun's commentary-had thirty-four chapters, not thirty-two. He adds

7See Yen Ling-feng g i , Chung-wai Lao-tzu chu-shu mu-lu c 4 f ~ j-

g , (Taipei:
chapter Chung-hua
divisions ts'ung-shu
are no longer wei-yiian-hui,
extant; 1957),
they are noted 27. Te-chih
by Chu K'ung's *commentary and
j 5 in his Lao-tzu

t'ung-i fan-lieh J j.- A ~j flJ.

8 The text is HY 693. In Cheng-t'ung tao-tsang, see xx, 16154-235. In the Tao-tsang text,
the noted essay is called '[Yen] Chiin-p'ing's explanation of the division of the two books'
( ZJ~i ~ J ). Yen Tsun is actually Chuang 4 Tsun. His name was changed by
Pan Ku to avoid the taboo on the name of Emperor Ming of the Han (Liu Chuang IJ 4,
r. A.D. 58-75). See Cheng-t'ung tao-tsang, xx, 16156, the appended commentary to Yen Tsun's
' Shuo-mu '.

SSee Yen Ling-feng's' " Ku-shen-tzu hsii " yii "shuo-mu " i wei-wu' ' ~J Jj f
U f , in Yeng Ling-feng (ed.), Wu-ch'iu pei-chai Lao-tzu chi-ch'eng
- (Taipei: I-wen, 1965), I, 6a-9b. Most convincing is his argument on p. 7a that
part II of the Tao-tsang's Ku-shlen-tzu chu has forty chapters, not thirty-two.
1o HY 693; Cheng-t'ung tao-tsang, xx, 16156.

This content downloaded from

103.62.153.18 on Mon, 27 Feb 2023 07:38:21 UTC
All use subject to https://about.jstor.org/terms
 504 ROBERT G. HEh'NRICKS

that the thirty-four was arrived at (based on the normal count

by:
- combining chapters 39 and 40,
- combining the last part of chapter 58 (starting with ' The people have been
deluded for a long time ') " with chapter 59,
- combining 67 and 68,
- appending the first part of 58 to the end of 57, and
- omitting chapters 75-81.12

I will return to this in the second part of this paper; there are in-
teresting connexions with the texts of Ma-wang-tui, especially with regard to
chapters 39 and 40 and 67 and 68.

3. Numerology of one sort or another-Yin-Yang, the Five Elements, the

Four Seasons-is involved in at least four other accounts of the chapter
divisions, two of them found, like the fragment of Liu Hsin's Ch'i-lilieh cited
above, in the commentary to Tung Ssu-ch'ing's (Sung) preface to his Tao-te
chen-ching chi-chieh. There we find the following statements:
(a) 'Ho-shang Kung divided the text into 81 chapters to match the extreme
(? perfect ?, 4) number of Great Yang [i.e. 81 = 9 x 9]. Book I has
37 chapters modelled on the odd numbers of Heaven. Book II has 44
chapters modelled on the even numbers of Earth.' 13 Since Ho-shang
Kung is traditionally believed to have lived during the reign of Emperor
Wen of the Han (179-157 B.C.), were this statement true, the eighty-one
divisions would date from that time. But whether there was a Ho-shang
Kung in the second century B.c. is a point open to question; the com-
mentary bearing his name seems to date from the third century A.D.14
Moreover, it is worth noting that there is nothing in Ko Hsiian's gi
(A.D. 184-246) preface to Ho-shang Kung's commentary to indicate that
Ho-shang Kung had so divided the text: 1i thus, this story would seem to
be late.
(b) 'Ko Hung [g jJ, 253-333 ?] and others made additions and omissions
[presumably meaning they added chapters on to one part of the text and
took some away from the other], saying that since Heaven completes things
with the four seasons, Book I should have 4 times 9-36 chapters. And
since the Earth completes things with the five elements, Book II should have
5 times 9-45 chapters. Together they match the number 9 times 9.' 16

" Tr. by Wing-tsit Chan, The Way of Lao Tzu, 203.

12 Hu's 'T'i' is included with the prefatory material to Yen Tsun's commentary in Yen
Ling-feng's Wu-ch'iu pei-chai Lao-tzu chi-ch'eng, I.
13 HY 705; Cheng-t'ung tao-tsang, xxI, 16875.
14 See Wing-tsit Chan's review of the evidence in The Way of Lao Tzu, 76-82.
15 For Ko Hsiian's 'Preface' see Yen K'o-chiin's 4 iJ ~ Ch'I1an shang-ku 8an-tai Ch'in,

Han, san-kcuo,
Chui-bun liu-ch'ao
4i 7 edition (Kyoto, 1972). wen ~_ _ - ~ ' J , II, 1460, in the
16 HY 705; Cheng-t'ung tao-tsang, xxI, 16875.

This content downloaded from

103.62.153.18 on Mon, 27 Feb 2023 07:38:21 UTC
All use subject to https://about.jstor.org/terms
 ON THE CHAPTER DIVISIONS IN THE LAO-TZU 505

That Ko Hung 'added and subtracted' (fl f= g) suggests to

did so to the eighty-one chapter text of thirty-seven and forty-f
a chapter away from part I and adding it on to part II. But
have been done in more than one way."7

However that may be, Ko Hung and his thirty-six/forty-five

followed by T'ang Hsiian-tsung t g (r. 713-756) (sometimes cred
being the first to establish the Tao and Te sections of the text a
chapters and sentences to accord with that division is) with his
commissioned Tao-te chen-ching shu a RT j. But a new ra
supplied for thus dividing the text. In notes prefacing the comm
read that chapters 1-9 illustrate the way of spring, 10-18 the way of
19-27 the way of fall, and 28-36 the way of winter, while 37-45
of part II) illustrate the virtue of jen f, 46-54 the virtue of li ji
virtue of i i, 64-72 the virtue of chih ', and 73-81 the virtue o
Finally, Ch'en Ching-yiian [ & jI (fl. 1258) says, in his' Tao-te c

tsang-shih
comes tsuan-wei
first and the k'ai-t'i'
Te book follows. gthe
I and II, j Itwo, divisions,
.,'The
on the birth and nurture [qualities] of the two primaries [Yin an
81 chapters reflect the extreme number of Great Yang. Theref
clarifies the Tao, modelling itself on Heaven; Book II clarifies Te
models the Earth. The numbers of Heaven are odd; therefore,
37 chapters: the numbers of Earth are even; therefore Boo
chapters.' 20

4. The 'Lung-hsing kuan' gg |i g text, engraved on a stone tablet in

I-chou & ~1+ in 708 (Ching-lung & gi 1) 1-a text which has much in common
with the Ma-wang-tui texts "2-has neither chapter numbers nor chapter
names. However, there are breaks in the text where chapters are to begin,
and these do not totally coincide with the divisions currently made. They
differ in three respects:

- chapters 4, 5, and 6 are here combined into one,

- the last two lines of our chapter 11 (' Therefore turn being into advantage,

1; i.e. by ending part I with our own chapter 36 and then beginning part II with 37; or by
removing a chapter from part I and placing it in part II; or by combining two chapters in
part I and splitting two in part II.
18 See Chu Ch'ing-yiian's Lao-tzu chiao-shih, 2, where a statement from Tung Chiung's
( j Tsang-shu chih & ~ is cited by Wang Ch'ang r @.
19 See HY 679, Cheng-t'ung tao-tsang, xIx, 15416.
20 Quoted by Yen Ling-feng in ' Pien Yen Tsun Tao-te chih-kuei lun fei wei-shu '
ff , in Wu-ch'iu pei-chai Lao-tzu chi-cheng, 7ab.
21 This is the base text for Chu Ch'ing-yiian's Lao-tzu chiao-shih. Chiang Hai-ch'ang notes it

as
ji ]the 'pei' text
} (Tokyo: in his
Kyilko Lao-tzu
~ -j, chiao-ku,
1973) notes this and Shima Kunio
as' Lung-hsing , $[5
kuan 4, in RBshi kasei
pei'.
22 For example, is written with 51 in both cases. Variant characters and variant
readings also coincide from time to time.

This content downloaded from

103.62.153.18 on Mon, 27 Feb 2023 07:38:21 UTC
All use subject to https://about.jstor.org/terms
 506 ROBERT G. HENRICKS

and turn non-being into utility ')23

'Therefore ', and
- the first two lines of our chapter 26
The tranquil is the ruler of the hasty
Judging from the context, these last
example, if we move the first two lines
begin with a 'Therefore', which would b
will be further discussed below.

5. Wu Ch'eng : g (Yiian, 1249-1331), in a colophon to his Tao-te chen-

ching
text (i.e.chu
the M
usual~.eighty-one
,27 notesand
theYen
existence of varied but
Tsun's seventy-two), ways
he of dividing
states that the
divisions have been made where they should not be. Thus, in his commentary
he has a text with only sixty-eight chapters, with thirty-two chapters in part I
and thirty-six in part II. The sixty-eight is arrived at by combining:
- the present chapters 5 and 6,
- 17, 18, and 19,
- 23 and 24,
-40, 41, 42, and 43,
- 57 and 58,
- 63 and 64,
-67, 68, and 69, and
- 73 and 74.

In addition, in Wu Ch'eng's text the last line of 78 (' Straight words se

to be their opposite' 28 TE - J ) is made the first line of 79.
Wu Ch'eng's ordering of the text is followed by Ming T'ai-tsu A jli (r. 1368
1399) in the imperially commissioned Tao-te cheng-ching chu,29 with only on
exception; in that text chapters 76 and 78 are also combined (with chapter
coming between 78 and 79), giving a text of sixty-seven chapters instead
sixty-eight, with thirty-two in part I and thirty-five in part II. The last
of 78 is still the first line of 79.

has6. Li Yiieh's ~chapters

seventy-eight j (fi. 780) Tao-te
instead chen-ching
of eighty-one, hsin-chu
with g , iin1)pa
thirty-seven
and forty-one in part II. Combined are chapters 43 and 44, 48 and 49, a
68 and 69.30

23 Chan's translation, The Way of Lao Tzu, 119.

2 ibid., 146.
25 See p. 2 in Chu Ch'ing-ytian's Lao-tzu chiao-shih where he cites the findings of H
I-hailan .f ~F k~
26 The third line of 26 is 'Therefore the sage travels all day ...' (Chan, The Way of
Tzu, 146).
2?7 HY 704, Cheng-t'ung tao-tsang, xxI, 16874.
28 Chan, The Way of Lao Tzu, 236.
29 HY 676, Cheng-t'ung tao-tsang, xIx, 15233-73.
30 See HY 692, Cheng-t'ung tao-tsang, xx, 16127-54.

This content downloaded from

103.62.153.18 on Mon, 27 Feb 2023 07:38:21 UTC
All use subject to https://about.jstor.org/terms
 ON THE CHAPTER DIVISIONS IN THE LAO-TZU 507

7. Finally, a few miscellaneous items should be noted.

(a) Pan Ku [ f J (A.D. 32-92), in the Han shu 'I-wen ch

lists four commentaries (?) on the Lao-tzu. They are: Lao-tzu Lin

chuan, ssu p'ien i Jy- i i E fg, Vg ~ (Mr. Lin's commentary on

Lao-tzu: four Sections); Lao-tzu Fu-shih ching-shuo, san-shih-ch
f~ , I t -,- t (Mr. Fu's explanations of the Classic Lao-tz
seven Sections); Lao-tzu Hsii-shih ching-shuo, liu p'ien
g J (Mr. Hsii's explanations of the Classic Lao-tzu: six Sectio
Hsiang's Shuo Lao-tzu, ssu p'ien - ~ -, mg J (Explaining the Lao
Sections).3" We do not know if the 'four sections ', 'six sections'
titles refer to the way the text itself was then divided, or simply to
of sections in the commentary. However, the number thirty-seven i
work is certainly appealing. Since this is the usual number of c
part I of the text, it is tempting to conclude that the thirty-seven/
division did already (end of Former Han ?) exist, and that Mr. Fu's co
was simply on the first part of the text.
There is another piece of evidence which might confirm a Han da

'thirty-seven' division. Sun I-jang g, g - points out that th

huo lun 4p. y- Ja g - (Master Mou reasoning against doubts) at
mentions Mr. Lao's 'Tao ching' in thirty-seven sections
-A~ ).Y3 The Mou-tzu, according to its preface, dates from the end o
century A.D. However, this date is not commonly accepted, and
place it as late as the fifth century.34

(b) Ma Hsii-lun, in passing, points out that Lu Te-ming j f, i1

in his Chuang-tzu t'ien-hsia p'ien yin-i 4 .f - ~ ref
Lao-tzu as a text written in '19 sections' (shih-chiu p'ien -t
Ma considers this a mistake for ' Sections I and II' (shang-hsia p
(c) Finally, Ma Hsii-lun also notes that some feel, on the b
'Chieh-lao' J ' chapter in the Han Fei-tzu ~.~ f- (Han Fei w
233 B.c.), that the text was then divided into fifty-five chapters 36
because the Han Fei-tzu itself has fifty-five chapters). But tha
argument. The 'Chieh-Lao' chapter itself, we might note, 'ex

a1 Han shu, 30, vI, 1729 (using the Chung-hua shu-chii edition [Peking, 1962]
32 We have no dates for the first three; Liu Hsiang's dates are 79-8 B.c. Th
titles again (see n. 3 above) seem to indicate that the Lao-tzu was considered a c
by the end of Former Han. Ma Hsii-lun thinks not. He argues that this is an a
the part of Pan Ku. (See his Lao-tzu chiao-ku, I, 13). For evidence he points to
Liu Hsiang's text is simply on the Lao-tzu, and he feels that the other three tex
before Liu Hsiang's so they are listed before it by Pan Ku.
3 Cited in Chu Ch'ing-ytian's Lao-tzu chiao-shih, 2.
4 See Kenneth Ch'en, Buddhism in China: a historical survey (Princeton: Pr
versity Press, 1964), 36-7.
35 See Ma Hsti-lun, Lao-tzu chiao-ku, T, 16.
36 See ibid., 20.

This content downloaded from

103.62.153.18 on Mon, 27 Feb 2023 07:38:21 UTC
All use subject to https://about.jstor.org/terms
508 ROBERT G. HENRICKS

from, according to present chapter division

59, 60, 46, 14, 1, 50, 67, 63, and 54.37

What can we conclude about the chapt

basis of this interesting but varied data ? I a
1. That the text was probably divided into
with thirty-seven in part I and forty-fo

Han
a (A.D. 9). That
commentary the
usually Hsiang-erh
credited ch
to Chang

grandson Chang Lu , (fl. 188-220) 38-i

argue against this.39 But the evidence we h
Hsin and 'Mr. Fu's explanations ') is to m
Moreover, I would argue, and it is well to
outset readers recognized distinct units
often correspond with present chapter d
of the text may have lacked punctuation
The evidence for this lies in (a) the punctua
of the Ma-wang-tui texts, and (b) the fac
in the 'Chieh-Lao' and 'Yii-Lao' chapte
same chapter boundaries as our own in at

2. If the above cited fragment from L

I think it quite possible that it was, in fact
chapters at eighty-one, with thirty-seve
(reading the thirty-four and forty-seven
to imagine that he had several copies
perhaps, with a slightly different punc
(ting chu ' g) after editing, at eighty-on
3. I am convinced that the number eighty
Yin-Yang speculation. True, explanations

3 The 'Yii-Lao' P9 chapter 'illustrates' lines

63, 64, 52, 71, 64, 47, 41, 33, and 27.

4 38
t1 See JaoKong:
(Hong Tsung-i's study,
Tong Nam, Lao-tzu H
1965).
9 Wing-tsit Chan (The Way of Lao Tzu, 75) poin
chapter divisions in Lu Te-ming's (556-627
(581-618) or early T'ang (618-907) origin of the div
40 That is to say, the first and last lines of a chap
cited in sequence, set off from lines from others ch
38, 58, 59, 60, 46, 50, 53, 54, 26, and 47, and one cou
possible, however, and I will raise this point again
59, and 60 as one unit, and the same applies to 53
4x This fragment is not found in the Ch'i-lieh
Ch'ian shang-klu san-tai Ch'in, Han, san-kuo, liu-c
42 I think this is Yen Ling-feng's conclusion a

shih-t'an
concludes .*I
that it J.
was ~f g 5 into
divided - g eighty-one
(Taipei: Ho-
chap

This content downloaded from

103.62.153.18 on Mon, 27 Feb 2023 07:38:21 UTC
All use subject to https://about.jstor.org/terms
 ON THE CHAPTER DIVISIONS IN THE LAO-TZU 509

to come after the fact. But the fact that eighty-one happens to e
nine, with nine being the 'perfect' Yang number, is too signif
coincidence. It is certainly germane to point out that Liu Hsia
supporter of Tung Chung-shu 2 44 g (179?-104? B.c.), the or
in the Han of Yin-Yang and Five Elements cosmology.43
4. If our sources are accurate, it would appear that the eigh
version of the text was not the only one used (e.g. Yen Tsu
chapters and K'ung Ying-ta's sixty-four, and Wu Ch'eng's six
over, even with the eighty-one division, there were different w
parts I and II (e.g. Ko Hung's thirty-six and forty-five).
To conclude this part of my study I would call attention to th
those texts with less than the normal eighty-one count, where t
the present chapters are occasionally combined, there is remarka
in the combinations. For example, chapters 5 and 6 are co
Ch'eng and Ming T'ai-tsu, while 4, 5, and 6 are combined in t
kuan' text; 39 and 40 are combined in the Yen Tsun text wh
combined by Wu Ch'eng and Ming T'ai-tsu; 57 and 58 are co
Ch'eng and Ming T'ai-tsu with the beginning of 58 attached
in the Yen Tsun text; and 67, 68, and 69 are combined by W
Ming T'ai-tsu, while 67 and 68 are combined in Yen Tsun
combined in Li Yiieh. Moreover, in footnote 40 I suggested t
59, and 60 might be read as a single unit in the Han Fei-tzu,
58 is tied to the beginning of 59 in Yen Tsun.

II. Some recent studies and the evidence from the text itself
In more recent times several attempts have been made
text 'correctly'. First, Yao Nai M gp (1731-1815), in his L

-r- i - the
changes (preface
chapterdated 1783),dividing
boundaries, maintains the eighty-one
and combining chapter
the thirty-seven count but
chapters of part I to give thirty-one, and making fifty chapters of part II's
normal forty-four.44 This is arrived at by:
- combining chapters 2 and 3,
- attaching the last two lines of chapter 5 to the beginning of chapter 6,
- combining 7 and 8,
- combining 14 and 15,
- dividing 17 into two after the line ' And the next are those who are despised'
(Chan, 130),
- adding the first line of 20 to the end of 19,
- combining the rest of 20 with 21,

43 At the end of Tung Chung-shu's biography (Han shu, 56, viii, 2526) we find : ' Liu Hsiailg
acclaimed Tung Chung-shu as one whose abilities were those of a minister of kings.' (Cited
from Fung Yu-lan's History of Chinese pIhilosophy, transl. by Derk Bodde. Princeton : Princeton
University Press, 1953, II, 17.)
* See Yao Nai, Lao-tzu chang-i (T'ung-ch'eng j ': Wu-shih ch'ung-k'an perI)" t
J f(J 4, 1870).

This content downloaded from

103.62.153.18 on Mon, 27 Feb 2023 07:38:21 UTC
All use subject to https://about.jstor.org/terms
510 ROBERT G. HENRICKS

- combining 22, 23, and 24,

- combining 30 and 31,
- splitting 36, with the last two lines (fr
the water '-Chan, 164) forming a separa
- combining 39, 40, and 41,
- dividing 42 into two, with 'People hat
(Chan, 176) starting a new chapter,
- dividing 45 into two, breaking after 'The
(Chan, 180),
- dividing 46 into two, beginning the second part with 'There is no calamity
greater than lavish desires' (Chan, 181),
- combining 49 and 50,
- adding the first three lines of 53 (to ' Fear getting off the road '-Chan, 194)
to the end of 52,
- splitting 54 after 'Thus from generation to generation his ancestral sacrifice
will never be suspended' (Chan, 195),
- adding the first four lines of 59 (to 'The people are disappointed and con-
tentious '-Chan, 203) to the end of 57,
- splitting 60 into two, with the first line (' Ruling a big country is like cooking
a small fish '-Chan, 207) standing alone,
- making the first three lines of 63 (to 'Whether it is big or small, many or
few'--Chan, 212) a separate chapter,
- making the fourth line of 63 (' repay hatred with virtue '-Chan, 212) a
separate chapter,
- combining the rest of 63 (from ' Prepare for the difficult while it is still easy'
-Chan, 212) with the first half of 64 (to 'The journey of a thousand li
starts from where one stands '-Chan, 214),
- making the first three lines of 67 (to 'it would have been small for a long
time '--Chan, 219) a separate chapter,
- combining the rest of 67 (from 'I have three treasures '-Chan, 219) with
68 and 69,
- splitting 72 into two, following ' Then what is greatly dreadful will descend
on them' (Chan, 226),
- splitting 73 into two, following 'Even the sage considers it a difficult
question' (Chan, 228),
- combining 74 and 75,
- making the last line of 78 the first line of 79, and
- making the last two lines of 79 (' The Way of Heaven has no favourites.
It is always with the good man '-Chan, 237) a separate chapter.
Some of these changes are new; some we have seen before.
Secondly, Ma Hsii-lun, having dispensed with the eighty-one chapter
count, while for the most part maintaining the present sequence of material in
the text but here and there rearranging some lines, settles on 117 different
sections.45 The details are not essential to the present discussion.
45 See his Lao-tzu chiao-klu, III, 643-74.

This content downloaded from

103.62.153.18 on Mon, 27 Feb 2023 07:38:21 UTC
All use subject to https://about.jstor.org/terms
 ON THE CHAPTER DIVISIONS IN THE LAO-TZU 511

On the other hand, Yen Ling-feng, who considered that th

sections of the present text are completely confused, rearrang
an attempt to recapture the 'original' text and ended up
fifty-four chapters divided topically into four sections: the
Tao, the 'principle' of Tao, the 'function' of Tao, and the' me
Finally, D. C. Lau, working within the framework of the prese
chapters, distinguishes 196 units in the text, each unit bein
kinds: (a) what he calls 'pre-existing passages', passages w
rhymed (and parallel, I might add); (b) exposition on those p
in prose, and often marked by a' therefore '; and (c) short, editor
e.g. 'Therefore he discards this one and takes the other '.Y
Whether D. C. Lau is right in making this three-fold distin
I do not wish to argue at the moment. For reasons that will
the pages that follow, I share Lau's conviction that the text a
was put together by an editor/compiler. But I am interested
the basic 'building blocks' the editor used to form his chapte
the case, as where the editor himself intended one chapter to
to begin.
The evidence of the Ma-wang-tui texts-dating as they do
200 B.c.--strongly suggests, with a few exceptions to be note
the arrangement of the 'original' text of the Lao-tzu, the s
and chapters, was for the most part what it is today, despite
of Yen Ling-feng and others.49 The lack of thematic developmen
of lines and phrases, was all in the text from the very outse
before us then is not, can we rearrange the text and put th
original order, but rather, can we read through the presen
sequence of material as it is, ignoring present chapter boundaries
in this way where the text naturally divides ? This is what I
In this I simply repeat the efforts of others like Wu Ch'eng

48 See Yen Ling-feng, Lao-tzu chang-chit hsin-pien f- i iJ ~j (T

wen-hua, 1954), Vol. I.
47 See D. C. Lau (tr.), Lao Tzu: Tao Te Ching (Baltimore and London
1953), especially the appendix, 'The nature of the work', pp. 113-74; al
introduction.

48 In the Ma-wang-tui texts, what to us is chapter 24 comes between 21 and 22, 41 is between
39 and 40, and 80 and 81 come between 66 and 67.
4 While he maintains the present eighty-one chapters and present chapter sequence,
J. J. L. Duyvendak, like Yen Ling-feng, considers that when the arrangement into eighty-one
chapters was done, some lines (being on slips of bamboo in bundles) were disarranged and
moved out of their original context (see Duyvendak, tr., Tao Te Ching, 5). Thus, following the
suggestions of Ma Hsii-lun and others, he makes some thirty shifts of material from one place
in the text to another (he indexes these on p. 16). Of these changes, only two are borne out by
the Ma-wang-tui texts: (1) the omission of the line 'It is only when one does not have enough
faith in others that others will have no faith in him' (Chan, 141) from the end of chapter 23;
and (2) the omission of the line 'Even the sage considers it a difficult question' (Chan, 228)
from chapter 73. But these omissions occur in other texts as well.

This content downloaded from

103.62.153.18 on Mon, 27 Feb 2023 07:38:21 UTC
All use subject to https://about.jstor.org/terms
512 ROBERT G. HENRICKS

There are three indications of chap

meaning, rhyme, and' characteristic endin
a distinct unit of meaning, thematically s
follows: the ideas are self-contained. W
sufficient indication of chapter change, s
rhyme, one does expect the rhyme to
tinuity of rhyme, therefore, could ind
rhymes in the Lao-tzu I have consulted tw
chiao-shih,51 and Bernhard Karlgren's s
I distinguish six common ways to br
(1) with the expression 'It's only becau
k ~f .. . ku j, or shih-i & L); 53 (2) w
therefore ' (i-ch'i iA ;if ... ku i') ; 54 (3
(shih wei ) ; 55 (4) with the expression
this' (ku ch'ii-pi ch'u-tz'u R & I) ;
do I know ... ? By this' (wu ho-i chi
and (6) with the conclusion 'Therefore
These are not sure signs of chapter close,
Working on this basis, for any one of t
of the three, I would conclude, very te
chapters in the present text are correc
8, 9, 10 (with the exception of the firs
to the end of 9), 11, 12, 13, 21, 23, 26, 32
50, 55, 56, 62, 66, 70, 73, 74, 76, 77, and 8
this conclusion, simply because space d
reasons for including chapters here wh
I will, rather, move on to the importa
cerning the other chapters. My refere
Chan's translation of the Lao-tzu.

1. Chapters 5 and 6. It is possible to see these as forming a single unit since

both deal with a common theme-the inexhaustibility of the empty centre.

o50 Although in the course of the analysis that follows I will indicate a fourth, ' common '
or ' standard' chapter format.
51 Chu gives his own opinions on the rhymes and also collates the conclusions of twelve
other scholars: Chiang Yu-kao j_ ff , Wu Yii : i, Ku Yen-wu j 43 g-, Chiang

Yung-chih _ 7j~ 2, Yao Wen-t'ien gJ 7 f]l, Teng T'ing-chen gj j iI, Li Keng-yiin

-, Liu Shih-p'ei ()J fj J, Hsi T'ung 4, Ch'en Chu ~ A, and Karlgren. He
also, on occasion, cites the work of Takeuchi Yoshio A [2) .
52 In GOteborgs H6gskolas Arsskrift, xxxvIii, 3, 1932, 1-45.
53 See chapters 2, 8, 15; also similar endings in 41, 60, and 75.
54 See chapters 7, 34, 66, and 71.
66 Occurring, not only at the end, however, in chapters 10, 14, 27, 30, 51, 52, 53, 55, 59,
68, and 74.
"* Chapters 12, 38, 72.
57 See chapters 21 and 54; also 50 for a similar pattern.
58 Chapters 24, 28, 29, 39, 44, 46, 47, 56, 58, 62, 63, 69, 70, 81.

This content downloaded from

103.62.153.18 on Mon, 27 Feb 2023 07:38:21 UTC
All use subject to https://about.jstor.org/terms
 ON THE CHAPTER DIVISIONS IN THE LAO-TZU 513

In chapter 5 this point is made with the image of the bellows, in chapter 6
with the valley. It will be remembered that 5 and 6 are combined in the
Wu Ch'eng and Ming T'ai-tsu texts, with 4, 5, and 6 combined in 'Lung-hsing
kuan'. And, since one of the editorial principles of the compiler of the text
seems to have been to string together chapters on a common theme, that too
would explain this connexion.59
Chapter 5 itself does not hold together very well. The first part, on the
theme of Heaven and Earth and the sage not being humane, seems to have
little connexion with the second, on the theme that Heaven and Earth are
inexhaustible, like a bellows. And I am persuaded by D. C. Lau that this
illustrates another common feature of the book, namely, that within many
chapters, the compiler brought together sayings whose only connexion is a
common theme or expression (here Heaven and Earth).60 The last two lines
(' Much talk will of course come to a dead end. It is better to keep to the
centre '-Chan, 107) are not connected by rhyme to what precedes, and Yao
Nai attaches these lines to the beginning of chapter 6, while Ma Hsii-lun sets
the lines off by themselves, making three different units of what is now
chapter 5.61 The rhyme at the end of 5 does not match that of 6, and chapter 6
is one of only a few chapters in the book to have a single rhyme throughout
(here the en sound-4, pEJ, *, etc.).62
2. Chapters 14 and 15. I believe 14 and 15 are separate chapters. But
attention needs to be called to the fact that chapter 14 ends up talking of
'holding to the Tao of old' (chih ku chih tao ai jjf ~), and chapter 15,
in the Ma-wang-tui version at least, begins with 'In antiquity those who were
skilled at practising the Way' (ku chih shan wei tao che - ;2 ).63
This is probably to be explained by the above noted editorial practice of
stringing together chapters with a common theme or expression. Nevertheless,
Yao Nai, for one, did combine 14 and 15.
3. Chapter 16. The first two lines of 16 in the Ma-wang-tui texts take the
form of definitions (text A, chih hsii chi yeh, shou ch'ing piao yeh I f~ 4j,

,- &perhaps
could ; text translate
B, chih'Going
hsii chi yeh, shou
to emptiness ching
is the centre;tumaintaining
yeh ~ i -tran-
!_); 64 we
quillity (reading jg as gi) is the surface' (text A), or 'To reach emptiness is
to go to the extreme; to maintain tranquillity is to oversee' (unless tu g is

59 For further discussion on this point, see below. Note that D. C. Lau (Lao Tzu: Tao Te
Ching, 23) draws attention to the fact that chapters dealing with the ' way of Heaven' tend to
be put together near the end of the book, chapters 71-81.
0so See D. C. Lau, Lao Tzu: Tao Te Ching, 166.
81 Ma Hsii-lun, Lao-tzu chiao-ku, III, 644.
82 For the theory that even ssu ( had an en sound at this time see Chu Ch'ing-ytian,
Lao-tzu chiao-shih, 17-18. For the rhymes of chapter 5 see ibid., 16.
63 See Ma-wang-tui Han mu po-shu , F J j & , " (Peking: Wen Wu, 1980),
I, 117, text B: text A has a lacuna. (This work hereafter cited as MWTHMPS.) But note
that Ma-wang-tui chapter 14 has 'hold on to the Way of the present' (J , g ).
64 MWTHMPS, I, 118.

This content downloaded from

103.62.153.18 on Mon, 27 Feb 2023 07:38:21 UTC
All use subject to https://about.jstor.org/terms
514 ROBERT G. HENRICKS

chia-chieh for tu g). Since chapter 15 en

(an 9) and not overflowing (pu-ying f; ;
lines of 16 as commentary on the end of
together very well without these chara
arguing against this is the fact that the fir
and the next two) are rhymed.65
4. Chapters 17, 18, and 19. On the basi
clear that chapters 17, 18, and 19 form o
should be read without a break. Kao H
on the basis of the Ma-wang-tui texts, h
18 and 19.66 That 18 follows from 17 is
(though not elsewhere) since what is now ch
a ku @, 'Therefore ',.67 Moreover, as we
bined early on in the Wu Ch'eng and Min
5. Chapter 20. This is self-contained w
line, @ Mj & (' Abandon learning and th
Chapter 19 ends with two four-ch
('Maintain plainness, Embrace simplic
desires', Chan, 132). That the opening li
and rhymes is a fact noted by many.68 The
making it the last line of the foregoing
Wu Ch'eng and Ming T'ai-tsu texts it is
chapter 20.70
6. Chapters 22 and 24. In the Ma-wang
us is chapter 24 comes directly before w
21, 24, 22, 23.71 The two (22 and 24) are
statements: 72

He who shows himself is not luminous.

He who justifies himself is not prominent.
He who boasts of himself is not given credit.
He who brags does not endure for long.

65 Karlgren (' The poetical parts in Lao-tsi ', 9) rhymes chi and tu g with tso 4 and
fu4.

66 See their article, 'Shih-t'an Ma-wang-tui Han mu

FE ff ' g 5f ( ( y, Wen wu, 1974, 11, p. 4.
67 See M WTHMPS, I, 118.
68 See Kao Heng and Ch'ih Hsi-chao as cited in n. 66. Also Wing Tsit-chan, The Way of
Lao Tzu, p. 135, n. 1. On the rhyming of yu * with p'u s and yi e see Chu Ch'ing-yiian,
Lao-tzu chiao-shih, 49.
69 Though Ma Hsii-lun (Lao-tzu chiao-ku, III, 649) puts it at the start of the present 19.
70 See Cheng-t'ung tao-tsang, xxx, 16826, and xIx, 15242.
71 See M WTHMPS, I, 119-20.
72 Tr. by Chan, The Way of Lao Tzu, 143. The first two lines are reversed in the Ma-wang-tui
texts.

This content downloaded from

103.62.153.18 on Mon, 27 Feb 2023 07:38:21 UTC
All use subject to https://about.jstor.org/terms
 ON THE CHAPTER DIVISIONS IN THE LAO-TZU 515

While 22 says specifically of the sage: a

He does not show himself; therefore he is luminous.
He does not justify himself; therefore he is prominent.
He does not boast of himself; therefore he is given credit.
He does not brag; therefore he can endure for long.

Kanaya Osamu & @ jt has already argued that the order of

particular, with 24 preceding 22, is better than the reverse.14 T
wang-tui arrangement of the chapters is preferable to the nor
Kanaya also argues for reading 24 and 22 as one, undivided chapte
very well be right. However, both chapters are set in a very com
pattern in the Lao-tzu, especially in part I, viz. 4-6 (or more) par
lines, followed by a prose conclusion, 'Therefore, the sage ...' o
'Therefore, the one who has the Tao ... (does so and so) '. This c
a case of two chapters brought together because of similarity of the

7. Chapter 25. This could divide into two chapters, the first be
summary of Taoist cosmology (' There was something undifferentiat
complete ... It may be considered the mother of the universe. I
its name; I call it Tao. If forced to give it a name, I shall ca
Chan, 144); the second being the matter of the Tao, Heaven, Ea
king being great, and one being modelled on the other. There is
connexion of ideas between the two and no rhyme connects the
'Therefore' at the start of the second part in most texts; but t
therefore in Ma-wang-tui.76 The connexion between the two sec
the fact that the Tao is named Great in the first part, while the
begins ' Tao is great, Heaven is great', etc.
8. Chapter 27. This is an odd chapter. It begins with the ' stand
i.e. five parallel, rhymed 7 lines ('A good traveller leaves no tr
A good speech leaves no flaws', etc., Chan, 147) followed by a 'Th
sage . . .' (here,' Therefore the sage is always good in saving men
and the prose conclusion has a characteristic ending (' This is call
the light', Chan, 147). But the chapter continues with 'Therefo
man is the teacher of the bad, And the bad is the material from wh

7 ibid., 139. Again, the first two lines are reversed in the Ma-wang-tui texts.

" Kanaya Osamu, 'Hakusho Rashi ni tsuite-sono shiryash6 no shohotek

= 7f- -cV ~D : ~~ ) ~ ? l JJ 1 [j , in Chi~goku tetsugakush
to mosalc . 4 g ~ ~ Z (Tokyo: S~bunsho, 1976), 1
75 ibid.
78 See MWTHMPS, I, 121.
7 Linguists are not agreed on the archaio rhyming of k'ai r (' A well-shut door needs no
bolt, and yet it cannot be opened ', Chan, 147) with the other finals in the series, chi ", che ,
ts'e @, and chieh . See Chu Ch'ing-yiian, Lao-tzu chiao-shih, 71, and Karlgren,' The poetical
parts in the Lao-tsi ', 12, for the various opinions. The Ma-wang-tui texts have ch'i "i instead
of k'ai; ch'i probably presents fewer problems. See M WTHMPS, I, 121.
VOL. XLV. PART 3. 36

This content downloaded from

103.62.153.18 on Mon, 27 Feb 2023 07:38:21 UTC
All use subject to https://about.jstor.org/terms
516 ROBERT G. HENRICIKS

may learn' (Chan, 147).7s This additional six-line section

called' ending: 'This is called the essential mystery

On the surface it is difficult for the reader to see any

any of the three sections of the chapter other than the fac
about 'good' (shan ) things.
9. Chapters 28 and 29. The body of chapter 28 is three
each with its own rhyme (' He who knows the male and k
etc.', 'He who knows the white and yet keeps to the bl
who knows glory but keeps to humility ... etc.', Chan,
ends 'He will be proficient in eternal virtue, And re
simplicity (uncarved wood) ', and this is followed by: 79

When the uncarved wood is broken up, it is turned into c

But when the sage uses it, he becomes the leading official
Therefore the great ruler does not cut up.

The three-line ending would seem to follow from wh

'uncarved wood' (p'u 4) is mentioned in both. But there
meaning between the end of chapter 28 and the three s
And the transition provided by the ' uncarved wood' is lost
texts where stanzas 2 and 3 of chapter 28 are reversed.80s
That chapter 29 is made up of two unrelated units and
as two chapters has already been argued by Kao Heng a
I wonder if we should not read the end of 28 with the begin
as one chapter ? In Chan's translation that would read: 8

(28) When the uncarved wood is broken up, it is turn

things.
But when the sage uses it, he becomes the leading official.
Therefore the great ruler does not cut up.
(29) When one desires to take over the empire and act on it (interfere
with it),
I see that he will not succeed.
The empire is a spiritual thing, and should not be acted on.
He who acts on it harms it.
He who holds on to it loses it.

The units read very well when combined. Moreover, both places talk about

78 ibid. The Ma-wang-tui texts have 'Therefore the good man is the teacher of the good '.
7 Chan, The Way of Lao Tzu, 149.
8o See MWTHMPS, I, 122.
81 See their article,' Shih-t'an Ma-wang-tui Han mu chung t'i po-shu Lao-tzu ', 4.
82 Chan, The Way of Lao Tzu, 149 and 151. For the second line the Ma-wang-tui texts
(MWTHMPS, I, 122) have 'is used' (I)) instead of'uses it' (Il Y). Thus the translation
should be ' But when the sage is used ', which is parallel to the uncarved wood being broken up,
and altogether makes better sense.

This content downloaded from

103.62.153.18 on Mon, 27 Feb 2023 07:38:21 UTC
All use subject to https://about.jstor.org/terms
 ON THE CHAPTER DIVISIONS IN TlE LAO-TZU 517

a 'thing' (ch'i g) which should not be interfered with, and government by

wu-wei , seems
chapter was to be the theme
put immediately after throughout.
28 because of It could
the be argued
' uncarved wood'that
con-this
nexion. Moreover, this probably explains why some later editor reversed
stanzas 2 and 3 of chapter 28-to make the end of 28 fit better wvith the
beginning.
In the second part of the present chapter 29 we have four parallel, rhymed
lines (' Among creatures some lead and some follow. Some blow hot and some
blow cold', etc., Chan, 151), followed by 'Therefore the sage discards the
extremes, the extravagant, and the excessive' (Chan, 151). Thus, this section
of the chapter taken by itself is in what I have noted to be a very common
chapter form.
10. Chapters 30 and 31. It is possible that 30 and 31 should be read together
as one. They are both about the use of weapons (ping A), and they both make
the same point-that warfare is an inauspicious thing, and that the good
general indulges in it with great sorrow and only when he has no choice.
Chapters 30 and 31 are combined, it will be recalled, in the Wu Ch'eng, Ming
T'ai-tsu, and Yao Nai texts.
11. Chapter 35. This ends by talking of the Tao as The Great and states
that 'All things come to it' (Chan, 160). It begins with 'Hold fast to the
great form, And all the world will come' (Chan, 162). The first four lines of 35
could hook on to the end of 34. But 34 seems self-contained. Moreover, new
rhymes are introduced (g and f against the A at the end of 34).3 Thus,
this is probably the start of a new chapter, placed here by the compiler because
of similarity in language and theme.
12. Chapter 36. The first eight lines of 36 are rhymed and parallel (' In
order to contract, It is necessary first to expand. In order to weaken, It is
necessary first to strengthen', etc., Chan, 164). The rhyme pattern in Ma-
wang-tui is ABAB, CCCC.84 This is followed by a characteristic ending,' This
is called subtle light' (Chan, 164), - uP. However, the chapter does
not end there, but continues: 85

The weak and the tender overcome the hard and the strong.
Fish should not be taken away from water.
And sharp weapons of the state should not be displayed to the people.

It is tempting to see this as someone's comment which has mistakenly

been copied into the text. Ma Hsii-lun sets these lines off from what precedes,
creating a separate chapter,"6 while Yao Nai has the last two lines stand alone.
But Karlgren, for one, sees a rhyme in the 'light' (minq l~I) of the 'This is

s* See Chu Ch'ing-ytian, Lao-tzu chiao-shih, 90 and 92.

54 That is, Asi ' (A--4-, B-- ), chang g, jo , ch'iang , then ch'ii , yit ,
to , and yi -. Also see Chu Ch'ing-yiian, Lao-tzu chiao-shih, 94.
s Chan, The Way of Lao Tzu, 164.
56 See his Lao-tzu chiao-ku, III, 655.

This content downloaded from

103.62.153.18 on Mon, 27 Feb 2023 07:38:21 UTC
All use subject to https://about.jstor.org/terms
518 ROBERT G. HENRICKS

called' line and the 'strong' (ch'iang g

which would argue for continuity.87
13. Chapters 40-43. We have seen sev
chapters in the series 39-44: 39 and 40
39, 40, and 41 are combined by Yao Nai
Li Yiieh; and 40-43 are read as one in
editions.

On the basis of the method of reading proposed here, that is, in terms of
ideas and rhyme, chapters 39 and 44 are self-contained, unrelated to what
follows and what precedes. Things are more complex in the case of 40-43. It
seems preferable to read the chapters in the order in which they occur in the
Ma-wang-tui texts : there, what to us is chapter 40 comes between what to us
are 41 and 42.8s The sequence is thus 41, 40, 42.
If we read the material in this way, two distinct possibilities emerge.
(a) Chapters 41, 40, and the first part of 42 (ending with the line 'The
ten thousand things carry the yin and embrace the yang ...', Chan, 176)
could form one continuous discussion of the nature of the Tao. The whole
thrust of 41 is to show that the Tao, like other great things, seems to be it
opposite, which is why people laugh when they hear of it. This point
summarized in the opening line of 40 : ' Reversion is the action of Tao. Weak
ness is the function of Tao' (Chan, 173). Then begins a more direct statemen
of what the Tao has done: 'All things in the world come from being. And
being comes from non-being. Tao produced the One,' etc. (Chan, 173 and 176
Then the end of 42, beginning with 'What others have taught, I teach als
(Chan, 176) might go with 43 to make a chapter: both emphasize the valu
of taking no action, and 43 ends with the same concern with teaching that we
find at the end of 42: ' Few in the world can understand the teaching without
words...' (Chan, 178).
This leaves us with the middle of 42 (' People hate to be children withou
parents . . .', etc., Chan, 176), lines which are similar in theme to those at th
end of 39. I am still undecided as to whether to tie this to the end of what
precedes or to the beginning of that which follows. Yao Nai has the end of
standing alone.
(b) The other way to break this section of the text up is suggested by
Kanaya Osamu. He would simply make two chapters of 41, 40, and 42, tying
the first two lines of 40 (' Reversion is the action of Tao. Weakness is the
function of Tao ') to the end of 41, and the last two lines of 40 (' All things in
the world come from being, And being comes from non-being ') to the beginning
of 42.89

87 See Karlgren, 'The poetical parts in the Lao-tsi ', 13. Also see Chu Ch'ing-yiian, Lao-tzu
chiao-shih, 94, where he cites the opinion of Ku Yen-wu. The last two lines are also rhymed:
yiuan jJ (water) rhymes with jen A (people).
88 See MWTHMPS, I, 102.
8' Hakusho R~shi ni tsuite ', 189.

This content downloaded from

103.62.153.18 on Mon, 27 Feb 2023 07:38:21 UTC
All use subject to https://about.jstor.org/terms
 ON THE CHAPTER DIVISIONS IN THE LAO-TZU 519

14. Chapter 46. It is difficult at first to see what the relationship might be
between the opening lines of 46 (' When Tao prevails in the world, galloping
horses are turned back to fertilize. When Tao does not prevail in the world,
war horses thrive in the suburbs', Chan, 181) and the four lines that follow
(' There is no calamity greater than lavish desires. There is no greater guilt
than discontentment', etc., Chan, 181). Yao Nai sets these opening lines off
from the rest, and text A of the Ma-wang-tui texts brackets them off from
what follows by means of punctuation.90
On the other hand, we have here another common chapter form in the
Lao-tzu, especially in part II, viz. illustration or exposition in prose followed by
further development or a moral in rhymed, parallel lines.' The point here
might be that war is ultimately the result of discontentment and the like.
I am inclined to leave 46 as it is.

15. Chapter 51. This looks very much like a chapter composed of two,
originally separate sayings (chapters ?) on the Tao, both of which begin in the
same way (' Tao produces them. Virtue fosters them'; and 'Therefore Tao
produces them and virtue fosters them', Chan, 190),92 and deal with exactly
the same theme-the production of things by the Tao and its subsequent
relationship to them. The ' Therefore' of the Wang Pi text at the start of the
second saying does not occur in the Ma-wang-tui texts. Instead, there is a
punctuation dot indicating the start of a new chapter or saying in text A.93
While the rhyme is not continuous from one saying to the other, the same
rhyme is used predominantly in both.94
16. Chapter 52. One could divide 52 into three sayings on the basis of
discontinuity of ideas and rhyme: (a) the first seven lines, from 'There was
a beginning of the universe' to ' Will be free from danger throughout his life-
time'; (b) the next six lines, from 'Close the mouth' to 'And to the end of
life there will be no salvation'; and (c) the last six lines, from 'Seeing what is
small is called enlightenment' to ' This is called practising the eternal '.95

90 Dots which seem to indicate chapter divisions, and which are placed at the beginning of
the chapter, are here found at the beginning and end of these opening lines. See MWTHMPS,
I, 103.
91 For the rhymes (yiL , tsu ,, te f, and tsu ,) see Chu Ch'ing-yiian, Lao-tzu chiao-shih,
121.

92 The Ma-wang-tui texts (MWTHMPS, I, 104) omit the 'virtue' (te ) from the second
of these lines.

93 See MWTHMPS, I, 104.

94 For the rhymes see Chu Ch'ing-yiian, Lao-tzu chiao-shii, 131-2. Three of the first four
lines are rhymed (with sheng 4f, hsing 7, and ch'eng JiJ). Then the second saying has an

AB 95pattern with
All translated sheng
by Chan, The4Way
-hais @,Tzu,
of Lao ch'ang -yirhymes
192. For the @, ch'eng
see Chu Jjj -shu ,, and yang -fu .
Ch'ing-yiian,
Lao-tzu chiao-shih, 134-5. The first part has one rhyme throughout: saih *i, mu :, mu ,
tzu T-, mu +, tai . The rhyme words in the second part are tui 9 (A), men J (B), ch'in

) (B), tui
ch'iang f, -, (A), ),shih
kuang yang(C), chiu s @.
3, ch'ang (C). In the third part we have just one rhyme: ming I1)j,

This content downloaded from

103.62.153.18 on Mon, 27 Feb 2023 07:38:21 UTC
All use subject to https://about.jstor.org/terms
520 ROBERT G. HENRICKS

Text A of Ma-wang-tui, however, pu

mouth', so apparently setting off the last
is also where Chan breaks the text in his translation.96
In the final analysis there is reason to see the whole as one continuous
unit since the lines near the end 'Use the light. Revert to enlightenment'
(Chan, 192) j 4 W, 4g $ j I7j, are parallel to the lines near the beginning
'And having understood the sons, Still keeps to its mother' (Chan, 192)

17. Chapters 57-61. These are all concerned with the correct way to rule.
In this they are distinct as a unit from anything said in 56 or 62. This could
conceivably be one long discussion on government, and the words 'rule',
' government ', and ' state ' are prevalent throughout. The opening lines read:
57 : ' Govern the state with correctness' (Chan, 201) JA ]E i ; 58 : ' When
the government is non-discriminative and dull, The people are contented and
generous' (Chan, 203) j1 R 3 @ , fl J W ; 59: 'To rule people and
to serve Heaven there is nothing better than to be frugal' (Chan, 205) )'f A
$f , A Y ; 60: 'Ruling a big country is like cooking a small fish'
(Chan, 207) j )z- I~ IJ ; and 61 : ' A big country may be compared
to the lower part of a river' (Chan, 208) } J 7i -i . But there is nothing
to prevent these, as far as the ideas or the rhymes are concerned, from being
five separate chapters on governing, strung together by the compiler because
of the similarity in language and theme.
Text A of the Ma-wang-tui texts is punctuated by a dot at the beginning
of our 57, setting it off from 56; there is no break indicated in the text between
60 and 61. For the rest-the connecting points of our chapters 57-58, 58-59,
and 59-60-there is no way of telling how these were understood by the copyist
of text A since the manuscript at these points is rotted away.97
It will be recalled that chapters 57 and 58 are combined in the Wu Ch'eng
and Ming T'ai-tsu texts: this is quite possible from a reading of the text,
though there is no connexion through rhyme.98 But the 'correct' (cheng IE)
and 'unorthodox' (ch'i 4 : Chan, 'surprise' in 'Operate the army with
surprise tactics ') of the opening of 57 are picked up again in the middle of 58
in the lines 'Is there no correctness (used to govern the world)? Then the
correct again becomes the perverse' (Chan, 203).
I also mentioned earlier that the 'Chieh-Lao' chapter of the Han Fei-tzu
since it comments in sequence on lines from chapters 58, 59, and 60, in between
its commentary on lines from 35 and 46,99 might reflect a reading of these
chapters as one. Such a reading is also possible on the basis of the ideas of the

96 See MWTHAMPS, I, 105, and Chan, The Way of Lao Tzu, 192.
9' See MWTHMPS, I, 106-7.
98 For the rhyme see Chu Ch'ing-yiian, Lao-tzu chiao-shih, 150, 153.
" See Han Fei-tzu chi-shih ~ (Taipei: Shih-chieh, 1963), I, 340-59.

This content downloaded from

103.62.153.18 on Mon, 27 Feb 2023 07:38:21 UTC
All use subject to https://about.jstor.org/terms
 ON THE CHAPTER DIVISIONS IN THE LAO-TZU 521

text. However, 59 appears to have only one rhyme throughout,

not connect with either 58 or 60.100
The Yen Tsun text divides chapter 58, appending the first pa
present 57, and attaching the rest to the beginning of 59. Ther
tinuity of rhyme between 57 and 58, nor between 58 and 59,101 w
argue against such an arrangement. But the break in 58 in the Ye
does occur at an interesting place, right after the lines ' Then the co
becomes the perverse, And the good will again become evil', that
where the 'correct' and 'unorthodox' of chapter 57 have bee
which is a neat way to tie things up. Arguing against this again,
the fact that Karlgren and others see a rhyme between the 'evil
the end of this line and the 'long time' (chiu #k) at the end of th
people have been deluded for a long time ', Chan, 203).1o2
18. Chapters 63-65. These are combined in the Wu Ch'eng
T'ai-tsu texts, and one can see why. Both chapters are on a singl
importance of dealing with issues while they are small and sim
letting things get out of hand. To be more precise, the second half o
'Prepare for the difficult while it is still easy', Chan, 212) and t
of 64 (to 'The journey of a thousand li starts from where one st
214) are on this theme, and it is these that Yao Nai combines.
Several things argue against the unity of 63 and 64. First, 63 h
chapter form, with exposition of a theme, here in prose, followed by
the sage' does so-and-so, and 64 seems to begin anew with four p
In addition, text A of the Ma-wang-tui texts has a dot at the begi
indicating separation. Lastly, there is no apparent connexion
though 63 ends with nan g, a rhyme which is resumed in lines 3
with p'an j and luan RL.
Within 63 itself, it is difficult to see the connexion between
five lines ('Act without action. Do without ado. Taste witho
Whether it is big or small, many or few, repay hatred with virtue ',
and the rest. Yao Nai and Ma Hsii-lun both distinguish the first
two separate sayings.
On the connexion of 64 and 65, this is suggested by text A of
which has no punctuation dot between the two and begins with k
'Therefore it is said' (that those who practise the Tao @ -)
other text begins this chapter in this way: text B of Ma-wang-
most other texts, 'Of old, those who practised the Way' (ku chih

100o See Chu Ch'ing-yiian, Lao-tzu chiao-shih, 156, then 153, 159. In 59 the rhym
se , fu 4, te @, etc.; at the end of 58 the rhyme is ssu and yao j ; at t
of 60, hsien s? and shen jI. Also see Karlgren, 'The poetical parts in the Lao-
101 Chapter 57 ends with shilh $,fu , y1t , and p'u 4. Chapter 58 begins w

J, ch'un ch'un '. . See Chu Ch'ing-yiian, 150, 153, and Karlgren, 17.
102 See Karlgren, 'The poetical parts in the Lao-ts ', 17. See also Chu Ch'ing-
chiao-shih, 153.
103 See MWTHMPS, I, 108.

This content downloaded from

103.62.153.18 on Mon, 27 Feb 2023 07:38:21 UTC
All use subject to https://about.jstor.org/terms
522 ROBERT G. HENRICKS

- Yj g ). Moreover, there is no obvio

(' People are difficult to govern because
Chan, 216) and the end of64 (' Therefore
He does not value rare treasures', Chan,
study and thought.

19. Chapters 67-69. 67-68 are combined

and 67-69 are combined in the Wu C
Yao Nai affirms this latter arrangement

three
is treasures',
obvious: Chan,
towards the end219;
of 69RW
we _ ). The
read reason
'Making lightfor connecting
of the 67-69
enemy will
destroy my treasures' (Chan, 222) i ~ . ~ W. And reading reveals
a continuity of ideas. These chapters make the same point as 30 and 31 on
the sage's attitude toward war; he has a deep love for his fellow man and
approaches conflict with great sorrow. Thus his three treasures, 'deep love',
'frugality ', and' not daring to be ahead of the world' (Chan, 219), are effective
in military matters (' For deep love helps one to win in the case of attack,
And to be firm in the case of defence ' (67), Chan, 219, and ' One who is skilful
in using men puts himself below them' (68), Chan, 221).
The connexion between 67 and 68 is affirmed by text B of the Ma-wang-tui
texts in which 68 begins with the words ' Therefore, one who is skilled at being
a knight' (& A ~ ? ).104 Text A does not have the 'Therefore', but it
does not punctuate between 67 and 68. There is, however, punctuation in the
form of a dot in text A between 68 and 69.105
20. Chapters 71 and 72. A syntactical connexion can be seen between the
end of 81, i ch'i ping-ping, shih-i pu ping L ; 9 9, B 9 (' Because
he [the sage] recognizes the disease to be disease, he is free from it ', Chan, 225),
and the beginning of 72, min pu wei-wei, tse ta-wei chih j 4 f- , flJ f E
(' When the people do not fear what is dreadful, Then what is greatly dreadful
will descend on them', Chan, 226). And as I have pointed out elsewhere,106
text A of the Ma-wang-tui texts punctuates in 72 immediately after the line
'Then what is greatly dreadful will descend on them', possibly connecting
that which preceded with chapter 71.
Against this connexion is the fact that chapter 71 reads well as it is, the
ideas are self-contained. Moreover, 72 is about oppression and 'dreadful'
things, which follow from the opening lines. Finally, Chu Ch'ing-yiian argues
for a connexion of rhyme in the beginning and end of 72.107
21. Chapters 78 and 79. These are clearly two distinct chapters. There is,
however, a problem with the last line of 78, cheng-yen jo fan YE & P t

1o4 ibid., 110.

105 ibid.

106 Henricks, 'A note on the question of chapter divisions in the Ma-wang-tui manuscripts
of the Lao-tzu ', 50.

o'07 See Lao-tzu chiao-shih, 184. The rhyme words are wei g, chih @, chih ], ai , and
kuei .

This content downloaded from

103.62.153.18 on Mon, 27 Feb 2023 07:38:21 UTC
All use subject to https://about.jstor.org/terms
 ON THE CHAPTER DIVISIONS IN THE LAO-TZU 523

(' Straight words seem to be their opposite', Chan, 236). This clea

with theNai
and Yao opening lines
texts are of 79
surely (J, @,in@,
correct ,,).xOs The
regarding thisWu Ch'eng,
as the Ming
opening T'ai-tsu,
of 79.
As such it is perhaps better translated as 'Words of correction have the opposite
effect', since 79 continues 'To patch up great hatred is surely to leave some
hatred behind' (Chan, 237).
Text A of the Ma-wang-tui texts seems to affirm this arrangement since no
punctuation occurs between the lines in question.109
22. Chapters 80 and 81. The Ma-wang-tui texts place what are our chapters 80
and 81 between our 66 and 67.110 If there is a reason why this arrangement is
superior to that of having them follow 79, I have not yet discovered what it is ;
nor has Kanaya Osamu.nl
Chapter 80 is self-contained, clearly set off from 79 and 81, as they are, in
the Ma-wang-tui sequence, from 66 and 67. The beginning of 81 (to 'He who
has extensive knowledge is not a wise man ', Chan, 240) seems to have little in
common with the end (from 'The sage does not accumulate for himself',
Chan, 240). That these are two separate sayings (chapters ?) might be con-
firmed by text A of Ma-wang-tui which punctuates between the two.12 Against
making this break there is the argument of Chu Ch'ing-yiian that there is a
continuity of rhyme.113

I have now accounted for all but three of the chapters, namely, 53, 54, and
75. Ma Hsii-lun 114 sets the first five lines of 53 off from the rest as a separate
chapter (see also above on Yao Nai). The break comes between 'But people
are fond of bypaths' (Chan, 194) and 'The courts are exceedingly splendid'
(Chan, 194). It is true that the connexion is not obvious between the opening,
unrhymed lines on staying on the 'road' (Tao) and leaving it, and the
following rhymed,115 parallel lines, describing the decadence of contemporary
courts. Nevertheless, this format-unrhymed, prose exposition followed by a
rhymed, parallel summary or description-as we have seen with chapter 46,
is a common one in the second part of the Lao-tzu. Moreover, the symbolism
of the 'road', the Way, of the opening lines, is resumed in the final lines of
the chapter, 'This is robbery and extravagance [tao i]. This is indeed not
Tao (the Way) [] ' (Chan, 194). The decadence of contemporary royalty is
thus probably meant as an example of what happens when people stray from
the Way.
On 54, Yao Nai separates the first three lines from the rest (to ' Thus from
generation to generation his ancestral sacrifice will never be suspended',

208 ibid., 195. 109 See MWTHMPS, I, 113. 110o ibid., 109-10.
111 See his ' Hakusho Rashi ni tsuite ', 190.
112 See MWTHMPS, I, 109.
11a See his Lao-tzu chiao-shih, 200. He argues that 'knowledge' (chih )
rhymes with 'accumulate' (chi ) and 'more' (to g) in the second.
114 See his Lao-tzu chiao-ku, HI, 662-3.

1i5 See Chu Ch'ing-yiian, Lao-tzu chiao-shih, 137. The rhyme w

hsf , and yE .

This content downloaded from

103.62.153.18 on Mon, 27 Feb 2023 07:38:21 UTC
All use subject to https://about.jstor.org/terms
524 ON THE CHAPTER DIVISIONS IN! THE LAO-TZU

Chan, 195). The first three lines share a rhyme which sets them off from the
rest.116 It is, moreover, difficult to see a continuity of ideas here.
Finally, text A of the Ma-wang-tui texts punctuates in the middle of 75,
between the lines 'Therefore they are difficult to rule' (Chan, 232) and 'The
people take death lightly because their ruler strives for life too vigorously'
(Chan, 242).117 Kanaya Osamu considers this to be a proper break.118s I disagree.
There is a difference in theme between the two parts of the chapter, that is,
between the first section on the people being difficult to rule and the second on
them taking death lightly. But the chapter is composed of three, syntactically
parallel statements about the people (R g , J 2 , A ~ M [E)
followed by a standard ending formula, fu-wei jJ If.

III. Conclusions
Much that has been said in part II is speculative and inconclusive.
can hope for little else, given the nature of our materials. Other sch
repeating the experiment of reading the text as it stands and searchi
natural breaks, will see connexions where I have not and breaks in wha
left intact. There is a further problem in the fact that the compiler of th
seems to have brought together in single chapters sayings similar in them
language, as well as stringing chapters together on the same principle,119
makes it difficult to distinguish between 'chapters' and 'sayings'. Was
the compiler's mind two 'chapters' on the Tao, or two similar 'pa
combined to make a chapter ?
I have not solved the problem of where to divide the text correctly
the less, taking the evidence reviewed here as a whole-the record of
others have divided the text, the Ma-wang-tui punctuation, and the re
a reading from the standpoint of the continuity and discontinuity of idea
rhyme-it is difficult not to conclude at least that:
1. chapters 17, 18, 19, and the first line of 20, should be read as one,
2. chapters 67, 68, and 69 should also be read as one, though we migh
to start in 67 with the line 'I have three treasures ',
3. 30 and 31 might also be combined,
4. the last line of 78 should clearly be the first line of 79,
5. the Ma-wang-tui order of 21, 24, 22 is the correct order to follow, and,
6. I would also argue strongly in favour of combining the end of 28 w
beginning of 29.
Still to be sorted out are the connexions between chapters 40-43,
and 63-64. In the end, I come very close to affirming the sixty-eight
version of the text established by Wu Ch'eng and followed by Ming T

11 ibid., 139. The rhyme words are pa , t'uo a, and cho .

117 See Mi WTHIMPS, I, 112.
18s ' Hakusho R6shi ni tsuite ', 185.
119 I noted earlier the 'way of Heaven ' chapters in 71-81; chapters 73-76 all deal i
way with death.

This content downloaded from

103.62.153.18 on Mon, 27 Feb 2023 07:38:21 UTC
All use subject to https://about.jstor.org/terms