The Genesis of An Icon The Taiji Diagram

The Genesis of an Icon:
The Taiji Diagram's

Early History
FRAN?OIS LOUIS
The Bard Graduate Center
taiji tu ;MSH, or "Diagram

name of the Supreme Ultimate,"
has come to refer to a number of Chinese cosmological images
THE
that explain the concept of a primal entity and its generative prin
of these taiji diagrams are associated closely with the
ciple. Most
individual thinkers who conceived or perpetuated them. But some
entered the learned discourse on cosmology at unknown points in
time, and their intellectual provenance is shrouded in myth and leg
end. The following study aims to clarify the provenance of one of
these It consists of the eight tri grams sur
diagrams. (bagua A#)
rounding a circle that illustrates the taiji (fig. 1). This circle is divided
into two intertwined dark and light halves to represent yin and yang.
Near are two dots, which are usually
its center interpreted as depict
ing yang within yin and yin within yang respectively.

Ming scholars, who were the first to discuss the history of this di
agram, generally considered it a divine revelation from the remote
past and believed that sages such as the Duke of Zhou and Confucius
still knew about its value for divination. The scholars recognized
references to the diagram by these sages in certain passages of the
But at some later point in antiquity, they presumed, the
Yijing.
This has gone through numerous
paper incarnations and I would like to thank those who
have commented
critically on the manuscript and helped prepare it for submission toHJAS:
Anne McGannon, Craig Clunas, Marion Lee, and the anonymous readers of HJAS.
145
146
JL& jfc& 4fc >>t

? 4
ir
1
v?* ^
f
f ???fr
.-urn^'i:
^7
?i
?l^tjA.iS^ *j*j?*
it
A* fl?Tc *!*
f?
flW-fi^-f-?lMÎ it Jt?I
Fig. 1. "The Taiji Diagram of the Heart of the Changes as Mysteriously Revealed by Fuxi,!
from Lai Zhide, Yijing Laizhu tujie (1688; rpt., Chengdu: Ba Shu shushe, 1989), p. 553.
THE TAIJI DIAGRAM'S HISTORY 147
diagram somehow vanished, for no references to it could be found

in texts of the Han, Six Dynasties, or Tang periods. Song sources,
on the other hand, seemed to indicate that the taiji tu had (re-)
emerged in Daoist circles of the tenth century, and that it was trans
mitted in secret by alchemists as part of their esoteric doctrine. Yet
because the diagram could not be verified in texts from this period,
some of the more discriminating Ming reviewers concluded that it
must have remained secret during the Song as well. Even the emi
nent and criticalQing scholar Hu Wei #]iff (1633-1714), who in
the 1690s conducted one of the most comprehensive studies on the
history of cosmological diagrams, subscribed to the secret trans

mission explanation.1 Hu's research
quoted today, is widely and the
taiji tu is therefore generally considered to have been passed down
secretly by certain Daoist adepts and Neo-Confucian thinkers of the
period before becoming more widely known in the fourteenth

Song
century.2 The critical opinions of later Yijing experts such as Zhang
Huiyan 5S?H however, who felt that secret trans

(1761-1802),
mission theories were unconvincing attempts to come to terms with
a disconcerting lack of tangible evidence, are rarely
acknowledged.3
According to Zhang, the taiji tu grew out of the cosmological spec
ulations of early Ming scholars.4
Why is the early history of the taiji tu so elusive? And how can
we locate the provenance of this image, which was celebrated by
men such as Hu Wei as a revelation of the prime mysteries of the
1
Hu Wei, Yitu mingbian HI8$?|#, Congshu jicheng chubian, vols. 438-439 (Beijing:
Zhonghua shuju, 1985), p. 81.
2 see Zhang
For summaries of the most frequent versions of secret transmission theories,
Qicheng, ed., Yixue da cidian Hr<P^C??A (Beijing: Huaxia chubanshe, 1992), pp. 486ff.; and
Wu Hua, ed., Zhouyi da cidian ^^^Ci?? (Guangzhou: Zhongshan daxue chubanshe, 1993),
pp. 719-20. The most recent Western contribution to the history of the image does not go
beyond early Qing scholarship and mainly relies on the important study of Isabelle Robinet
on the concept of taiji, which, however, is concerned with pre-Ming issues and only hints at
the diagram's history in passing, see Stephen Little, Taoism and the Arts of China (Chicago:
The Art Institute, 2000), p. 131; Isabelle Robinet, "The Place and Meaning of the Notion
of Taiji in Taoist Sources Prior to the Ming Dynasty," History of Religion 29.4 (1990): 374-76,
388.
3
Zhang Shanwen, Zhouyi cidian )r1 Hi?ft (Shanghai: Shanghai guji chubanshe, 1992), pp.
99-100.
4
^Mi&ff? (1888), inWuqiubeizhaiYijingjicheng
Zhang Huiyan, Yitu tiaobian feJcffiSJs^
KlJ5Jc vol. 146 (Taipei: Chengwen chubanshe, 1976), p. 23b.
148 FRAN?OIS LOUIS
Yijing, of the Daoist Huang-Lao School, and of alchemical practice

all in one?5 This study proposes to abandon the traditional search
for a specific inventor, which has dominated explanations of the dia
gram's origin. Instead it will view the traditional explanations as

evidence for the construction of an intellectual pedigree necessary
for the diagram's acceptance in the cosmological discourse of the
Ming and early Qing periods. The murkiness of the taiji tu's early
history is seen here as the result of understanding the diagram as a
primeval unit instead of recognizing its main constituent elements?

the trigram circle and the dynamically divided yin-yang circle?as
separate entities with an early history independent of each other.
Categorically different in their pictorial qualities, these two graphic
signs, it is argued here, communicate on different cognitive levels

and originally functioned in disparate systems of visual expression
and symbolic relevance. While the trigram circle was a prime sym
bol essential for divination and cosmological analysis, the yin-yang
circle evolved in the context of decorative imagery and became a
standard symbol with a characteristic iconography
diagrammatic
only after with the
its combination trigram circle. Textual data point
to the fourteenth century as the time when these two graphic enti
ties were first combined, but the earliest material evidence readily
available today dates only to the sixteenth century, reflecting the
increasing popularity of the image toward the end of the Ming
period.
THE ICONOGRAPHY OF THE TAIJI TU
Although its name as well as the design of the central circle dif
fered somewhat from case to case in Ming and early Qing illustra
tions, the basic of the taiji diagram remained the same
iconography
(figs. 1, 11, 12, 19, 20, 21). It is always shown with the white half
of the central circle towards the top, the whirl rotating clockwise,
the two antithetical dots near the center, and the eight trigrams lo
cated in the same positions. A closer look at the symbolism of these
specific iconographie elements shows that the design of the central

swirls and the positions of the trigrams are carefully coordinated.
5
Hu Wei, Yitu mingbian, p. 81.
qian
? ? S ggen S
kun i$
Fig. 2. The circular xiantian arrangement of the trigrams (Fuxi sequence).
In the taiji tu the trigrams are arranged according to the so-called

xiantian %Ji or "prior to heaven" system. Unlike other ways of dis
playing the trigrams in a circle, the xiantian circle depicts the wax
ing and waning of yin and yang while at the same time pairing each
trigram with its respective opposite across the circle (fig. 2). In this
arrangement the trigrams are understood to illustrate the interac
tion between heaven and earth and the creative principles underly
ing the cycles of nature, such as the phases of the moon, the daily
and yearly course of the sun, and the seasons. The trigrams in the
xiantian circle are read in a specific sequence, which starts with qian
$? at the top. Because broken lines symbolize yin and solid lines yang,
qian represents pure yang, being made up of three solid lines. It

stands for heaven, light, or the south, and is consequently placed
at the top of the trigram circle, which can be read like a map.6 The
6
Traditional Chinese maps usually depict south on top, north at the bottom, east on the
left, and west on the right. In the 1688 illustration (fig. 1) the reading according to the direc
tions of the compass is indicated in the text below the diagram. Lai Zhide Jfc5$\W>, Yijing
Laizhu tujie BUFABA? (Chengdu: Ba Shu shushe, 1989), p. 553b.
150 FRAN?OIS LOUIS
xiantian trigrams are then read counterclockwise following their grad
ually decreasing yang valence, through dui j? and li 8|, which both
consist of two thirds yang and one third yin, to zhen H, which con
sists of one yang line at the bottom and two yin lines at the top.
Because these four
trigrams all have a yang line at the bottom, they
are considered to be essentially yang. The yin trigrams all have a
broken yin line at the bottom; they are still read in decreasing yang
increments, but now the sequence evolves in a clockwise sense start
ing with xun ft in the southwest (top right) and going through kan
%Kand gen S to kun i$. This shift in direction is the result of hav
ing the yin trigrams arranged in opposition to the yang trigrams on

the other side of the circle.
The same information o? yin and yang waxing and waning in jux
taposition is also visualized in the design of the swirls of the central
taiji circle. from the top to the bottom, while yin de

Yang decreases
creases to the top. What
from the bottom the trigrams express in a
conceptual and numerical manner, the central design visualizes in

a much more illustrative mode. Essentially, the trigrams lay out in
eight points what is traced as a smooth curve by the dividing line
between the yin-yang swirls. Several Ming and early Qing depictions
of the taiji tu explain the exact correspondence between the xiantian
trigrams the design of the taiji circle. The
and 1688 print illustrated
in Figure serve as our reference here.7 Right next to the pure
1may
yang trigram qian at the top, the yang swirl is white all the way from
the circle's edge to its center. The widest part of the black yin swirl,
on the other hand, corresponds to the trigram kun, which represents
the or north. The of the cor
pure yin, earth, darkness, explanations
respondence between the swirls and the trigrams

directly are written
behind the trigrams. Underneath qian is written
"pure white?pure
yang," underneath kun, "pure black?purejym."8 Next to dui in the
southeast, the central circle matches the trigram's composition by
showing "some black and a lot of white." The black and white dots
near the center of the circle are seen as visual paraphrases of the
7
This illustration appears in the Yijing Laizhu tujie, an enlarged edition of Lai Zhide's
1599. Cf. Larry James Schulz, "Lai Chih-te (1525-1604)
{\S2S-\W*)ZhmyijizhuM%iW?Lo?
and the Phenomenology of the "Classic of Change" (I Ching)" (Ph.D. diss., Princeton
University, 1982), pp. 43, 251-56. See also figure 20.
8
Lai, Yijing Laizhu tujie, p. 553b.
trigrams li and kan, which mark the east-west axis. Li, in the east,
consists of a yin line between two yang lines. Within the central cir
cle, this configuration is represented by the black dot within the
white tip of the yang swirl?or as the explanation in the 1688 print
states, "li: yin is within two opposites?the black within white." The
case of kan is just the opposite: the white dot is within the black tip
on the left.
The connection Ming and early Qing commentators saw between
the design of the central taiji circle and the yin-yang code of the xian
tian trigrams must be recognized as a primary reason for the diffi
culties in determining the origin of the taiji diagram. For viewing
the taiji diagram as a coherently designed unit allowed commenta
tors to explain its provenance as the provenance of either of its com
ponents. The most influential of these explanations equated the
discovery of the taiji tu with that of the xiantian trigram circle,9 pre
sumably because Chinese commentators considered the xiantian tri

grams, not the taiji circle, to be the semantically dominant element
of the taiji diagram. The determinant role of the trigrams is not sur
prising. The eight trigrams have enjoyed prime status as explana
tory symbols of nature's concrete forces ever since the Yijing became
the standard source for divination
and cosmological speculation.
The only other symbols of equal importance in Chinese cosmolog
ical reasoning were numbers. More telling than words, trigrams
were seen as abstractions of the structure of the
revelatory universe,
as they were primary images (xiang ^) designed at the beginning of
civilization by Fu Xi f??, primeval sage and China's first creator
of culture.10
9
This view is summarized and adopted, for example, in a very thorough and useful new
study on the early reception of the Yijing by Hermann G. Bohn, Die Rezeption des Zhouyi in
der Chinesischen Philosophie, von den Anfangen bis zur Song-Dynastie (M?nchen: Herbert Utz Verlag,
1998), pp. 138-39.

10
On Fuxi as the creator of the trigrams, see Xici B2. For summaries of recent theories
concerning the origin of the trigrams, see Paul Fendos, "Fei Chih's Place in the Development
of'I Ching' Studies" (Ph.D. diss., University of Wisconsin), 1988, pp. 26-58; Dominique
Hertzer, Das alte und das neue Yijing: Die Wandlungen des Buches der Wandlungen (M?nchen: Die
derichs, 1996), pp. 137-41; and Mark Edward Lewis, Writing and Authority in Early China
(New York: SUNY Press, 1999), pp. 281, 471 n. 135.

152 FRAN?OIS LOUIS
THE XIANTIAN TRIGRAM CIRCLE
Theyin-yang valence of the xiantian trigrams, and thus the specific

sequence in which these trigrams are read, comes about through a
process that is simultaneously arithmetic and graphic and is best
a diagram is read from
explained through (fig. 3).11 The diagram
the bottom up and shows taiji separating into yang and yin (yang is
represented by a solid line on the right and yin by a broken line on
the left). These symbols are divided again into their respective yang
and yin complements, and these are divided once more. The result
ing trigram permutations are then read from bottom to top and from
ft A
8kun 7 gen 6 kan 5 xun 4 zhen 3 li 2 did 1 qian
tai yin shao yang shao yin tai yang
taiji
Fig. 3. The creation of the xiantian trigram sequence.
11
This diagram is based on the example in juan 10 of Huang Zongxi IftTjsWt (1610-1695),
Song Yuan xue'an ^Tt^^ft (Beijing: Zhonghua shuju, 1986), vol. 1:414, where it is credited
to Shao Yong SPÎ (1012-1077). Zhu Xi (1130-1200) had already described and illustrated
this type of diagram, see Huang Zongxi, ibid. 1:387-91; Bohn, Die Rezeption des Zhouyi, p.
449.
yang to yin. This system of binary permutations is justified by a

famous passage in the Xici zhuan ^|?fl|, or "Commentary of the
Appended Statements," one of the early Yijing commentaries, most
likely written during the third or early second century B.C.12 The
passage states
that "the taiji or Supreme Ultimate13 generates the
two^z Hi, or modes [i.e., yin andyang\. The two modes generate the
four xiang, or images.14 The four images then generate the eight tri
grams, gua," which portend good fortune and misfortune. And
"good fortune and misfortune constitute the great field of action,"

or, as we might call it, social life.15 Interpreting this passage with
combinatorial logic results in a linear sequence of the trigrams which
are divided into a yang and a yin group, With yin continually wax
ing and yang continually waning.

By the twelfth century the xiantian trigram order was hailed as a
sublime, naturally perfect arrangement, for it provided the most
systematic known way of organizing the trigrams. To leading

twelfth-century intellectuals, most notably Zhu Xi %kM (1130-1200),
this order, whether in its linear sequence or in the inverted pairing
of the circle, represented a natural principle of organization, corre
sponding to the original conception of Fu Xi.16 Southern Song
12 see Willard
On this commentary, J. Peterson, "Making Connections: 'Commentary of
the Attached Verbalizations' of the Book of Change," HJAS 42.1 (1982): 67-116. For a sum
mary of all the Yijing appendices, see Fung Yu-lan, A History of Chinese Philosophy (Princeton:
Princeton University Press, 1952), vol. 1:379-95.
13 as "Supreme Ultimate"
The term taiji has been translated (Bodde, transi, of Fung Yu
lan, A History of Chinese Philosophy, vol. 2:435 and passim); "Great Primal Beginning" (Baynes,
transi, of Richard Wilhelm, The I Ching or Book of Change [Princeton: Princeton University
Press, 1950], p. 318); "Great Ultimate" (Richard John Lynn, The Classic of Changes. A New
Translation of the I Ching as Interpreted by Wang Bi [New York: Columbia University Press,
1994], p. 65); "Supreme Pole" (Joseph Needham, Science and Civilisation in China, vol. 2, sec
tions 8-18 [Cambridge: Cambridge University Press, 1956], p. 460), and "Grand Terminus"
or "Great Extreme" (James Legge, / Ching. Book of Changes [1899; rpt., New York: Dover,
1963] pp. 13, 373). Interestingly the term does not appear in the earliest extant version of
the Xici zhuan, the manuscript unearthed in tomb no. 3 at Mawangdui and datable prior to
168 B.C. the term daheng ^?L,
Instead "Great Constancy," is used, which underscores the
concept in the world principle.

of time See Edward L. Shaughnessy, / Ching. The Classic of
Changes (New York and Toronto: Ballantine Books, 1996), p. 198.

14
The word xiang has been translated variously as "emblem," "image," "representation,"
or "figure." See Peterson, "Making Connections," 80-81.
15
Xici, A 11.
16
Zhu Xi, Zhuziyulei #c^tg^ (Beijing: Zhonghua shuju, 1986), 65.1612-19 ("The xian
tian diagram drawn with Fuxi's trigrams"); Zhang Shanwen, Zhouyi cidian, pp. 277-78.
154 FRAN?OIS LOUIS
Neo-Confucian writings thus promoted it as a reflection of the most

authentic layer of meaning encoded in the Yijing, implying that it
also reflected most accurately the principles of nature. Although the
Yijing actually makes no explicit reference to the xiantian sequence
at all, late Song scholars nevertheless claimed that the circular xian
tian arrangement was referred to in a passage of the Yijing's Shuogua
|ft#fi| commentary: "Heaven [qian] and Earth [kun] establish posi
tions. Mountain [gen] and Lake [dui] circulate their material force.
Thunder [zhen] and Wind [xun] give rise to each other. Water [kan]
and Fire [li] refuse to destroy each other. In this manner the eight
trigrams alternate with each other, so that to enumerate what has

passed, one follows their progress, and to know what will come, one
moves backward through them."17 This passage is far from clear,
however, and interpreting it as a description of the xiantian circle?
as happens in the still influential Yijing translation by Richard
Wilhelm?is far from convincing.18 The order in which the trigrams
are mentioned here does by no means correspond to their positions
in the xiantian circle. Originally the passage may well have been
understood as a guide to interpreting those four hexagrams which
were composed of the complementary trigrams (i.e., nos. 12, 41,
32, and 63).19
Among the several trigram sequences described in the Shuogua

commentary, only one explicitly arranges the trigrams in a circle
according to the points of the compass. Known since the Song period
17
Shuogua, 3. The first publication which links this passage to the xiantian trigrams is appar
ently Zhu Xi, Zhouyi benyi ffi^î^ (Siku quanshu edition), tu, 5. However, this book was
first printed in 1265 and the authenticity of Zhu Xi's authorship of the introductory illus
tration, where this statement appears, is uncertain. For Zhu Xi's commentaries on this Shuogua
passage, see Zhuziyulei, 77.1971-72, where no reference to the xiantian trigrams is found. For
a fine survey of Zhu Xi's approaches to the Yijing, see Joseph Adler, "Chu Hsi and
Divination," in Kidder Smith jr. et al., Sung Dynasty Uses of the I Ching (Princeton: Princeton
University Press, 1990), pp. 169-205.

18
Wilhelm, The I Ching, pp. 265-66. More recently this view has been propounded again
by Schuyler V. R. Cammann, "The Eight Trigrams: Variants and their Names," History of
Religions 29 (1990): 301-2.
19
Legge had already expressed his unease with the xiantian interpretation of this passage,
see Legge, / Ching, p. 424. For alternative translations, see Shaughnessy, / Ching, p. 219;
Lynn, The Classic of Changes, pp. 120-21; and Kidder Smith Jr. and Don J. Wyatt, "Shao
Yung and Number," in Kidder Smith Jr. et al., Sung Dynasty Uses of the I Ching, p. 118.
kun J$
zhen dui ?
>\8
*"*" wmm *
gen H qian
kan jfc
Fig. 4. The circular houtian arrangement of the trigrams (King Wen sequence).
as the houtian \%^ or "After Heaven" order (fig. 4), this arrange
ment was understood to describe the cyclical changes of the seasons
and days and, by extension, explain the workings of the world.20
The trigram zhen represents spring and the beginning; its position
on the circle is east, and from there the sequence is read clockwise.
The most recognizable elements in this constellation are li in the
south and kan in the north, symbolizing fire and water respectively;
20
Shuogua, 5. Thispassage reads in the translation of Lynn, Classic of Changes, p. 121 (quoted
here without the annotations): "The Divine Ruler comes forth in Zhen and sets all things in
order in Sun, makes them visible to one another in Li, gives them maximum support in Kun,
makes them happy in Dui, has them do battle in Qian, finds them thoroughly worn out in
Kan, and has them reach final maturity in Gen. . . . Zhen corresponds to the east. . . . Sun
. . .is the
corresponds to the southeast. Li. trigram of the south." Cf. Legge, / Ching, pp.
425-26; Wilhelm, The I Ching, pp. 268-71. For an insightful exposition of possible meanings
of this sequence based on explanations of the Yijing in the Han apocrypha, see Fung, A History
of Chinese Philosophy, 2:102-6.

156 FRAN?OIS LOUIS
they are the only trigrams positioned opposite each other as a com
plementary pair, one representing yang, the other yin. The other
trigrams, though not paired as opposites, are still ordered around

li and kan so as to form and a yin group respectively
a yang (fig. 4).
Their yin-yang valence, however, is assessed according to a differ
ent, much simpler principle than that found in the xiantian system,
simply working with even (i.e. yin) and odd (i.e. yang) numbers. To
determine a trigram
whether is yin or yang according to the Yijing,
one counts of lines (a yang line is one, a.yin line two),
the number
then tallies whether the total is an even or odd number.21
It was this trigram cycle that served as the standard directional
and cosmological circular arrangement in early China. Referred to
in Han texts as the "proper order,"22 it is, with minor variations the
only directional trigram sequence found on archaeological material

up to the early Southern Song period. It appears most frequently
on objects whose function closely to cosmological
related specula
tion, to the transcending of life and death, and the invocation of
good fortune, such as late Han and Six Dynasties "diviner's
boards,"23 stone lids of tomb epitaphs from the late Six Dynasties
through the Five Dynasties period (fig. 5),24 late Six Dynasties
Buddhist stupas,25 and, most prominently, bronze mirrors dating
21
This manner of determining the yin-yang valence can be observed throughout the Yijing,
see e.g. Shuogua 10.
22
Fung, A History of Chinese Philosophy 2:103-4.
23 one from an Eastern Han
Two diviner's boards, tomb and the other in the Shanghai
Museum datable to the Six Dynasties, have part of the houtian circle trigrams inscribed on
them, see Lian Shaoming j?SO?n, "Shipan zhong de simen yu bagua" ?SÔ^BSP^HA
&, Wenwu 34. For wider-ranging discussions of such boards, see the articles by
(1987.9):
Donald Harper, "The Han Cosmic Board (Shih)," Early China, 4 (1978/79): 1-10, Christopher
Cullen, "Some Further Points on the Shih;' Early China, 6 (1980/81): 31-46, and Donald
Harper, "The Han Cosmic Board: A Response to Christopher Cullen," Early China, 6 (1980/81):
47-56. This circular
trigram arrangement also appears on turtle-shaped inkstones made of
earthenware. Such
objects have so far been dated to the Eastern Han period, yet on stylis
tic grounds a dating to the tenth century seems more convincing. For an example of such an
inkstone in the Minneapolis Institute of Art, see Ezekiel Schloss, Art of theHan (New York:
China Institute in America, 1979), no. 52. For a comparable excavated Tang example, see
Song Yanyan ^^^, Tangdai yishu ?ftSl? (Xi'an: Shaanxi renmin yishu chubanshe,
1991), no. 73.
24
For a tenth-century example from the tomb of a scholar named Du Jiyuan f?a?7C, who
died in 940 in Nanjing, see Wenwu ziliao congkan ?ft?^?fJ (1987.10): 160.
25
Eugene Y. Wang, "What do trigrams have to do with Buddhas? The Northern Liang
stupas as a hybrid spatial model," Res 35 (Spring 1999): 70-91.
157
<
a g|y ?
9& ?3K h^
tz A
^^? ^_ w
Fig. 5. A drawing of the lid of the epitaph of Du Jiyuan, who died in 940 (excavated in
Nanjing). From Wenwu ziliao congkan (1987.10): 160.

158 FRAN?OIS LOUIS
from the eighth to the twelfth centuries (fig. 6).26 It was also still
the norm at the Song court in 1118, when emperor Huizong ffi^
(r. 1100-1125) began building his famous garden in the northeast
of Kaifeng, which he named
Genyue fiai, "Northeast Marchmount"
or "Marchmount of Gen," in accordance with the directional sym
bolism of the houtian trigrams.27 Only from the Southern Song period
onwards is the xiantian trigram circle also found as an occasional
decorative symbol.28 Among the earliest archaeological evidence for
the xiantian circle is a mid-thirteenth-century silver cup discovered
in the tomb of a Southern Song scholar named Shi Shengzu j?$?fi
(1191-1274) in Quzhou, Zhejiang (fig. 7A). As attested in Shi's epi
taph, he was an official of rank five with a vivid interest in the
YijingP
Aside from the archaeological evidence, there is ample textual evi
dence for the beginning popularity of the xiantian trigram circle dur
ing the Southern Song period. Don Wyatt has recently discussed
some of the Song textual history of the xiantian diagrams and con
cluded that one of the earliest precise descriptions of the circular
diagram was written around 1200 by Lou Yue ft? (1137-1213),
while less explicit but nevertheless unambiguous references to the
image can also be found in Zhu Xi's Lou not men

writings.30 only
tions the directional positions of all the xiantian trigrams, he also
states that only "since the advent of our [Song] dynasty have there
26 see Ma Chengyuan
For examples, %%WiW>,ed., Zhongguo wenwu jinghua daquan. Qingtong
'
Juan rpmJCf?lE^JKik W?# (Beijing: Wenwu chubanshe, 1993), nos. 1293, 1313; Wu
Shuicun ÎtR#, Jiujiang chutu tongjing AtC?Hltffl^ (Beijing: Wenwu chubanshe, 1993),
nos. 72, 81, 82, 92; Cheng Changxin @^0f and Cheng Ruixiu ??^??j;^, Tongjing jianshang
IP?ll?ilt (Beijing: Beijing yanshan chubanshe, 1989), nos. 53, 56; Kong Xiangxing TL^J?
and Liu Yiman ??!j?^?, Zhongguo gudai tongjing 41 l? r?ff^$P??t (Beijing: Wenwu chubanshe,
1984), pp. 165-66.
27
Huizong took the geomantic advice of a Daoist practitioner named Liu Hunkang !*!)$?
J? (1035-1108). See James M. Hargett, "Huizong's Magic Marchmount: The Genyue
Pleasure Park of Kaifeng," Monumenta S?rica 38 (1988-89): 1-48 (on the naming see pp. 7-8).
28 on
The houtian trigrams remained popular designs throughout Chinese history, especially
ritual objects.
29 mu chutu qiwu" SrtC?S?'J'N
Quzhou shi wenguanhui, "Zhejiang Quzhou shi Nan-Song
fljf?*?ffi?iSiKj, Kaogu (1983.11): 1004-11.
30
Don J. Wyatt, The Recluse ofLoyang: Shao Yung and theMoral Evolution ofEarly Sung Thought
(Honolulu: University of Hawai'i Press, 1996), pp. 195-201. See also note 48 for Zhu Xi.
159
Fig. 6. A drawing of a bronze mirror with the houtian trigrams, carrying a date correspond
ing to 758. Diameter: 21.0 cm. Private collection, New Zealand. Courtesy of the owner.
Drawing by Zoi Kotitsa.

160
Fig. 7. The decoration on a silver cup excavated in 1974 in the tomb of Shi Shengzu
(1191-1274) inQuzhou, Zhejiang. From Kaogu (1983.11): 1005.
A: Outside wall: The xiantian hexagram circle with top band indicating the 64 hexagram com
binations.
B: Inside bottom: Diagram with the Five Elements.
C: Outside bottom: Taiji circle.

existed explanations of the xiantian and houtian."31 Wyatt percep

tively recognized that intellectuals like Lou and Zhu were well aware
of the relative novelty symbols, yet insisted on their
of the xiantian
primordial perfection and timelessness, which resulted in the para

doxical tension in their texts "between the simultaneous modernity
and antiquity of the diagrams."32
Undoubtedly, antiquity of the xiantian trigrams

the manufactured
was a means of promoting the new arrangement.33 The proponents
of the new xiantian system must have been acutely aware that this
sequence caused a profound conflict in the interpretation of the indi
vidual trigrams. After all, both the xiantian and the houtian arrange
ments the of nature in the same manner?as cir
represent cycles
cles the trigrams are directionally
in which positioned and grouped
into a.yin and a,yang section. Yet in each arrangement the same tri
grams possess contradictory meanings. Half of them have opposing
yin-yang valences and dui axe yin in the houtian cycle, hut yang in
{li
the xiantian cycle; kan and gen axe yang in the houtian cycle but yin in
the xiantian cycle), and all of them represent different directions.
Emphasizing the primordiality of the xiantian system was crucial for
its adherents to have it accepted as an alternative mode of inter
preting the signs of heaven. In fact, the adoption of the suggestive
terms "Before Heaven" and "After Heaven" to distinguish the two
trigram cycles was so effective that it eventually led to the anachro

nistic perception that the younger sequence was the older one, and
34
vice versa.
31
Lou Yue, Shoo Kangjieguanwu bian HPlUfStl^ll, translated after Wyatt, Recluse ofLoyang,
p. 199. Certainly by early Southern Song times the contrast between xiantian and houtian tri
grams was well established, as seen for example in juan 4 of Zhang Xingcheng's 5MfTJ5^ (jin
shi 1132) Huangji jingshi guanwu waibian yanyi M1S$?t??l^^lift'T?? (Shanghai: Shangwu
yinshuguan, 1935).
32
Wyatt, Recluse ofLoyang, p. 199.
33
Presumably unaware of the archaeological evidence, Wyatt concluded that not only the
xiantian but also the houtian diagram was novel to the Song philosophers, Wyatt, Recluse of
Loyang, p. 199.
34
A temporal reading of the terms was emphasized through the association of the xiantian
order with Fu Xi, and of the houtian order with the less ancient King Wen ?, the founder
of the Zhou Dynasty. The Yijing originally used the terms xiantian and houtian to refer to two
noble ways of behavior in relation to the knowledge gained from the oracle, not as two dif
ferent ways of divining. Wenyan, hexagram 1: qian; Wyatt, Recluse ofLoyang, pp. 194-5; Lynn,
Classic of Changes, p. 138.
162 FRAN?OIS LOUIS
As is well known, Chinese historiography traditionally credits the
philosopher and diviner Shao Yong S?Î (1012-1077) with the con
ception of the philosophical framework for the xiantian system.35
Twelfth-century scholars, however, including his son Shao Bowen
SPi?ffl (1057-1134), did not cite him as the inventor of the xiantian
circle. Instead they reported that Shao derived his insights from
already existing diagrams. These pre-existing, so-called xiantian dia
grams 9c^M were thought to have been transmitted to Shao Yong
through his teacher Li Zhicai ^?^f (d. 1045), from whom they were
traced back via Mu Xiu %fe (979-1032) and Zhong Fang #$ (d.
1014) to the Daoist monk Chen Tuan ffiff (872/895-989).36 Chen,
a heavily mythicized figure, takes the crucial position of a spiritual
inceptor of the Neo-Confucian Daoxue tradition, which credits him

with first making public certain elements of esoteric Daoist knowl
edge previously developed and transmitted within alchemist tradi
tions. He is said to have employed cosmological charts and images,
tu H, for meditative purposes?an activity which, by the tenth cen
tury, had of course become a well-established Daoist and Buddhist
practice. Chen's few known writings confirm that he was well ac
quainted with the processes of alchemy, and also that he concerned

himself with diagrams.37
We do not know what Chen Tuan's xiantian tu looked like, as none
of his discussions of such a diagram survives with illustrations.38 Song

35
The term xiantian may only have been assigned to Shao's cosmology by his immediate
followers, see Wyatt, Recluse ofLoyang, pp. 179ff. There is abundant literature on Shao Yong
and his xiantian theory. Aside from Wyatt's recent book, see e.g. Fung, A History of Chinese
Philosophy 2:451-74; Wing-tsit Chan, A Sourcebook in Chinese Philosophy (Princeton: Princeton
University Press, 1963), pp. 481-99; Anne D. Birdwhistell, Transition toNeo-Confucianism:

Shao Yung on Knowledge and Symbols of Reality (Stanford: Stanford University Press, 1989);
Smith and Wyatt, "Shao Yung and Number."
36
For quotes and literature on this lineage, see Fung, History of Chinese Philosophy, 2:440;
Livia Knaul, Leben und Legende des Ch'en T'uan (Frankfurt a.M./Bern: Peter Lang, 1981), p.
34; Birdwhistell, Transition toNeo-Confucianism, pp. 207-8; Smith and Wyatt, "Shao Yung
and Number," 110; Wyatt, Recluse of Loyang, p. 199. Regarding Chen's dates, Knaul, Leben
und Legende, p. 145, mentions various Daoist records according to which Chen was 118 sui
at his death, which would indicate a birth year of 872. Traditionally his birth year is given
as 895.
37
Knaul, Leben und Legende, pp. 35-6. Also see the special issue of Taoist Resources 2:1 (1990),
subtitled A Memorial to Chen Tuan, especially the contributions by Livia Kohn, "The Life of
Chen Tuan After the History of the Song," and "Chen Tuan in History and Legend," 1-7,
8-33.
38
Recent discussions of Chen Tuan's diagrams, such as the one by Li Yuanguo, are based
scholars were already quite vague in their descriptions, and most

presumably had never seen Chen's pictures. All we are told is that
his graphs used numbers, lines, and trigrams, and that they func
tioned without text.39 The transmission of the tu from the tenth to
the eleventh century evidently was a secretive affair, forming part
of an intimate master-student relationship that stood at the core of
finding individual enlightenment and personal divinatory insight.40

How long the xiantian tu had existed in secret as tools for the alche
mist-diviner-philosophers we do not know. Four generations reach
ing back to Chen Tuan would well seem possible. But whether Chen
was really the source of the diagrams will remain a mystery.
What we can say with certainty is that the circular houtian trigram
arrangement had become ever more popular since the mid-Tang
as demonstrated a marked increase, even the commer
period, by
cialization, of artifacts decorated

with it.41 In certain
of intel areas
lectual life in the Tang and Five Dynasties periods, moreover, tri
grams arranged in a circle started to become accepted for
symbols
rendering the waxing and waning of the moon?a use of the Yijing
line statements which had
previously been reserved to hexagrams.
Circular arrangements of twelve hexagrams a continuous
showing
waxing and waning of yin and yang are already recorded in some of
the early Yijing apocrypha, notably in the Yiwei zhilan tu Mf?fl?l^,
which is associated with the first century B.C. authors Meng Xi ??
H and Jing Fang j?f%, and was commented by Zheng Xuan W%
The first book to use trigrams in a circle to express
(127-200).42
a cycle of waxing and waning yin and yang was an alchemist trea
tise entitled Zhouyi cantongqi MI?3#|pI|g, which employs six of the
eight trigrams. The diagram in this treatise must be considered a
on traditional attributions only. Li, for example, does not discuss the historical authenticity
of the diagrams he ascribes to Chen Tuan, see Li Yuanguo: "Chen Tuan's Concepts of the
Great Ultimate," Taoist Resources 2:1 (1990): 32-53, esp. 46ff.
39
Zhu Bokun fci?M, Yixue zhexue shi J^g^? Huaxia
(Beijing: chubanshe, 1995),
2:11-12.
40
On Shao Yong's secretive treatment of his xiantian fortune-telling skills, seeWyatt, Recluse
ofLoyang, pp. 202-4.

41
See notes 23, 24, 25 and 26.
42
Bohn, Die Rezeption des Zhouyi, pp. 59-101; Fung, History of Chinese Philosophy, 2:106-118;
Michael Nylan, The Canon of Supreme Mystery, by Yang Xiong (Albany: SUN Y Press, 1993),
pp. 14-17.
164 FRAN?OIS LOUIS
particularly important predecessor of the xiantian circle, for Zhu Xi

explicitly refers to it when discussing the provenance of the xiantian
tu.*3 The book, attributed to a second-century Daoist, Wei Boyang
S?ES, has a very obscure textual history. The oldest extant version
was produced in 947 and commented by Peng Xiao ^^ a
(?-955),
high official active in Sichuan under the Later Shu dynasty
(934-965).44 Parts of the text certainly existed since the late Six
Dynasties, but much was emended during the Tang and Song, and
no standard edition was ever established. In the Cantongqi we find
a combination of Yijing principles and alchemist methods. In one
section the trigrams are correlated to the phases of the moon, most
likely in order to indicate different stages and

of firing intensities
during the alchemist
production of elixirs.45 The
cycle begins in the
east with zhen, and continues through dui, qian, xun, ana gen to kun
(fig. 8).46 These six trigrams are read from the bottom line up; in
this way yang lines increase until, after the moon is full on the fif
teenth day of the month, yin lines begin to increase from the bot
Fig. 8. The Phases of the Moon, from Yu Yan (1258-1314), Zhouyi cantongqi fahui (Wenyange
edition).
43
Zhu Xi, Zhuziyulei, 65.1617.
44
On the complex history of the Cantongqi, see Fabrizio Pregadio, Zhouyi cantong qi: Dal
Libro deiMutamenti all'Elisir dOro (Venezia: Cafoscarina, 1996); for an English summary, see
http://helios.unive.it/~dsao/pregadio/book.html (July 1996); Fukui Kojun, "A Study of Chou

I Tsân-Cung-chH," Acta Asi?tica 27 (1974): 19-32; Joseph Needham, Science and Civilisation in
China, vol. 5:3. Chemistry and Chemical Technology (Cambridge: Cambridge University Press,
1976), pp. 50-57.
45
Needham, Science and Civilisation in China, vol. 5.3:60-66.
46
Fung, History of Chinese Philosophy, vol. 2:426-30.
torn up. Li and kan, the trigrams traditionally the sun

representing
and the moon and hot and cold, are excluded because the cycle
explains their waxing and waning. According to Fung Yu-lan, they
are to be imagined within the circle.47
The Cantongqi also contains other concepts which foreshadow the
later directional xiantian trigrams, such as the emphatic use and link
ing of the trigram pairs qian-kun and li-kan. Li and kan here are
understood as a secondary display, eryong Zlffi, of qian and kun, rep
resenting the same creative principles.48 But while these pairs in the
Cantongqi are intended to explain the human body and various al
chemist and meditative practices,49 the Neo-Confucian reader was
able to discern in them the directional
xiantian patterns. In his anno
tated edition of the Cantongqi, Zhu Xi explicitly bases his interpre
tation of these two trigram pairs on the directional xiantian arrange
ment, even though the original text never mentions north and south,
nor left and
right.50
To be sure, medieval alchemist books such as the Cantongqi do
47
Ibid, 2:430. One might also argue that li and kan were excluded because their linear com
position did not neatly fit into the sequence of waxing and waning yin and yang lines.
48
In Zhu Xi's edition of the Cantongqi, the first chapter states for example: "Heaven and
earth establish the positions, and the changes take place between them. Heaven and earth
means qian and kun. Establishing the position means the position of a.yin and yang
arranging
pair. The Yijing speaks of kan and li. Kan and li are the second display o? qian and kun. The
lines of the second display do not have a definite position. They float to all six vacancies."
Wu Shuping Îf?ff, comp., Zhouyi cantongqi kaoyi; Zhouyi cantongqi fahui; Zhouyi cantongqi
fenzhangzhuA<g*|^#?, ?S#I^^??, JSA*Hlg#*!? (Tianjin: Tianjin guji
chubanshe, 1988), p. 11. The explanatory parts in this passage may well have been emen
dations by Peng Xiao or some other tenth or commentator.
eleventh-century
49
For a Daoist reading of the Cantongqi, see Yu Yan's fir?& (1258-1314) edition Zhouyi can
tongqi fahui j?|Ji^?|Rl?|cl?ff (preface 1284) in Zhengtong Daozang (hereafter Daozang; rpt.
Taipei: Yiwen yinshuguan, or the commentary
1962), pp. 625-27, by Chen Xianwei of 1234,
Daozang, vol. 628.
50
Zhu Xi, Zhouyi Cantongqi kaoyi, shangbian (first published 1197), 4a, inWu, comp., Zhouyi
cantongqi kaoyi, pp. 7-8. The Cantongqi begins like this: "Qian and kun are the gate and the
door to the changes, they are the father and mother of all trigrams. Kan and li assist
[them
like a] wall revolving around a nave on an upright axle tree. These four trigram pairs are
like a bag and a tube." Zhu Xi reinterprets this clearly three-dimensional illustration (surely
written with the human body in mind) in the following two-dimensional manner: "The posi
tion of qian and kun is at the top and at the bottom, while kan and li ascend and descend
between them. Therefore it is called 'changes.' In the xiantian position qian is south, kun north,
li east, and kan west. That is why this image has the form of a city wall and its rising and
descending sides are like the revolving wheels on the naves of a chariot axle?one at the bot
tom and one at the top."
166 FRAN?OIS LOUIS
not contain the xiantian sequence itself, but they clearly indicate that
certainly by the tenth
century the groundwork had been laid for the
xiantian system. Considering the new functions assigned to the tri
grams in alchemist literature and the popularity of the houtian circle
during that period, Shao Yong, or his

teacher Li Zhicai, must have
been able to conceive of the xiantian sequence without much diffi
culty. There are no indications in Five Dynasties or Song sources
that the formation of the xiantian trigram circle came about in con
junction with a dynamically divided yin-yang circle, comparable to

the one we find in the taiji diagram. The texts explain the genesis
of the xiantian circle entirely in terms of trigrams and numbers?
symbols whose code was recognized as basically arithmetic.
There is no evidence from the Song period that the xiantian
trigrams at the time had any connection to a yin-yang circle with
swirling halves. The mid-thirteenth-century silver cup from Shi
Shengzu's tomb, for instance, still combines the xiantian circle (fig.
7A) with the kind

o? yin-yang circle (fig. 7C) that had become pop
ular through Zhou Dunyi's MWC?& (1017-1073) classical taiji diagram
(fig. 9).51 That circle is static and does not reflect the waxing and
waning of yin and yang. Even Yuan scholars apparently did not feel
compelled to combine a dynamic circle with the xiantian
yin-yang
trigrams, as is suggested by a short treatise entitled Bao yihan sanmi
jue }g??SHSifc, written by the Quanzhen ^:H master Jin Yueyan
as and his assis
&?^ (also known Jin Pengtou &?M, 1276-1336)
tant, the famous and Daoist
scholar-painter practitioner Huang
Gongwang ??M ( 1269-1354).52 Here the trigram circle is discussed

as a sequence from yang to yin (fig. 10). The trigrams are numbered,
51
The taiji circle is chased on the outside bottom of the cup. The inside bottom of the cup
shows the diagram of the relation of the Five Elements, also taken from Zhou Dunyi's taiji
tu. See note 29. For thirteenth-century textual sources where the xiantian circle is discussed
as containing taiji, visualized by Zhou's taiji circle, see Robinet, "Place and Meaning of the
Notion of Taiji," p. 399.
52
Bao yihan sanmijue in Daozang, 321:1a. The Bao yihan sanmijue is a Daoist text that com
bines explanations of neidan practices with traditional cosmological theories based on the Yijing.
Like the Cantongqi, it appealed to classically educated scholars, and was in fact conceived by
two scholars of the Quanzhen sect who rigorously promoted a syncretistic philosophy. Cf.
A. John Hay, "Huang Kung-wang's "Dwelling in the Fu-ch'un Mountains": The Dimensions
of a Landscape" (Ph.D. diss., Princeton University, 1978), pp. 11-12; Caroline Gyss
Vermande, La vie et l' uvre de Huang Gongwang (1269-1354) (Paris: Coll?ge de France, 1984),
pp. 35-51.
167
?fcto&
Fig. 9. Taiji Diagram, redrawn after Zhou Dunyi (1017-1073), from Joseph Needham, Science
and Civilisation in China, vol. 2, sections 8-18 (Cambridge: Cambridge University Press, 1956),
p. 461.
168
*
? IHHHHHlHHHHHHHBiBiBBMHHBBBBMBMaiBVBaBa
Fig. 10. "Fuxi's First-Drawn Prior to Heaven Diagram," from Jin Yueyan and Huang Gong
wang, Baoyihan sanmijue, in Zhengtong Daozang (Taipei: Yiwen yinshuguan, 1962), 321, la.
and the dynamic of change they express is indicated in their center

with an angular s-line. The text explicitly states that this line
describes the changes.53 Nevertheless, the line marks only six instead
of eight corners, and thus does not exactly replicate the dynamic
stages described the trigrams. Assuming
by that the fifteenth-cen
tury Daozang edition, which makes this book available, did not
gravely distort the fourteenth-century Yuan original, one cannot but
suspect that the authors simply were unaware of the dynamically
divided yin-yang circle; otherwise they would certainly have em
ployed it instead of the unconventional and unsystematic line.
THE TAIJI CIRCLE
Ming and early Qing authors were quite mystified when it came
to determining why early depictions of the xiantian trigram circle
never included the taiji circle with the characteristic, interlocking
yin-yang swirls. None of the major compendia, in which the taiji tu
could be expected, included it. It appears neither in the Xingli daquan
ttH^C?: of 1415, the standard
Ming compilation of orthodox texts
of, and commentaries on, Song Daoxue nor in the
philosophy,54
Daoist Canon published in 1444 and 1445. Its earliest appearance
was discovered by Hu Wei at the end of the seventeenth century in
an etymological called Liushu writ
study benyi Aitîl (figs. 11, 12)
ten by the Zhejiang scholar Zhao Huiqian f?jgiR (1351-1395).55
That book was first printed in the Hongwu era (1368-1399) and
has four prefaces by early Ming literati. Two of the prefaces are
dated; one was written by Bao Xun jS?tfe in 1380, one Zhao him
by
self in 1378.56 No Hongwu edition, however, appears to have sur
vived. The earliest extant versions date to 1517 and 1520 respec
tively, and thus are over a century younger than the original.57
53
Bao yihan sanmijue, lb.
54
My thanks to Peter Ditmanson for pointing out this compilation as a potential source.
55
Hu Wei, Yitu mingbian, 79-80. Hu's book appeared in 1706. For Zhao Huiqian's biog
raphy, see Edward L. Dreyer, "Chao Ch'ien," in L. Carrington Goodrich and Fang Zhao
ying, eds., Dictionary ofMing Biography, 1368-1644 (New York: Columbia University Press,
1976), 1:124-125.
56
The other prefaces are by Lin You $;? and by Xu Yikui #!?Mt
(1356-1402) (1318-c.
1400).
57
The 1517 edition was issued by Shao Fen B?jt, the 1520 edition by Hu Donggao ?fjjfC
170
?*? **.. >"

^%?>?:.wii^
M -nf ?fe ^4 TV ^
s?fe
JC?
1.___
Fig. 11. "River Diagram of the Spontaneous Process of Heaven-and-Earth," from Zhao
Huiqian (1351-1395), Liushu benyi (1520 Hu Donggan edition).

?) :? ?K i) j? *.
9
^
I
*
#
i
?&*t
*M? ^ ? i)
f &* .&.
1) ?feAr
Fig. 12. "River Diagram of the Spontaneous Process of Heaven-and-Earth," from
Huiqian (1351-1395), Liushu benyi (1751 Wenyange edition).

172 FRAN?OIS LOUIS
taiji tu appears near

The the beginning of the Liushu benyi in an
illustrated introductory section(tukao H#) on
the origin of written
Chinese characters.58 In the tradition of classical Yijing interpreta
tion, Zhao explains that in the earliest forms
of writing the trigrams
and the characters for the natural elements such as fire, water, and
wind were written the same way and therefore must have the same
origin.59 Linking his pal?ographie observations to the xiantian tri

gram charts of the Song, Zhao correlated the origin of the Chinese
script with the origin of cosmological symbols and diagrams, and
fittingly began his section of tables and charts with our taiji dia
gram. His illustration is entitled tiandiziran hetu 55ift?^MH, "River
Diagram of the Spontaneous Process of Heaven-and-Earth." To

Zhao Huiqian this was a primal image of natural creation compa
rable to one of the divinely revealed charts of mythical times, the
famed hetu MH, or River Diagram.60 After associating his diagram
to such divine images he provides the earliest known explanation of
its provenance: "As for this diagram, it has long been said that Cai
Yuanding received it from a recluse in Shu [Sichuan]. He kept it
^. Of both these editions several copies are preserved. In addition there are copies of aWanli
era edition by Yang Junkuang ?fffl? of 1610, see Shen Jin (Chun Shum) tfc??, Meiguo Hafo
Daxue Hafo-Yanjing tushuguan zhongwen shanben shuzhi HlllPpW^^Bp^ÂM^f?^X???
H^ (Shanghai: Shanghai cishu chubanshe, 1999), pp. 77-78. Wang Zhongmin ??S men
tions an extant Hongwu edition in his Zhongguo shanbenshu tiyao ^llsi?ir^?lfiiil (Shanghai:
Shanghai guji chubanshe, 1983), p. 57. Wang describes a copy he saw at the library of Beijing
University; but since

the published catalogues of Beijing University do not list such an edi
tion, Wang appears to have been mistaken, see Beijing daxue tushuguan, ed., Beijing daxue
tushuguan cang guji shanben shumu ^bA^CiPH^:t?SK"?'ffS^^@ (Beijmg: Beijing daxue
chubanshe, 1999), p. 43. I am indebted to Philip Hu for his advice during my search for
early editions of the Liushu benyi.

58
Zhao Huiqian, Liushu benyi, "Tukao," la.
59
For a fine summary of the origins of this etymological tradition, see Lewis, Writing and
Authority in Early China,pp. 272-8.

60
The highly
complex issue of the hetu being considered the prototype of all cosmological
diagrams (and therefore also the ancestor of the taiji tu) does not need to be addressed here.
For introductory discussions of the hetu, see Michael Saso, "What is the Ho-tu?" History of
Religions 17:2 (1977), 399-416; John B. Henderson, "Chinese Cosmographical Thought: The
High Intellectual Tradition," in The History of Cartography, vol. 2:2, Cartography in theTraditional
East and Southeast Asian Societies, edited by J. B. Harley and D. Woodward (Chicago: University
on the
of Chicago Press, 1994), pp. 213fi?. For analyses o? hetu numerology and speculations
hetu's connection to the dynamic taiji circle see Schuyler Cammann, "Some Early Chinese
Symbols of Duality,"History of Religions 24:3 (1985): 234, 252-53; Lars Berglund, The Secret
ofLuoShu. Numerology in Chinese Art and Architecture (Lund: Lunds Universitet, 1990), pp. 174,
189, 383ff.
secret and did not transmit it, so that even Master Zhu [Xi] did not
see it. Now I have obtained it from Mr. Chen Bofu and take plea
sure in its perfection. It has the Great Ultimate containing yin and
yang, ana yin and yang containing the Eight Trigrams. The subtlety
of spontaneous process truly is the origin of written characters of all
ages and the pivot of creation. How divine it is!"
Zhao here discusses the image not as an illustrative diagram in a
book, but rather as a rare physical object, an old painted chart which
fell into the hands of a nowadays obscure collector Chen and which
supposedly had a history reaching back to the scholar Cai Yuanding

W;7?1j? (1135-1198) and beyond. Cai was one of Zhu Xi's close
friends and most outstanding pupils, and also one of the main com
mentators of Shao Yong's basic work, the Huangji jingshi SUIS tit.
Moreover, he is traditionally considered the person responsible for
compiling the charts that now form the first two introductory chap
ters of a heavily augmented and commented later edition of Shao's
book.61 a taiji diagram with the dynamically
But divided circle never
appeared in any known Huangji jingshi edition. Zhao Huiqian's com

ment in fact suggests that he was not aware of any substantial lit
erary history of this image, as he describes it as an antique chart
which, although known to a small circle of local scholars, had not
yet been widely publicized. Quite possibly because of the relative
novelty of this diagram, Zhao published it with precise annotations
on how to understand its design.62 To Zhao, who felt that the cen
tral circle expressed perfectly the meaning of the xiantian trigram
circle, it must have seemed most improbable that Cai Yuanding
should have been unaware of it. And so we once again encounter
the secret transmission theoryalready employed in explanations of
the xiantian tu?this time placed into the twelfth and thirteenth cen
turies.
Zhao, in reconstructing the provenance of the picture, may have

recollected parts of a story that had already been circulating for a
61
Shao Yong, Huangji jingshi shujie jSfiffl&tFffP (Siku quanshu edition, 1779), preface A,
12bff.
62
Zhao's annotations anticipate later Ming and Qing readings, as discussed above in the
section on the iconography of the taiji diagram, see note 7 and fig. 1. Zhao's comments behind
the trigrams read (figs. 11, 12): "Qian is the place of pure yang; dui the place of two parts yang
and one partan, li has jwz in between yang," etc.
174 FRAN?OIS LOUIS
century or so. Around the beginning of the Yuan period, Yuan Jue
Utt makes mention of three charts in the possession of
(1266-1327)
the Southern Song official Xie Zhongzhi ?8#? (1226-1289).63
Those charts could also be traced back to Sichuan, a
supposedly
Chen Tuan some of his younger
region where had spent years.64
Yuan Jue explains that Zhu Xi, eager to find out what Chen Tuan's
charts really looked like, had Cai Yuanding search for them in
Sichuan. Cai returned with three charts, but none of
supposedly
them were made The charts are said to have remained in
public.
the Cai family collection for several
generations, until they were
away and eventually came into the possession of Xie Zhongzhi.
given
Hu Wei seems to have been the first to explicitly connect this story
to the image in the Liushu benyi, but whether one of these three thir
teenth-century images actually looked like the one in the Liushu benyi
we cannot know.65
In thelate eighteenth century the young Zhang Huiyan (1761

wrote an astute comment on Hu Wei's findings and rejected
1802)
the Cai Yuanding transmission theory altogether for lack of evi
dence. Instead, he concluded that diagrams like the one in the Liushu
benyi originated from the cosmological speculations of early Ming

scholars.66 Unfortunately, his little book on this subject,
insightful
Yitu tiaobian WjMf?ff?, has only recently received wider recognition.67
Zhang's drastic revision of the idea that twelfth- and thirteenth-cen
tury intellectuals, especially those steeped in Shao Yong's xiantian

theories, must have been familiar with the taiji diagram, can be sup
examining more closely the pictorial nature of taiji dia
ported by
grams in Song cosmological discourse. Taiji depictions of unam
Song date can be differentiated from the yin-yang circle of

biguous
on a conceptual but also on a perceptual
the Liushu benyi not only
level.
63 vol.
Yuan
Jue, Qingrong jushiji fj|?gd:?fi (Beijing: Zhonghua shuju, 1952), 7,
Hu Wei, Yitu 78. Xie is Xie see Zongxi,
21.373-74; mingbian, p. Zhongzhi Fangde, Huang
Song Yuan xue'an, pp. 2845-46.
64 see Knaul, Leben und Legende des Ch'en T'uan,
On the data for Chen Tuan in Sichuan,
pp. 22, 153.

65 Yixue zhexue shi, 2:12.
Hu, Yitu mingbian, pp. 73-80; Zhu Bokun,
66
Zhang Huiyan, Yitu tiaobian, 23b.
67 Yitu tiaobian (Shanghai:
Zhang Shanwen, Zhouyi cidian, pp. 99-100; Zhang Huiyan,
Shanghai guji chubanshe, 1995).
Song cosmological diagrams employ only a very limited vocabu

lary of visual signs consisting of writing, trigrams, numerical signs,
circles, boxes, and lines linking individual parts of a chart. These
are to content, and aesthetic con
signs arranged solely according
siderations play no role at all. The taiji is variously represented by
a simple circle, a circle with a central dot, numerical signs (fig. 13),68
concentric circles or a circle split into the semi-circular
(fig. 14),
versions of the trigrams li (left) and kan (right) respectively (fig. 9),
as in the classical taiji diagram of Zhou Dunyi In all
(1017-1073).69
these cases
the pictorial signs are abbreviated, devoid of represen
tational elements, and so general that they usually cannot be com
prehended without textual explanation. The only significant recog

nizable symbols aside from text are numbers and the binary line
statements that compose the trigrams. Conceptually, the Song taiji
diagrams thus functioned within the same framework we have
already encountered for the xiantian diagrams: their main informa

tion was defined textually in relation to the Yijing and through the
arithmetic qualities of the trigrams or of numbers.
The yin-yang circle of the taiji tu, on the other hand, communi
cates on an essentially geometrical basis. Its two curved, interlocking
geometric shapes depict a rotating, self-creating cycle of comple
mentary opposites, of mutually dependant entities whose beginnings
are the other's The interaction of dual opposites in this
endings.70
particular design is self-evident; the cosmological symbolism of yin
and yang?and especially of their waxing and waning?is not. This
symbolism requires explanation such as is provided the xiantian
by
68
Compare the examples, including a numerical taiji tu based on Chen Tuan's explana
tions in the Longtu xu ?H/?' in volume 143 of Wuqiu beizhai Yijing jicheng ^?R??ifrf J?$?Hl
J5Jc(Taipei: Chengwen chubanshe, 1975). See also the works by the Yuan author Zhang Li
3?3?, Zhouyi tushuo waibian Wi^jW^Wt^Wi, shang 3, Daozang, vol. 69; and his Da Yi xiangshu
goushen tu Jk<?jMW(.$Q*WM, shang Iff. in Daozang, vol. 70; for reprints see also Wuqiu beizhai
Yijing jicheng, vol. 143.

69
On Zhou's well-known diagram, which had found doctrinal acceptance among Confucian
scholars, see Michael Lackner, "Die Verplanung des Denkens am Beispiel der fu," in
Lebenswelt und Weltanschauung im fr?hzeitlichen China, edited by Helwig Schmidt-Glintzer
(Stuttgart: Franz Steiner, 1990), pp. 133-56; Fung, History of Chinese Philosophy, 2:434-51;
Joseph Needham: Science and Civilisation in China, vol. 2, sections 8-18, pp. 460-68.
70
For thoughts on such images and their visualization in various cultures, see Nathan
Sivin's contribution in Joseph Needham, Science and Civilisation in China, Vol. 5:4 (Cambridge:
Cambridge University Press, 1980), pp. 374-88.

X
mm
VA
to
~ ?E.* * J*~ r
?ihAil* 4-?. *
*, J. j? * * ^ 1
ft ^&^Jt J
a.
Fig. 13. (left) Tizy? Diagram, from Liu Mu (1011-1064), Yi shujunyin tu (1681, Tongzhitang edition, rpt. i
Fig. 14. (right) Taiji Diagram, probably of Tang origin. From Cheng Daye, Chengshi moyuan (1606; rpt. Bei
trigrams or additional text. Seen

individually, the two halves are
basically meaningless shapes, likened by many Chinese commenta

tors to fish (or commas inWestern In contrast, the only
literature).
other taiji circle that depicts yin and yang, the one made famous by
Zhou Dunyi (fig. 9), uses traditional, precisely defined symbols of
yin and yang: the trigrams li and kan.71
Curiously, Zhou Dunyi's taiji circle fails

visually to represent
the
dynamics of yin and yang as described in the textual
commentary
appended to the diagram.72 Zhou's static design simply serves as a
reminder that yin and yang emerge from the taiji, but does not
attempt to illustrate the interaction of one taking shape as a result
of the characteristics of the other. Zhou communicates this creative
process verbally. To the left of the circle he writes: "yang?motion";
to the right "yin?repose." The concept o? yin "being the root" of
yang and vice versa is explained more
elaborately in the commen
tary. One wonders why Zhou Dunyi chose a static taiji symbol, if
the dynamic taiji circle could have expressed his ideas much more
emphatically. Was it because he wasof a dynamic

not aware
taiji
circle? Possibly. Or did he know of the latter, but considered it inap
propriate as a diagrammatic sign? Most likely not. For while some
of the images he was familiar with may have resembled the design
of the yin-yang circle found in our taiji tu, no Song scholar would
71
Li and kan, symbolizing fire and water, are the only complementary trigrams in the hou
tian circle and it may well be because of their positions in that circle that they came to rep
resent the juxtaposed yang andern sections in other circular yin-yang depictions. The li-kan
circle has been traced back to the Tang period. It appears without a white circle in the cen
ter in one of the texts of Zongmi tk$j (780-841), the fifth patriarch of the Huayan IjijSt
School of Buddhism. Zongmi used it to denote alayavijnana, the store-consciousness in which
the true and the false, i.e. enlightened and phenomenal existence are blended. The alayavi
jnana produces all phenomenal existence, yet is never distinct or apart from the true, enlight
ened existence. Cf. Lackner, "Die Verplanung des Denkens," 142-43; on the alayavijnana and
other ideas of dualism current in medieval Chan and Huayan Buddhism, see Alfonso Verdu,
Dialectical Aspects in Buddhist Thought. Studies in Sino-Jap??ese Mahayana Idealism (Lawrence:
Center for East Asian Studies, University of Kansas, 1974).
72
InWing-tsit Chan's translation: "The Great Ultimate through movement generates yang.
When its activity reaches its limit, it becomes tranquil. Through tranquility the Great
Ultimate generates yin. When tranquility reaches its limit, activity begins again. So, move
ment and tranquility alternate and become the root of each other, giving rise to the distinc
tion o? yin and yang, and the two modes are thus established." Chan, A Sourcebook in Chinese
Philosophy, p. 463. See also Fung, History of Chinese Philosophy, p. 435; and Julia Ching, The
Religious Thought of Chu Hsi (Oxford: Oxford University Press, 2000), p. 22.
178 FRAN?OIS LOUIS
have read them as symbols of the taiji, or o? yin and yang splitting
from the taiji as the two creative modes.
DEPICTIONS OF COSMIC ENERGY
Between the ninth and fourteenth centuries, images of circles

divided into two halves by an s-line or an inverted s-line were by
no means abstractions
of cosmological concepts, comparable to the
symbolic graphs in used

Song diagrammatic discourse. On the con
trary, they were representational depictions of the concrete physi

cal world and formed part of larger pictorial contexts. They appear
only being small, they are hidden away within
rarely; the surround
ing imagery. No common design elements other than some form of
s-shaped dividing line links these circular motifs (figs. 15, 16, 17).
And features such as two central dotsor the contrasting colors and
particular shape of the yin-yang swirls?iconographie characteristics

of the taiji circle seen inMing and Qing illustrations?are conspic
uously absent. In fact, until about the twelfth century, a common
iconography was not established through the design within the cir
cle but through additions outside it. Most commonly, flaming halos
(and sometimes also mythical animals) identify these circular designs
as blazing spheres and discs. Clearly, the specific design of such
was determined an artisan to aesthetic con
spheres by according
siderations, not by an intellectual to clarify cosmological specula
tions.
Early images of flaming spheres divided into two (and sometimes
three) curved sections appear in northern China, most prominently

on ritual utensils from the Liao empire (907-1125) (fig. 16); their
meaning is, however, still largely unstudied.74 The earliest datable
example I have been able to trace is on the lid of a silver reliquary
73
The history of designs that anticipated such images of spheres and gems cannot be exam
ined here. Suffice it to say that dual, dynamically opposed pictorial entities which are not
enclosed by a circular frame can be traced back to antiquity. In many cases these designs
can be interpreted on a symbolic level within the larger system of correlative thought.
74
Fran?ois Louis, "Shaping Symbols of Privilege: Precious Metals and the Early Liao
Aristocracy," Journal of Sung-Yuan Studies (forthcoming). For a rare datable Song example of
1085 on the forehead of a mythical stone beast along the spirit road of the mausoleum of
emperor Shenzong (1048-1085), see Ann Paludan, The Chinese Spirit Road: The Classical Tra
dition of Stone Tomb Statuary (New Haven: Yale University Press, 1991), pp. 148-49.
179
Fig. 15. Decoration on the lid of a silver reliquary casket excavated in 1987 at the Famensi,
Fufeng, Shaanxi. Tang, 871-873. Length: 20 cm. From Wenwu (1988.10): 21.
180
f \ 3 ? 1 f <***
Fig. 16. A composite drawing by the author of the silver funerary crown of Liao prince Xiao
Shucheng (d. 1017/18). Excavated in 1986 in the tomb of the Princess of Chen in Naiman
Banner, Zhelimu League, Inner Mongolia.
Fig. 17. A drawing of a lacquered box excavated in 1979 in Zhenjiang, Jiangsu. Southern
Song, thirteenth century. From Wenwu ziliao congkan (1987.10): 186.

manufactured for the Tang emperor between 871 and 874 to en

shrine a finger bone of Buddha ?akyamuni (fig. 15).75 The gem is
depicted within slender streaks of clouds with two dragons swirling
around it. Thedragons' bodies and postures are identical, their
heads thrown back, snouts gaping at the sphere between them. The
combination of the antithetical elements water and fire, depicted as
clouds and halo, and the polarized positioning of the dragon pair
leave little doubt that we are presented here with the theme of cre
ation through the interaction of opposite entities.76 Yin and yang meet
as a roaring clash of elemental powers, their transformational force
regulated only by the ornamental symmetry of the design. The cos
mological significance of this imagery on the lid is further empha

sized by the iconography on the sides of the reliquary, where the
four cardinal directions are indicated through the depiction of the
four Heavenly Kings (tianwang ^1). It is a Buddhist to
cosmology,
be sure; Vaisravana, the King of the North, resides on the front
panel of the casket.

Having recognized the cosmological symbolism inherent in the
lid's main imagery, can we
assume that the bipartite design of the
gem, because it mirrors the inverted and thus dynamic juxtaposi
tion of the dragons, itself represents the yin-yang dualism?77 The
answer is yes, but it requires explanation. Given the specific pur
pose of the reliquary, we may identify the flaming gem as a repre
sentation of the cintamani, a wish-fulfilling gem variously said to be
obtained from the dragon-king of the sea, a dragon-fish (makara), a
bird-king some other mythical animal, or from the relics
(garuda),
of the Buddha.78 Because we know from archaeological evidence
75
This casket formed part of a set of eight containers, stored one within the other, dis
covered in 1987 in the crypt of the Famen Temple in Fufeng, Shaanxi. The casket was the
outermost metal case. A sandalwood case originally enclosed it, but had decayed and could
not be recovered, see Shaanxi sheng Famensi kaogudui, "Fufeng Famensi ta Tangdai digong
fajue jianbao"ftaf?n#*Jf ftJftS???*, Wenwu (1988.10): pp. 20-21.
76
On the^m ana yang symbolism of dragons, see Jean-Pierre Di?ny, Le symbolisme du dragon
dans la Chine antique (Paris: Coll?ge de France, 1987).
77 as a design choice and its
On the idea of inverted symmetry perception as "dynamic,"
see Ernest Gombrich, The Sense of Order: A Study in thePsychology ofDecorative Art (Ithaca: Cornell
University Press, 1979), pp. 138ff.

78
Cf. Mochizuki Shink? M^fi?, ed., Bukky? taijiten AS^Cl?A (Tokyo, 1958-63), pp.
4132-33; Ding Fubao TUfl?, Foxue da cidian fM^lftA (Beijing: Wenwu chubanshe, 1991),
p. 550.
182 FRAN?OIS LOUIS
that the liturgical aspects of worshipping this particular Buddha relic

were entrusted to chief priests of the Esoteric School,79 the various
types of cintamani recorded in Esoteric texts of the eighth and ninth
century are particularly relevant. In those texts one can indeed find
descriptions of gems considered to be dualistic embodiments of

Buddha's transcendence, their wish-fulfilling creative power being
explained by their bipartite nature. The eminent Japanese patriarch
K?kai 3?#J (774-835), for example, who introduced the Chinese eso
teric teachings to Japan, discussed the view that the cintamani con
sists of two antithetical entities. He corrects a traditional notion that
a true wish-granting gem ismade
from dragon's liver and phoenix's
brain, and explains that the duality in fact reflects the laws of spon
taneous process inherent in the That?gata's body.80 It is thus quite
possible that the priest who conceived of the decorative scheme on
the Famensi reliquary recognized the notion of yin and yang in the
cintamani. In this way, the gem could be interpreted as a magic sym
bol not only of Buddha's central position in the universe, but also
of his transformational power, which transcends the cycle of life and
death and promises spiritual salvation in a future existence.
A similarcosmological symbolism was most likely intended in
Liao depictions of the flaming sphere, which are usually accompa
nied by a pair of dragons or phoenixes (fig. 16). Yet, given the rep
resentational nature of these gems, their interactive halves should
be understood as concrete rather than abstract symbols of yin and
yang. The most obvious choice for such representational symbols in
a flaming sphere would have been the contrasting elements water
and fire. Support for this proposed reading is found, for example,
in a flaming gem with a red and a blue or green half that appears
79
Han Wei $$f?, "Famensi digong Tangdai sui zhenshen yiwuzhang kao" ^P^^f J??a J??
ftÂ#?C#|l|?#, Wenwu (1991.5): 27-37; Fran?ois Louis, Die Goldschmiede der Tang- und
Song-Zeit (Bern: Peter Lang, 1999), pp. 140-50.

80
Goyuigo fEP?er, T. 77:2431, 413a. One of Amoghavajra's (705-774) sutras similarly rec
ognizes the interaction between yin and yang in two types of cintamani and expresses this in
terms of the polarity of water and fire. Amoghavajra states that cintamani produced by peo
ple are thought to embody the virtue of fire, while those which come from dragon palaces
are endowed with the virtue of water. Ruyi baozhu zhuanlun mimi xianshen chengfo jinlun zhou
wangjingmM%ftmmm^M?&B&f?VR5.&, T.19:961, 332a.
on top of a sedan chair in a Liao tomb mural in Kulun painted

around 1080.81
By the thirteenth century the choice to divide a circle into two
swirls had become a more widespread design principle, now being
used also in southern China. A circular lacquered beauty case with
two interlocking inserts discovered in a thirteenth-century Southern
Song tomb (M6) in Zhenjiang, Jiangsu, may exemplify the grad
ual abstraction of this design principle from representational depic
tions into simple geometric forms (fig. 17). Meanwhile the tradition
of depicting concentrated cosmic
energy continued?be it in images
of flaming balls or swirls of clouds. Chen Rong's g|? (jinshi 1235,
d. after 1262) famous handscroll of Nine Dragons Afila in the
Museum of Fine Arts, Boston (fig. 18), for example, presents both
these images in the exact center of the composition. A great vortex
of intertwining dark and light clouds is unambiguously intended to
depict nature's creative material force (qi |R). The contemporary

colophons make several references to it. Right next to this whirl a
young, "inexperienced" dragon has managed to grab a smooth gem
with a flame shooting forth that is associated with this center of
on
energy.
It seems
likely that by the fourteenth century, images of ethereal
whirls and bipolar gems had become stock depictions of organic cos
mic energy. The presence of such imagery in the corpus of identi
fiable visual forms appears to have created the necessary pre-con
ditions for the conception of a two-dimensional abstract yin-yang
circle with intertwining black and white halves which could be com
bined with the xiantian
trigrams. The Siku quanshu edition of the
Liushu benyi, in fact, preserves a design of the tiandi ziran hetu (fig.
12) which appears much more like a whirl than a precise transla
tion of the yin-yang valence expressed through the xiantian trigrams.
81
Cf. Wang Jianqun :Et?f!? and Chen Xiangwei Bjxfflftj, Kuhn Liaodai bihua mu J?l?iiS
ftHS? (Beijing: Wenwu chubanshe, 1989), color plate 2:2. See also the swirling elements
emanating from the beaks of two juxtaposed phoenixes on a set of jade
plaques datable to
the tenth century, in Ren?-Yvon Lefebvre dArgenc?, Treasures from theShanghai Museum, 6000
Years of Chinese Art (San Francisco: Asian Art Museum, 1984), no. 63.
82
Wu Tong, Tales from theLand ofDragons: 1000 Years of Chinese Painting (Boston: Museum
of Fine Arts, 1997), no. 92, pp. 90-95, 197-200.
Fig. 18. A detail from the handscroU Nine Dragons by Chen Rong. Southern Song, dated 1244. Ink and
cm. Museum of Fine Arts, Boston (Francis Gardner Curtis Fund; 17.1697). ?2002 Museum of Fine Arts,
THE ICONIC TAIJI CIRCLE
Perhaps the most striking conclusion to be drawn from the above

brief survey of early circular designs divided into two interlocking
halves is that circles with two colored halves and two antithetical
dots do not seem
to appear before the Ming period. The design in
fact begins to proliferate in decorative as well as literary contexts
only in the sixteenth century. As this section will show, it was only
during the later half of the Ming period that our specific taiji circle
became truly iconic, that is, widely recognized as a cosmologically
meaningful design. The new icon was the result of the circular
whirling image of creative energy merging with the xiantian trigram

circle into one cosmological diagram. As a result of this combina
tion generic images of interlocking swirls were modified into the spe
cific abstraction of a circle with a clearly defined appearance that
would eventually retain its abstract cosmological symbolism even
without the trigrams.
Judging from the Liushu benyi, it was during the fourteenth cen
tury that a small number of intellectuals began to abandon concerns

over the diagrammatic usefulness of an organically designed yin-yang
circle.However, because this circle evolved from a pictorial tradi
tion, and because its geometrically complex design contrasted
sharply with the numerically encoded trigrams, its transfer into dia
grammatic discourse hinged upon a systematizing and rationalizing

of the design. This need for explanation is amply; demonstrated by
the annotations in the Liushu benyi and later illustrations on how to
understand the relationship between the circle's design and the xian
tian trigrams. The subjectivity of the design, particularly the way the
swirls intertwine at the center of the circle, must have
recog been
nized as mathematically inaccurate and
quite different from the
more precise symbolic codes of trigrams and numbers. Indeed, the
history of the taiji image in the Ming to a
period is large extent dom
inated by scholars trying to come to terms with this To
inaccuracy.
remedy the subjectivity of the circle's design, some scholars of the
late sixteenth and early seventeenth centuries began to reconstruct
it, basing themselves on the increasingly familiar methods of ge
ometry, as introduced most influentially by the Italian missionary
Matteo Ricci (1552-1610). The result of this new geometric
186 FRAN?OIS LOUIS
approach was a distinction between what may be labeled an "objec

tively" and a subjectively designed version of the taiji tu.
Such late Ming approaches are best exemplified in the Tushu bian
Bli?&i, a massive, encyclopedic collection of diagrams and writings
drawn from over two hundred earlier publications.83 Zhang Huang
3ggf (1527-1608) compiled the book between 1562 and 1585 in an
attempt to sort out the true Confucian illustrative materials from
those of Daoist and Buddhist provenance. Because of financial con
straints, he never published his seven-thousand-page compendium;
only after his death was one of his disciples able to afford the cost
of having the wood blocks carved, and the book could finally be
printed for the first time in 1613.84
Zhang Huang included two images of the dynamically divided

circle surrounded by the xiantian trigrams. In one illustration the
circle is subjectively
designed and the image is called "Ancient taiji
Diagram" tu
(gu taiji S^ffiH) (fig. 19); in the other the circle is con
structed with a sensitivity to geometric principles in order to match
more precisely the yin-yang flux indicated the
by trigrams (fig. 20).
This latter image, which we shall discuss first, is called "Diagram
of the Trigrams Drawn Prior to Heaven" (xiantian huagua tu 5fe^4
$M?). The circle here is divided into four sectors and the intertwining
yin and yang halves are aligned along the diagonals, resulting in a
sliding join rather than a hooked one as seen in the gu taiji tu (fig.
19).85 The quarters of the circle represent the four xiang, the second
stage in the binary exponential process which links yin and yang to
the eight trigrams. The four xiang are called greater or lesser yin and
yang respectively (cf. fig. 3). But while this manipulation of the cir
cle helped to clarify to some extent the joining of the two halves in
the center, the splitting o? yin and yang is neither visually convinc
nor do the of the four to be
ing, positions resulting xiang appear
related to those of the trigrams. The sectors are merely superim
83
The first Western author to point out the Tushu bian in connection with the history of
the taiji tu was Joseph Needham: "The Institute's Symbol," in Biologist. Journal of the Institute
of Biology 24, (1977): 71-72.

84
Goodrich and Fang, Dictionary ofMing Biography, vol. 1:83.
85
The bad quality of the print is due to a flaw in the 1613 printing block?a wide scratch
near the trigram kan on the right. The same flaws seen in the Columbia University version
also appear in the copy reprinted in 1971 by Chengwen chubanshe in Taipei.
187
& w it i $^*l,
?A* ??
ir '
' &$#
]-
1
Fig. 19. "Ancient Taiji Diagram," from Zhang Huang, Tushu bian (1623), 1:1a. Courtesy of
C.V. Starr East Asian Library, Columbia University.
188
Fig. 20. "Diagram of the Trigrams Drawn Prior to Heaven," from Zhang Huang, Tushu bian
(1623), 2:5a. Courtesy of C.V. Starr East Asian Library, Columbia University.
posed over the circular design, and information about the four xiang
must be conveyed textually.86
In seventeenth-century adaptations of this xiantian image the cen
tral circle is sometimes divided by four lines into eight sectors (fig.
These lines function as guides for creating two sine curves with
21).
continuously decreasing radius, helping render the
trigram code
more accurately in the curvilinear mode. In the example illustrated
here from the yin axid yang halves join along the dui-gen diag
1688,
onal.87 Eventually, as knowledge of Western geometry increased,
their outline would become a true sine curve that uses the qian-kun
or north-south axis as a guide for alignment. In a detailed article
explaining the astronomical and mathematical dimensions of circu

lar taiji tu Li Shicheng ^ffc?c, a Nanjing physics professor, has re
cently presented the trigonometric decoding of this most accurate
diagram (fig. 22).88

Nowadays, one also finds the diagram with the central circle's
dividing curve simply constructed from semi-circles. However, this

construction is geometrically unrelated to the xiantian trigrams,
because the curve
does not continually decrease in radius.89 While
this design has appeared in decorative contexts since the seventeenth
century, its use as a literary diagram is more recent. It has mainly
been employed by twentieth-century Western scholars such as
Marcel Granet
(1884-1940) to explain Chinese cosmological thought
In China, its diagrammatic use is a phenomenon of the
(fig. 23).
late twentieth century: it occurs mostly in connection with the prac
tice of qigong M#J and traditional Chinese medicine.90
86
Zhang Huang, Tushu bian (1623), 2:5a. It is theoretically possible to recognize the four
xiang in the four design elements that make up the central circle: the two fish shapes and the
two central dots. But such an interpretation was apparently not yet envisioned around 1600.
Even a hundred years later, Hu Wei still understands the central dots in relation to the tri
grams li and kan and interprets them as the sun and the moon which had traditionally been
represented by these two trigrams, see Hu Wei, Yitu mingbian, p. 81.
87
Published in the Yijing Lai zhu tujie, where it belonged to one of the additions to Lai
Zhide's 5fcffl^ (1525-1604) original work entitled Zhouyi jizhu, first published in 1599. See
note 7.
88
Li Shicheng ^tfcSt, "Lun taiji tu de xingcheng ji qi yu gu tianwen guancha de guanxi"
^*?HW^^??H*^X?^WIIS*, Dongnan wenhua jfCgfXffc 85/86 (1991/3,4):
17-19.
89
For see C.A.S. Williams, Outlines of Chinese Symbolism and Art Motives
examples, (rpt. of
third revised edition of 1941, Rutland: Turtle, 1974), pp. 121, 150, 188.
90
Li Shicheng mentions a 1990 publication by an agronomist and qigong adherent named
190
ll> ? X fu
&w
m
i
4h
% in a M * *pv*v??.> & ^ a ?n4M&?|

- & *
? tf? ?9& & 'N,*. tf *^"v F& &
Fig. 21. "Diagram of the Trigrams Drawn Prior to Heaven," from Lai Zhide, Yijing Laizhu
tujie (1688; rpt. Chengdu: Ba Shu shushe, 1988).
191
-â?a-*if{?LH^wsi?$fL?
mm
'Wi-? 1*
*8 * w< ? Mii:U
BAR SB?? ff
./ iff" ? \*r
?*
*
V
?ffHT ES. !*JL*" :
/?3 =
2 32
'?-32Xjr
a =? , r^p??nor
p=M (??=0*2*) 16
P?*-2P,f>oO<*<0-eo)-l-*,2-r2=O
16 16 32" 16"
i99o. ti.se.
Fig. 22. Geometrically constructed xiantian diagram, by Li Shicheng, 1990. From Dongnan
wenhua (1991.3/4): 18.
192 FRAN?OIS LOUIS
Fig. 23. Taiji Diagram, from Marcel Granet, La pens?e chinoise (Paris: La Renaissance du
Livre, 1934), p. 280.
It is significant that Zhang Huang placed the xiantian diagram

(fig. 20) only into the second juan of his compendium, while he saw
the so-called "Ancient Taiji Diagram" fit for the very beginning as
the appropriate first image (fig. 19). The importance that latter
image had gained by the early seventeenth century as a universal
symbol of creation, as a primal tu, could hardly have been stated
more explicitly. In Zhang's "Ancient the
Taiji Diagram" trigrams
appear as small characters, dwarfed by a giant central circle with
sharply interlocking black and white halves. The trigrams are still
an essential part of the of this image?relevant to main
iconography
tain the directional positioning, the clockwise rotation, and the two
dots near the center
of the taiji circle?but they are visually subor
dinate. Moreover, the analytical rigor we saw in the depiction of
the xiantian diagram's central circle is here sacrificed in favor of a
blatantly subjective design. The prominent central circleempha
sizes recognizability over mathematical exactness. As a result, it
subsumes the full load of associations of the entire diagram into one
self-explanatory pictorial symbol.

Three lengthy commentaries celebrate this image. Rather than
treating it like a diagram that is part of an argument, these texts
Han Yongxian ft^cff as one of the first scholarly attempts in China to explain the geome
try of the semi-circle-based design adhering to Neo-Confucian terminology. See Li Shicheng,
"Lun taiji tu de xingcheng," 19-22. For another example, see Wang Rongkui EE?Ê et al.
eds., Zhouyi baodian Jr|Jg Jfft (Hohhot: Neimenggu daxue chubanshe, 1998), vol. 4:3722.
discuss the image as a symbol, as an icon that can be interpreted

in manifold ways because it functions independently of text. It is
described as a symbol
of the way to physical as well
as moral per
fection, as an abstraction of the principle workings of nature and
the beginning of all creation. It is also presented as a symbol for the
origin of culture, as an image that creates basic insights and thus

symbolizes all scholarly pursuit. Two hundred years earlier Zhao
Huiqian, in order to claim this image for the Confucian elite, had
to connect it to the Song literati's philosophical discourse with dia
grams. By Zhang Huang's time, it could be separated from the tra
dition of intellectual diagrams and be perceived as a timeless sym
bol. There was no longer a need to explain its uncertain provenance
with theories of secret transmission. Rather, it was accepted as a
suggestive image which evokes, elucidates, and embodies funda

mental cosmological and moral ideas described in the Yijing and
later reinterpreted and
popularized by Song philosophers. One of
the commentaries bluntly states that "it is unknown who painted
or under which dynasty this came about. Because it
[this image],
has been circulating for a long time, it is known as the Ancient Taiji
Diagram. When looking at the opening passage of the Xici com
mentary in the Yijing, this diagram and that text elucidate each
other. The passages in the Shuogua commentary 'heaven and earth
determine the position' and the resulting intermingling are explained
by this picture."91
The visual distinction in Zhang Huang's Tushu bian between the
subjective "Ancient Taiji Diagram" and the objective xiantian dia
gram is paralleled by the differentiation of the origin of the two
images. The subjective circle is perceived as a popular symbol of
unclear provenance while the analytical circle is traced back as part
of the xiantian tu to famous members of the Chinese intelligentsia.
Around 1600 scholars were
increasingly thus aware of the popular
nature of this taiji circle and some, such as Zhang Huang, attempted
to separate it from traditional Confucian illustrative discourse by
treating it as a primary symbol rather than as a diagram. Others,
like Lai Zhide, felt the imprecision of this taiji tu to be too unsatis
factory and devised more rationally designed versions that became
91
Tushu bian, 1:4a.
194
V$ $k A
1.?
JUS ?? rt. v5)

?.\fk ?3?
4
C3
-ft:
'V
?s 4LiD xa
AI asi%Vi te ?a T,
?9 r?Sf! A
'ir
4L'
mm
Fig. 24. "Taiji River Diagram," from Lai Zhide, Laizhu Yijing tujie (1599; rpt., Zhanghua
Shi: Yichun chubanshe, 1969).
very influential in later Confucian discourse on cosmology (fig.

24).92
CONCLUSIONS
This paper set out to clarify the provenance of the taiji tu by treat
ing its two main components, the xiantian trigram circle and the taiji
circle with its interlocking yin axid yang swirls, as separate pictorial
entities. This resulted in the following significant conclusions. The
xiantian trigram circle is not an ancient chart, as is often implied by
its association to Fuxi the Yijing, but was conceived
and during the
Northern Song period as a result of cosmological and alchemical
speculations with numbers and trigrams. It was successfully pro

moted during the twelfth century by intellectuals of the Neo-Con
fucian Daoxue school, most effectively Zhu Xi, whose writings se
cured the ultimate acceptance of this trigram arrangement. There
is no evidence throughout the Song period that the circular xiantian
arrangement was in any way connected to a taiji circle with inter
locking yin axid yang swirls. The earliest known text that illustrates
a diagrammatic synthesis of these two entities is the Liushu benyi, an
etymological treatise written in the 1370s, but apparently only avail

able now in editions of the early sixteenth century. The book claims
that the actual, physical chart it illustrated dated from at least the
twelfth century and had passed secretly through of
the collections
various noted Song and Yuan intellectuals; the image itself, how
ever, was understood as having been invented in archaic times by
Fuxi. Thus a myth created during the twelfth century to promote
the xiantian was here for a more com
trigram sequence appropriated
plex diagram.
Ever since the
taiji diagram first appeared in the early Ming
period, scholars
discussed the yin-yang circle not as a later addition
to the xiantian trigram circle, but as an intrinsic part of it. To Ming
and early Qing scholars the image simply presented the informa
tion encoded by the trigrams in an alternative visual mode. A look
92
Lai Zhide, Yijing Lai zhu tujie (Chengdu: Ba Shu shushe, 1989), p. 484; Schulz, "Lai
Chih-te," pp. 141-68. Ironically, later interpreters of Lai's work did not make his distinc
tion and hence included the taiji diagram into his book, see note 7.
196 FRAN?OIS LOUIS
at the pre-Ming history of the yin-yang circle, however, reveals that

a abstract with two swirls of
diagrammatically image complemen
tary colors and two central dots had not gained any currency.
Circular designs with an implied symbolism nevertheless
yin-yang
existed beforethe fourteenth century as part of Buddhist or Daoist
cosmologies in the form of flaming spheres and atmospheric whirls
depicting concentrated, creative energy. Such representational cos
mological imagery can be traced back to the ninth century, but it

seems to have gained wide recognition only by the thirteenth cen
tury. These early designs should not be confused with the taiji cir
cle used since the Ming period in diagrammatic discourse. They
should instead
be considered as predecessors which paved the way
for the conception of a graphic, abstract taiji circle designed like the
images of swirling energy. The specific iconography of two anti

thetical dots near the center and
the contrasting colors of this circle
were determined by the positions of the surrounding xiantian tri
grams. With the acceptance of the taiji diagram as part of the schol
arly discourse on cosmology over the course of the Ming period
came the recognition of the dynamically divided circle as an iconic
symbol. Since the late Ming period design variations of this circle
have proliferated in contexts far beyond the confines of Confucian
ideology. Today, designs alluding to the taiji circle abound, while
their symbolism tends to be far removed from anything we en
countered in Ming and early Qing writings.93 Nevertheless, the
popularity of the design would be unthinkable without Ming dia
grammatic discourse on cosmology that created and justified the
prototype.
93
For an assembly of illustrations of nineteenth and twentieth century uses of the design,
see Li Shicheng, "Lun taiji tu de xingcheng," 9.

The Genesis of An Icon The Taiji Diagram

Uploaded by

Copyright:

Available Formats

The Genesis of An Icon The Taiji Diagram

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

The Genesis of An Icon The Taiji Diagram

Uploaded by

Copyright:

Available Formats

The Genesis of an Icon:

The Taiji Diagram's

taiji tu ;MSH, or "Diagram

ing yang within yin and yin within yang respectively.

JL& jfc& 4fc >>t

diagram somehow vanished, for no references to it could be found

history of cosmological diagrams, subscribed to the secret trans

taiji tu is therefore generally considered to have been passed down

secretly by certain Daoist adepts and Neo-Confucian thinkers of the

period before becoming more widely known in the fourteenth

Huiyan 5S?H however, who felt that secret trans

Yijing, of the Daoist Huang-Lao School, and of alchemical practice

gram's origin. Instead it will view the traditional explanations as

primeval unit instead of recognizing its main constituent elements?

Categorically different in their pictorial qualities, these two graphic

signs, it is argued here, communicate on different cognitive levels

increasing popularity of the image toward the end of the Ming

THE ICONOGRAPHY OF THE TAIJI TU

specific iconographie elements shows that the design of the central

Fig. 2. The circular xiantian arrangement of the trigrams (Fuxi sequence).

In the taiji tu the trigrams are arranged according to the so-called

qian represents pure yang, being made up of three solid lines. It

xiantian trigrams are then read counterclockwise following their grad

ing the yin trigrams arranged in opposition to the yang trigrams on

taposition is also visualized in the design of the swirls of the central

taiji circle. from the top to the bottom, while yin de

conceptual and numerical manner, the central design visualizes in

respondence between the swirls and the trigrams

sumably because Chinese commentators considered the xiantian tri

1998), pp. 138-39.

(New York: SUNY Press, 1999), pp. 281, 471 n. 135.

THE XIANTIAN TRIGRAM CIRCLE

Theyin-yang valence of the xiantian trigrams, and thus the specific

tai yin shao yang shao yin tai yang

Fig. 3. The creation of the xiantian trigram sequence.

yang to yin. This system of binary permutations is justified by a

Appended Statements," one of the early Yijing commentaries, most

"good fortune and misfortune constitute the great field of action,"

ing and yang continually waning.

systematic known way of organizing the trigrams. To leading

sponding to the original conception of Fu Xi.16 Southern Song

concept in the world principle.

Changes (New York and Toronto: Ballantine Books, 1996), p. 198.

Neo-Confucian writings thus promoted it as a reflection of the most

trigrams alternate with each other, so that to enumerate what has

Among the several trigram sequences described in the Shuogua

University Press, 1990), pp. 169-205.

of Chinese Philosophy, 2:102-6.

trigrams, though not paired as opposites, are still ordered around

only directional trigram sequence found on archaeological material

Nanjing). From Wenwu ziliao congkan (1987.10): 160.

taph, he was an official of rank five with a vivid interest in the

image can also be found in Zhu Xi's Lou not men

Drawing by Zoi Kotitsa.

B: Inside bottom: Diagram with the Five Elements.

C: Outside bottom: Taiji circle.

existed explanations of the xiantian and houtian."31 Wyatt percep

primordial perfection and timelessness, which resulted in the para

Undoubtedly, antiquity of the xiantian trigrams